Prospective Control and Analyzing Data from an Ecological Perspective with Andrew Wilson

Welcome back to finding the edge podcast. I'm Garrett, boy, am joined with Garrett Baker and Robert Fry. And today we have on Andrew Wilson, Andrew Wilson is a professor at Leeds Beckett University. And we brought him on to discuss a tweet that we had seen where he was talking about how a lot of sports science articles. The way that I interpret it was just like sports scientist. Just collect a bunch of data and report it with no real consideration. Ation as to why.

And I think that is something that I noticed for myself, like, it stood out to me, because I was, I'm the type of person that just wants to collect a ton of data so that it's there. So later on, I can go back and I can question it. I can learn from because if you don't collect data, it's you're having to basically remember based upon your recollection recollection of the past as to what occurred and draw a conclusions and inferences from that. And so it's me better to have data.

Because sometimes you can complete days and events and all that sort of stuff. And so I wanted to bring and Ron to kind of discuss more of like the intent and the thought behind that. That question also because Andrew has done a ton of research in ecological Dynamics as well as just knows the field really well. And so I want to bring him on to talk about perspective, control and some of the other work that he has done. And yeah. Yeah, to continue on with

Garrett's point. You know, we talked we talked quite a bit about understanding like certain statistical analyses and perspectives and like one important thing we talked about was like, the data that we collect is there. We just have to ask my questions and being able to apply it being in, want to apply it and then apply it with a rationality ecological Dynamics. Yeah, it's a good to go after that. I thought I thought this conversation was actually Really, really fascinating.

Like as we're going live, I took well over page and notes and then notes often notes to be completely honest. So that's that's how good I think this this conversation really was and that idea of like asking better questions.

I think just kept coming up and I think it's something to really think about is actually asked this question recently about, where do I think like all this data that were collecting in the pro pro orbs, where it should go or where it may go and to me, it's like, it should start, we should start asking better questions of what to do with the data.

Or start having a theory or a lens to look at the data through versus like having all this data and then trying to trying to put it on people or looking at this very confined thing and then trying to say that skill to me that's that's missing pieces of it. And I think Andrew goes to a pretty well and how he's approaching his collection of data and how he's looking at things is really interesting and it's a different way than most scientists are at this point.

So it's really cool. I think to we discuss the importance of having a theory because the theory is going to shape how you see the data and we all have a framework by which we are looking at things and sometimes we're not even aware of the lens or the framework or the theories that were actually trying to approach the data with.

But a lot of times to we've heard how a lot of professional organizations, they click on collect a ton of data and they don't know what to do with it. And so, Of the things that we discuss in today's episode is okay. What do we do with data? Especially if we use a theory like ecological Dynamics, how do we then begin to examine the data through an ecological lens? And so, we begin to discuss this topic and we're hoping to have Andrew back on to explore this

idea. Further of, how can we apply ecological, Dynamics with data? And that's been one of the big themes of this podcast. And so, I hope you guys enjoy. Today's podcast with Andrew Wilson.

Also, if you want to learn more about ecological Dynamics, one of our favorite topics here and you want something that's a little bit more systematic, and of like what are the basics of ecological Dynamics. And some of the other motor learning theories that are out there and how ecological Dynamics is different, make sure to check out emergence is course, the movement Academy intro.

It it is a college style course that has that It's a, it's a 12-week course with 12 weeks worth of lecture along with a discussion board and then there might be some other little projects that they have. You kind of put together in a discussion board, they'll be other people that you can bounce ideas off of they'll also be an emergence team. Member to that will also be there to be able to answer your

questions. And so, this is a really good starting place for somebody who's looking for something that's more in depth, but also interactive. And so, I highly recommend that you guys. Check out the movement Academy intro. You can use the code edge, 7 to get 7% off on this course. And highly recommend it. Make sure to use the code edge 7 to support this podcast and also to support emergence.

To Jump Right In. I guess one of the one of the struggles that I have is like, people still are quite attached to the traditional models of thinking and then, but the way that I think about it is that well this is the main thing that they interact with, right? Like this is usually the first thing that they learn and so it's kind of like you know the chick that comes out of the egg. The first thing that it sees becomes as mother type of a thing.

And so, once you've gone down a line of thinking and you've actually started to build You could say Your Castle becomes hard to shift to something else and so I have to acknowledge that and then go. Okay. Now we have to build a convincing argument I guess you could say like for why this other path is better. And so, I guess that's, that's where I guess I find my struggle when when talking to people from this approach or about an

ecological. Approach and then perspective control the challenge actually becomes. I don't feel like I understand like I can articulate well enough and simply enough like why why it is that this this way is is better. And the other way actually has a whole bunch of embedded assumptions that make it actually difficult to practically work out and so I guess, Oh the other element to is that it's a little bit newer and so like there's the argument that there's less evidence for

it, right? And so like well and then how does it how does it practically work out you know like and I think the practicalities of it and how it scales is it's very subtle because if you're using a predictive model and then versus using an ecological approach and doing things for like a perspective control perspective, How you design your, practice designs initially, or maybe in your most ideal State, actually look very similar because both need information.

And so, that's where I think the struggle is delineating something because I think the underlying theories matter a lot, especially once you start getting into The when you start actually having to deal with the real world, that is very constrained, you're going to go for things that are a little bit

easier. And so if you if you think that you can just go with a predictive model and you can break things apart because you just need to give the the brain the information and then it can just put it, put it together later. Then you when you're when you're under constraints like you won, You would just go with it.

The path of least resistance. That's that's at least my thought of like why you're under pinning Theory matters because your decisions under stress are going to change or under constraint is going are going to change. Yes. There's quite a few things going on in there. Did you have any references to where to go? So know what stands out most to you and your personal experience? I mean. So the first thing is this idea that people tend to come to the ecological approach.

Second is really common, it's really funny. Actually, we talked about a lot as a conference last year, for the first time in ages, reading out with his friends, in the Netherlands and like we were all people who came to the ecological approach via a different route. So we all had to do the thing of unlearning.

A whole bunch of stuff we can because we encountered it and decided that we would kind of resonated to it and thought it was a good idea, but yeah, there really is just sort of that fact of having to go through that process of unlearning and and it kind of mirrors the history of it. So there's that history of ecological approach being the Plucky Underdog and being bad replacement. Being the alternative and so we're the ones that had to make the case to kind of Justify Your Existence.

But here's the thing, right? That argument at least in the scientific literature that justification for our existence and saying, look at actually it's a viable approach. It happened 30 40 years ago, right? So the field itself has been productively taking along developing an Evidence base for a lots of these different things for a long time. So it's that weird situation where actually like a lot of those fights.

Academically happened. So long ago and got an all the reasons for saying that we were at least a viable alternative. They're all their heart of the problem is that they are often all there in Michael survey's writing. God bless him and everything that he does. But he's not, like you have to work to get into to get to the meat of what Turvy is talking about. It's always worth it because they're in Quebec Clarity in that position that he brings to his good.

But there is an issue of accessibility, right? And the right there is an issue of some of those things. So I mean there's been a lot of root interesting moves come up in the last law, so there's been an increasingly large number of useful and accessible books Rob Graves. Obviously doing gangbusters on this run. He's got a couple out now. And his podcast is obviously been about doing this for a long time, but there's a there's now

so terribly. He's writing his lectures on perception, but was just came out and then there's the lectures on action, which is coming out. And that's kind of pitched as kind of a graduate level class on theories of perception and action so good, but quite detailed. But there's also now we're going to be our have their undergraduate scaled textbook as came out last year and actually I've just been reading through it and it's good, it's great, right.

It's Engages with all the material and tries to convey some difficult ideas. But it engages with them and it's pitched at sort of an undergraduate level to try and make it an accessible topic for undergraduate psychology students to encounter in a sensible kind of way.

So like I think you're right I think it's a really real issue but I also think it's the Vincent great strides in the last few years about Ram creating accessible access to information about the ecological approach and how it works and what it is and why it's interesting Yeah.

And honestly a big chunk of that has been driven by the, by the interest of applying it for ecological Dynamics to sports, because when you're doing it sort of within the academic field, you can get away for a bit longer with just being still a bit jargony orbit technical. But as soon as you try and apply it, then people very sensibly, come back and say, wait a minute, how do I actually turn this into a concrete thing? So that pushback I think has been really useful.

I think it's been a really useful sort of spur. Yeah, so that's part of it and then I guess so one of the other things you were talking about and they're the sort of There's this issue of your underlying theory of how all these things work and how important that is. And that's one that certainly one of the big discussion points that comes out in within the bounds of sport Science and coaching and things like that

people to talk about. Well, how how much do I have to care about the details of the academic theory that underpins this in order to inform my practice, and should I care? And does it matter? And you were certainly coming up saying. Yeah, probably does matter and I would strongly Endorse that right? I'm not a coach. I'm a scientist. I'm an academic. So of course, I think our stuff is important but actually I actually think it is because I

think it matters. I think you had a nice phrase in there. You said you just let what was it your decision making Under Pressure kind of flips to what you know. Right. And and if what you know or if kind of the core of of the motivation of all your coaching techniques is from a more traditional approach as soon as you hit and you put your trying to develop an ecological one. As soon as you hit any kind of bear. Mayor or any kind of complication which can't quite get it to work.

There is just that instinct to go back to what, you know, you can at least do independent of how much, you know, whether it works or not. Yeah. So I think that's, that's really interesting.

I mean for for myself I've seen people with minimal knowledge actually like just having a rudimentary understanding of the concepts actually do quite a lot with with that and then just even to if you look at some of the most effective coaches if you look if you begin to like begin to investigate a little bit more into like what they do, oftentimes it's very ecological, it might not be 100% ecological and I think that's where Maybe I

could see people then using that as evidence for why they could, you know, sort of sit the fence and use both sides, but I just think of it from the standpoint of if you want to be able to improve what you're doing. Because oftentimes like if if somebody becomes successful, a lot of times people just want to copy, whatever it is that they're doing. And So eventually like because even in I think in pro baseball

once analytics became popular. But he had an analytics team, so now it becomes a question of how you use analytics, like the ones who are have a better system for using analytics will then therefore have the advantage. And so like that's where to me all these things come into play of.

Like, if you want to reproduce your success, you have to have an understanding for why that underlying success happens because one of the that's triply important for the ecological approach, of course, because it's very much, not a kind of a cookbook recipe, kind of scenario for your training situations. It's very much around.

And the design of constraints is very much a, an interactive process where you have to be fairly firmly embedded in your training environment with your athletes co.design. And with them, it's kind of baked into the idea. There isn't a Playbook of constraints that you can apply in order to get specific behaviors to emerge. It's unfortunately more annoying than that, you can get stuff

reliably, but you can't. Yeah. And so having an understanding of where your where those things are coming from in the motivation for them. The theoretical kind of framework that's generating that as an idea. That's that's how you generate the next idea for trying to get the thing that you're trying to get to work to work. Absolutely? well, in I I feel like in some ways I did this to myself. I was so hyped up on all this stuff that I didn't get enough sleep. So thoughts are going to get

dropped and the middle of this. But I think Baker. Because or actually Robert because maybe maybe we should go on the the data question in terms of collecting a ton of data because I think this is, this is Heart of it.

Like to me, I think it's important from just the like your Tweet on being careful, I guess, you know like just collecting a ton of data without having any consideration of Y and Reporting it. I think that's the thing because like, when I've done stuff and I have a like, I take something from a paper or whatever. And then I want to collect data.

I don't do what I do with a lot of like the sports science findings, whether you're talking about something like motifs and the the the ideal workload Management number that's sweet spot for acute to Chronic workload.

Yeah. I've seen criticism of it and so hence where it allows you ask the question of like well maybe I can actually ratchet that up or maybe I don't have to follow that as rigidly and like, We if we if we're to do more research, what we actually find the reality be somewhere different.

And so to me, I don't, I want to collect data first and a bunch of data before I say exactly and Report, like this is what, how we're actually going to use it because if I, if I work with my assumptions first I will, it will exclude some things that will actually work. Not that you don't Have an intention and guide where you're going and where you're looking.

But in the same way of Rob, kind of talks about it, like keeping your affordances open like not coming to too much of a conclusion too fast like keeping yourself. Open to the data especially when it's new, right? And that I think is where the the nuances is like if it's if it's something that's established and we've seen this recreated pattern and we want to go this direction, then I think

it makes sense. But I also think When we use analytics, something that we've talked about before on this podcast is good Hearts law of like when a measure becomes a Target it ceases to be a useful measure. And so like trying to balance those two things and so I guess that's where I think what really resonated to me was there's this this need or this, this felt need to actually have to report the data initially. To show value right instead of just absorbing in an observing. So okay.

There's a few things that occurred to me on that front of me, do a little framing here. So, The tweet that kind of sets us off and got us chatting about having a chat was me, just being a bit grumpy on Twitter and which I was just sort of. Identifying I've been reading a lot of papers from Sports Science journals like you for a couple of projects that I'm working on and this is a thing that I've noticed over the years and lots of different sort of context as well.

And it really sticks out to me as a psychologist, prove it for a variety of reasons, Sports Science, academic, papers, collect and report, a lot of data especially biomechanics kind of work, right? They just they reported on the numbers. Very little discussion about why they measured those numbers or what exactly. They think those numbers mean which and it sticks out to me like a sore thumb because I'm busy trying to figure out what all my numbers mean? When I measure stuff, like, what

am I measuring the right thing? And then my measuring am I actually successfully measuring something? That's telling me something. Interesting about the thing that I'm studying is a big question, it's a key question. And part of the problem like data collection is easy bit, right? But and so that's the problem is that you have to kind of use. Is it in service of a good question? And good questions.

Don't actually come from data, good questions, come from Theory. And, you know, there's lots of philosophy of science on this and there's lots of ways of people. There's lots of ways of sort of thinking that this, but let me give you a kind of a concrete example, right? So in Psychology actually so you know how much those just saying as a psychologist.

I have a theory mostly as an ecological psychologist that I have strong opinions about Theory because in Psychology, there's also that of a tendency to collect data and interpret it in a fairly free form kind of way, it's whatever. Mesa statistics, do it cetera. And there isn't sort of a single kind of cord. There are core theoretical commitments and cognitive psychology but not to the point where they're generating specific hypotheses or the times

under the big. It's just kind of a constraint on how you talk about your data and in science. What that gets you is a replication crisis, right? That's what it got psychology. The replication crisis is the problem where people were running experiments, finding statistically significant results and not, Not sufficiently interrogating. The relationship between the study that they run on the data that they were getting and asking sort of really they weren't asking important enough

questions. And one of the one of the results was that you end up with a bunch of bunch of results that we just statistically accidents. People got lucky there was also some, you know, there's some dodgy sort of practices that were underpinning this but a big chunk of this and a lot of people talk about this as kind of a lack of theory.

People weren't asking hypothesis-driven questions, they were just collecting data and running experiment, running the next experiment running the next experiment. And so and sports science at journals. I kind of see this quite a bit where it's just I collected a bunch of data and I analyzed it and I found some differences and I told you about it and then I'm just sitting there going. Yeah. But what does any of it mean?

Why did you run that day? What why did you think that the velocity of that joint angle was the thing that was going to be the most important? Well, it was the thing that you measured, right? That's that's not a good reason. So there's that and that element of that frustration and then they It shows up, I think that the connection to analytics and pro sports as well. I think is really interesting as

well. One of the things that's really intrigued to me. Lately is discovering just how much data there actually is in sports Amino. And you lots of sports baseball and all that sort of stuff. And then the lead of this and collecting data. But now it's really, really easy to get movement kinematics out of video, right? And, you know, automatically digitizing an entire, you know, like English football game to digitize in the position of the players all over, over the pit.

For example, you know, first of all, it's been used to create all kinds of cool displays and you can see people passing and when they're doing play-by-play and all that sort of stuff and that's kind of cool. But it's really easy. Now to generate those numbers automatically, like the computational tools exist to do that digitizing and nice and nice and robustly. And the and the thing about big data is that you can find kind of anything you want and a big

data set. And you only find things that are interesting. If you use that data and service of answering good questions, and data doesn't tell you a good questions to ask data data stages where you go with your good question. But you have to have that good question ahead of time and scientifically that good question comes from Theory which means that if you want to be asking good questions, this is why it's important.

Don't understand your theory and to have a theory that you are explicitly talking and thinking about as your source of your questions because actually as a scientist it's the source of my questions to go. Run my next experiment as the coach, the theory is the source of your questions to try and figure out what the next round of constraints are that you place to try and make your training session, do more of what you want to do. For example has to be a place

where those come from, right? And so that's kind of the problem and it's a I think it's a problem that occurs at a few different levels of few different scales. Like I said, shows up in slightly different ways. It shows up, shows up as a potential replication crisis and potential disaster academically, and it shows up as just kind of people just reporting stuff and you know, just there's so much, there's so many numbers around. So the question is, what do you

want to do with them? And it's really easy to ask stupid questions. It's really easy, really easy to ask God questions, I that's and and so that's why it's really useful to have a good theory and some explicit commitments to try and say no. Okay, the reason why I'm going to ask that question rather than this other question is because this is the one that makes most of the squid. This is a better question because of my various commitments and it's okay to rule things out.

That's the other thing. A good theory does having a good framework having a good understanding of what it is. You're trying to tackle. It's okay to rule things out ahead of time and say, I'm not going to ask that question because that question doesn't make any sense. Someone else wants to ask it because they think it makes sense, then that can be their problem and that's fine.

And if they find something interesting man, then you have to have a conversation but It's okay to rule things out as no I'm not going to do that and that's another thing that crops up I think in sports and sports science literature doesn't and coaching and of the coaches and interactive people don't like ruling things out. People don't like saying I'm no I'm not going to I'm not going to try and do that for sort of theoretical reasons over because that doesn't suit my constraint.

People seem to get very nervous about that whereas scientifically you know being able to do that is there's a huge help. All right, it keeps you focused to keeps you grounded? It keeps you pursuing a particular kind of coherent account. So yeah, so there's a lot going on there and there's anything in particular to pick up on ya and to kinda pick up on that. You brought up a lot of great points.

So I think one, one thing that is a big struggle right now in the analytics world is we treat a lot of things as prescriptive and what I mean by that is, you know, we just report data and Say, here it is but we don't, we don't do the second Dairy thing of, you know, what to do with it.

We just say, you know, hey X is y and I think part of it too is, you know, there with all these numbers, like, there's a big lack of knowledge and / understanding with it because we don't, we have this big constraint called time especially in the coaching. Yeah. Because once you, No in colleges that say maybe have certain form of analytics, but don't have the staff behind it to be able to analyze you two numbers on a full-time role.

They're going to cut corners and with that being said, then it becomes an issue of it. Get good Hearts law but then it also becomes an issue. So like in the statistics world this happens a lot. I'm sure you've run across this a lot quite a bit and Drew you know, a lot of Annotations will try to fit an r-squared model. Try to just force variables to be like, oh my arse words did, so that means this, but then they fail realize, obviously the classic quote, no correlation

does not imply causation. So I think if given more time or at least in educational aspect to understand like what these Analytics mean and how we can apply them would help a lot. But yeah right now we're just dealing with a big constraint which is simply foot time given you know the amount of resources available. Yeah. Hi. I'll often say never send an engineer to do a psychologist job.

Definitely, never seen a psychologist to do an engineer's job, but don't do it the other way around either. And by that, I mean there's a lot of really clever, a sophisticated technically minded, engineering people who have built the capabilities to accumulate, all this data really quickly and efficiently, and effectively, and reliably. And that's amazing that they were the wrong people to come up with things to do with the data, right?

Because what they tend to do is they tend to generate numbers based on things that they can quickly generate from the data set as opposed to trying to find the data, what the data can tell them about something that they actually want to know about. They don't know how to ask behavioral questions.

So my favorite example of this right now is a football raisin in English sort of European football soccer is, you know, you can get these data sets very quickly, you know, these clubs just send off the video footage of all their training sessions. For example, a company company will digitize it reduce a bunch of metrics and just send them back spreadsheet, right? And all of those metrics. Acts are things like, you know, number of times person, X touched ball number of times

person acts passed the ball. Although, you know, some interesting numbers, right? Why the thing is what and those, then become the numbers that people are using the data for What why are those two numbers will? Those are the numbers that are easy to pull out some engineer. Would look at go. Yes, I can count the number of times somebody touched the ball. That's the thing I can do and I know the algorithm for that, that's fine.

But actually, the more interesting thing is not, how many times did the person pass the ball? What you really want to know is how many times did they pass the ball? When passing the ball was a good option? How many times did they not pass the ball? When passing the ball? Was not a good option and then the various combinations, right? And that gets you more complicated instead of just asking. You know, how long, you know, how many times did person X touch and then pass the ball.

It's when that person had a ball, what will the opportunity while the affordances in front of them for passing, right? What were the gaps? And here's the thing, all of that, all of those kinds of questions can be answered with the exact same data set, right? So instead of just taking the data, from an individual and Counting, how many times they do things with the ball. Now, you have to start asking questions when the person had the ball.

Where was everybody else that data exists in the data set. But no one's thinking to us, as well, when I say no, and people are starting to think about how to ask these questions. They're harder questions, and then the kinds of questions that you only think to ask. If you think affordances are interesting and there any of the kinds of things, if you start thinking of it more relationally, where you where the skill of a player is not just, did they pass the ball

with that? They pass the ball, when passing the ball was a good option, right? How many times did they try to pass the ball when it really wasn't on, right? Those are too, you know. Very distinct questions, but it's all. It's all potentially answerable. This is the thing that intrigued me. Is that the data is there. You just have to ask different questions. I would like I said, there are people out there who are trying to ask and answer these different questions.

But the reason why those people are coming up with those different questions is that they're ecologically driven, right? They understand that they're coming at it from an ecological perspective of coming at it, from an understanding of more dynamical systems and thinking more dynamically and thinking about process, rather than outcome, So the good questions aren't coming from the data? Write the data is there and can be interrogated in a bunch of

different ways. So and that gets back to the idea of just, you know, generating, you're a square foot or whatever it is on whatever your data. Statistics are great but what statistics are is, it's a linear model. You're trying to model your data with a set of linear components. If it kind of model, a nonlinear system and you're only going to get an approximation out of your model unless your model of the data can handle non-linearity,

right? So a general linear model is useful and Powerful because the maths is straightforward and it works surprisingly well, but the reality is that actually the things that were kind of interested in especially once you Getting into Pro Sports, especially when you start getting into more complex, you know, sports with multiple teams and lots of different people and lots of different moving parts

that's a man. That's immediately a nonlinear complex system and you need to be approaching the data from an entirely from a, radically different perspective in order to be getting any sort of meaningful answers out of the data. But I did, like I said, I just want to emphasize, right? I get kind of thank you for this but I am optimistic. And the reason I'm optimistic is The technology that now exists to generate the data quickly and efficiently produces data but can be interrogated.

Using these tools we just need to join up more and then the other thing gets of, you know that you're talking about the the the time for the coaches. Coaches weren't involved in the development of the interface for pulling out which variables right? This is the engine, somebody sent the engineers to do the coaches in the psychologist job. So the engineers didn't develop metrics based on what the

coaches wanted to know. The coaches are having to learn how to do stuff with the numbers that the engineers were able to easily pull out of the data. So that's and that's the wrong way around. That's a bad design process, right? So the ideal situation, which again, is happening out there in the world, but it needs to happen. What needs to happen?

More is that They technically minded, people need to be asking need to be applying those technical skills to answering questions brought to them by the behavioral experts boat ecologist by coaches by athletes and people actually know what's going on. Take those questions, apply those technical skills to answering those questions. And then building the nice flash interface that takes the data automatically generated those numbers instead of different,

right? And so that's again, it's a more iterative design process but it's and it's it's a mystery to me, why? It's not the norm, right? People people seem to have a very They don't seem to think that coaches should be involved in the design of the of the analytics product. They might talk to them about the usability of the interface and have that kind of user discussion, but they don't seem to be involving them in the development process and we and the name, the net result is a

bunch of numbers. Some of which are interesting and might correlate the something interesting, some of which aren't, and then you hit people. And then of course if you've got coaches with limited time, trying to figure out which numbers are good and which numbers are bad and eventually throw the whole thing out, The waste of time. So anyway, that's my picture. There's any analytics companies out there that want to play.

Come talk to me, right? I can solve all your problems and you can solve or Mine by being able to implement these things. Yeah, and then to kind of like bring some more like to it, you know, like like you mentioned in terms of football. So stats bomb allowed users to use open source, tracking data for the World Cup.

So within that, you know, I know this happens a lot in terms of, you know, understanding like football, both European and American is Is, you know, affordances, it's like oh, that guy was open when I release the ball or pass the ball, but then he wasn't so like being able to analyze that data. Highly crucial on that end. But you know, I just thought that was a good point on that and but, you know, like you said to that that technical model in terms of Engineers, like, that's

another great point. But you know, the fallacy that we're underway Right now is we want things quickly, you know, if we don't if we go to like a fast food place and say, you know, hey I want a hamburger with fries, like we don't want to like break it down and say, like, I want this beat, you know, this Patty this way. This Predator cooked medium, etc, etc. So I think part of it, too is just adding a little bit of patience into our system. Yes. Yeah, absolutely.

Because these things do take time and they are little miss here and it is the relationship between the number and performances harder, right? I mean, this is the other fingers that this, the other thing that I know that it's provides, this kind of that illusion of Simplicity and control. Oh, I got someone to run faster. Therefore, they're a bit of footballer. Looks like now actually, you've got them to run faster.

With a runoff that right, do anything, useful on the football field is an entirely separate question, right? And so you know it's delightful that you were able to get them to run faster and then measure that And sure being able to run fast, definitely correlates with some performance things on the football field with your a squared of, you know, whatever it is. But yeah, we can probably do better than that. Actually, that's the other thing I would like, like, that's it, right?

Sure, you can do that. And you can get some information out and it's clearly working because some companies and some teams are doing things like this, and getting benefits from it, right able to Jitney able to push things forward. I just feel like we can probably do even better. I mean I that to me is the the whole point there of like I think we can do better than what we're currently doing.

But when people's jobs are on the line, I think they want, they want something that is more proven like okay you want to do this, and this is where to me, the whole report system comes in is like there's a game being played of a person with a nicer Report with something that's easier to grasp. Is more likely to get move forward.

Then something that is a little bit more complex that takes a little bit more time to like really grasp and understand and that's that's where I think the challenge at least to me is even if you use experts are like coaches to develop technology, they wouldn't develop what the bat sensors were telling us if that makes sense. Like everybody tries to take their understanding of a mechanical model of an like a

traditionalist approach of like an ideal way to move. what they've come to understand like to be the biomechanical best optimal and they try to shove that into what the technology spits out cuz I saw that happen when the bat sensors came out of people who are trying to infer like over there hand staying inside and even though it the technology didn't necessarily demonstrate that but it was just interesting to see that because it even to if you believe that a batter should actually swing

down The ball. Well then you're going to change how you actually look at it. Even though all your data coming in doesn't most often doesn't reflect that you're going to try to push the data in a certain direction. And so to me, that's where there's something to be said of observing what's going on First and not immediately, assuming that it's wrong in that you need to fix it.

Like and that's, that's where I see like, sometimes the, the danger of taking your preconceived notions and trying to fit everything to that. That and change reality to fit your your theory. And so that's where I think it's like, it's you why theories are important and like because like for example, when you start talking about science, like I almost would like to hear your thoughts on like the myths of science that people because it they have, right.

You talked about, I mean I've heard of like these different theories of like the frame problem. And in this idea of using that idea to like, explain why is it that we struggle to be on the The same page.

And this is me like, I mean I've only just interacted with this so like I don't have a deep understanding of it, but I know that this is a whole theory that helps and and part of what's useful of something that's well-constructed is that it gives you proper constraints to keep you to direct you. More towards your goal of what we actually all want but because we've been sold, oftentimes this a like pop science. Understanding of things like to me, everybody talks about like

well can you explain it? Simply and it's like if your goal is always Simplicity and to explain it simply not that that's necessarily a bad goal. Oftentimes you strip out the what something actually means in order to get that Simplicity. And so you actually you lose you don't actually have the whole thing, a really good Simplicity. Actually transmits almost the whole thing to a person to be able to actually Stand it.

Yes, so yeah, this is a thing. So one of the hats that I wear in my day job as an educator, right? So, I'm a lecturer at a university, so I teach. But I'm also interested in successfully communicating and educating people about ecological approach.

Through my papers are interacting on social media, Etc. So, this issue of how best to educate people about these things is very much on my mind and then one of the things that the specially comes up in the coaching Realm, And it especially gets thrown around in social media. Is this idea that we could logical psychology notes? It's too hard. It's too full of jargon and all that sort of stuff.

Can't you make it simpler? Can't just rip it down and I think your point is very well taken, but actually, There's, there's being clear and there's simplifying and those are, those can easily be two different things. Simplifying can often just be trying to take a square peg in jamming through a round hole and you if you know you succeed but when it comes out, it's not a square peg anymore, right?

It's being rounded up. So I always try and reply to people and say, look, I get that there are technical terms in the Act and the ecological approach. For example, we talked about things that don't feature in everyday language.

We talk about attractors and Paris dynamical systems and we talked about affordances, we talk about perceptual information, we talked about Attunement and all these kinds of self-organization and part of the issue was these Concepts don't just float around in day-to-day conversation the way a lot of psychology in psychological terms, pleasure

around in Daily conversation. So they are technical terms in that sense, but part of the issue was their only jargon if they're kept impenetrable and I can help for example, bring someone along for the ride of learning, what those words mean and how to use them, but the person's gotta meet me halfway, right? They that that like there's a there's a responsibility of the learner To come towards me as I am coming towards them and then we figure out how best to get

this to work. And the thing is without that with just the demand to make it simpler, make it consumable for me, that's not how that's not, how Learning Works fundamentally that's not how learning what that is. Is it's just an objection that's designed to make it so that I can succeed. So it's kind of a it's a bit of a bad faith move and you know em but here's the thing right?

It comes up you know students will come to me and say I just wanted simpler and part of my job as convincing them that actually we both have Some responsibility here, I'm willing to do my part right and come in and try and help those communicate. That, that's fine. It's okay that but it's also, it's okay that the stuff is initially a bit tricky and new and that if you want to actually understand, if I can't, I can't give you that. I can't give you knowledge. Well, let's face it.

That's a very ecological notion right there. I can't give you knowledge and have you and make, you know something. Right. What we can do is between us come to an agreement about what's going on. And the terms of just the vocabulary, the terms, the language is just it's just a vocabulary that enables us to talk about it as well as you know, I've got as a whole tool. You know, toolbox of dynamical systems modeling stuff that enables me to talk about what

I'm doing using maths, right? And sometimes that stuff's actually way more appropriate than language because language is actually, you know, it's a whole other kettle of fish there as to whether or not, some of these things are very hard to verbalize. Verbalize right. And but whereas they're very, you know, it's very straightforward. If you can just wrap it in the equation and just get once you spend a little bit of time, learning what the equation means, right?

So, yeah. So there's, that's it. Like, it is, it is tricky and there's a lot to it does. Require it requires effort from everybody and at the present of the other person on the other end doesn't want to make the effort, but there's only so much you can do to help them learn if they think that learning is transmission of Knowledge from me too. Amazon is so much. I can give them that. That works on a few different

levels as well. I mean, for me, like, if I were to say something like affordances, like I'm communicating to you. A whole lot of information through one word like and I began to realize this in like religious texts and language, right?

Like if you use a word like Calvinism or like, whatever, like you're communicating a ton of information through one word and so it cuts down on On how how much you have to go over and cover, like, you can you can then grab something and, and, like, move with a whole toolkit of understanding. And like, and to me, that's the that's where you can't sit

there. And just be like, well, I want the simplest version of it's like, okay, I can maybe give you an analogy and that can a good analogy can take you really far. I think, you know, because you can. And that's, that's where, you know, to jump back a little bit the way that I began to think about. Like, when I encounter a problem, like, Especially like one that we're talking about. Like there's so much complexity

there. I almost feel like I have to take an ecological approach to understanding that issue and that's where I actually think the real power is, is like, this is something that like Sean. I think, you know, of a theme that I hear from from, from Sean Miska. So much is like this name should become your form of life. Like when you begin to actually become ecological thread through it, helps you make sense of the world a lot more. And at least something that for me.

Me personally, like, I've kind of helped myself to or came to an understanding of like if I understand something, I don't get to be mad about it. Because if I understand something, then I have to now problem-solve to either become more. I guess now, to use the ecological language, become better fitted to this reality of, like, what I understand. And so that to me, you know, in a long security this way is like, how I began to like try to

parse through this stuff. But like because there are certain challenges of like just understanding human psychology of most people because of the there's so much information out there like the in order to function. It's easier to grab something like a surface level understanding and to run with it. And that's where I think like the power of ecological Dynamics, you could say is like you can you can take a very surface level understanding and run with it and you'll get

really far. Like I mean, I've just seen people with Who don't actually understand ecological is very deeply but they'll they're doing a lot of live a B's, you know, facing a pitcher. They're blending their training like they might do quite a bit of strength training to work on improving your bat speed. But just like you said to me bat speed is something very similar to just Sprint speed. Okay, great. You can run really fast right

ahead. It does give you now you have more action capabilities, but at the same point, it's not come To anything game like yet. And so, until you've actually learned how to harness that ability, like, that's because the way that I look at is like everybody's trying to justify their existence because there's, there's this, there's this element of if you don't understand what I do you're easily fireable. And like, that's the, that's the

thing of like, why I understand. Like, oh, you actually have to be able to for somebody because of the, the, the way that I like When I was first encountering, the ecological approach people were were like, okay? So what then becomes my role as a coach, if I'm just sitting back and like, creating this environment, I sit back and I don't say anything like what a what am I like? I have no job and it's like no

that's not actually true at all. But one quote that I'd found years ago before I ran into the ecological approach actually resonated really really well. To what I think an ecological approach is from a constraint lead and all this sort of stuff, what you're doing? The best leader is one who when everything is said and done. The people say we did it ourselves. Yeah, and so like if that's true that like when I'm doing my stuff really, really well, it's

almost as if, right? I am the, the stagehands for a play, like in all the behind-the-scenes people like a really good production happens when you don't know I exist. But as soon as as soon as everything breaks down like, boom, it's like everybody knows you exist and like, and so that's where I think it gets hard to begin to justify to other people your Since when you're, when you're at your best, when people don't notice you, yeah.

And Yeah, but look, I'm fully onboard with all of that sort of idea, but it doesn't pose a bit of a quandary as I understand it in the world of professional sports. Right? Because you if you if you're not seem to be being the thing, that is generating the success. Then the question will then why are you here? Seems like a very straightforward question to ask and like I get that as a challenge, right?

That's a that's a that's a that's a very real constraint that people are operating under two and and and part of the ecological approach to education of any kind coaching, physical skills and, you know, I try and teach as ecologically as I can to, and a big chunk of what I do is not being in charge of things, being being a protein, 8 constraining and guiding and managing but not Not being the one responsible for the success of the session and yeah, lot of coaches.

I've talked to talk about that as a thing that they value and that they've come to value. And it's one of the reasons why people who like the ecological approach often, like the ecological approach is because it aligns with that value of thinking that their job is to be a little bit invisible. Their job is to just be the measure of their success is not how visible way are the measure of their success is, how. Well. Buddy else does.

That's a, that's a very particular, it's a very particular value set for an educator and it's and it's not one that everybody shares for lots of reasons. And so if you don't have that or if you live in a world where that value set is not value, then it's a very real problem and it's a thing that comes up a lot, right? There's a need to be able to show results. I need to be able to show that the things that I'm doing are producing results. And it's, yeah, it's a tricky

one. It's a tricky, one, part of what we need with, you know, we need people in positions who were doing stuff to be spending more time about just talking about what they're doing and how they're seeing the success. So that the we're developing ways of talking about what counselors succeeding. Yeah. It's a tricky space to be in and I absolutely get that as a constraint as an obstacle for people adopting. I mean because like when you leave, you're talking about like

results like okay. I was the one in charge and so I guess I'm responsible for everything that then. Results from it. And because they're when you get into the complexity of it, how can you actually really deduce? How much of it has to do to, which part of the thing that you did? Because if you're doing multiple things and lots of things, like, how can you pin it back on? I mean, unless you have a theory to kind of guide you towards and highlight what things you think are important.

I mean, because we're this is it right? This is one of the interesting things that's been developing alongside all kind of the theoretical rationale from the ecological approach and Flying that, that sort of mechanistic, understanding of what perception action and learning and skill.

Acquisition, our clients Sports big logical, Dynamics, one of the other interesting things that's been developing in parallel with that is develop an ecological ways of talking about the activity of operating within that space, right? So, Marcus Allen and Carl woods and Keith, David's, obviously and lots of other people and change form. And things are writing papers that are about. What it's like to be a person operating in most cases and what counts as succeeding and not succeeding.

And so a big chunk of those papers. One of the interesting things about those papers is that they're busy having to develop a vocabulary for articulating, what counsel success, right? Because the success is coming, is this happening in lots of different ways? And the other thing about, of course, is that one of the other key elements of the ecological approach is that frankly, the coach is not the sole determinant of success. Yes. All right. Like I was saying about education, right?

You know, I could be the greatest teacher in the world, right? But if you're not listening then I can't help. You know. So there is so part of the success is success or failure on where those things are coming from is distributed across the entire institution, across the team, across the academy across the across the institution, that that is supporting the players

Etc, right? If all your players are I don't know, fold your players are tired because they're working two jobs in order to be able to do something. Then that like there's only so much. You can coach pass that food like as a dumb example, right? But these things are you know, so this is this is where you're trying to get to this complexity, right? Complexity is the deal and that's the other nice thing.

I'd get steps that point you were saying earlier about one of the things that I certainly, like, I think you were saying,

as well. One of the things I like about the ecological approach is that it is at its heart, a brave attempt to front up to the Actual complexity of the actual thing that you're trying to study orange intervene on and it's an attempt to, to meet what you're, what you're interacting with on its own terms rather than to take it and to jam through the tools of your general linear model, because that's what you've got access to.

So it's funny actually sort of historically and academically the ecological approach has been a place where, you know, we will run Stetson to our and overs and stuff. Like that. But, you know, to be honest mostly what we end up doing is internet because a big chunk of our training is about trying to front up to the complexity of

stuff. People are ecologically, minded, Biden, dynamically, motivated, Tinder, try and build analysis, tools that suit or go and find analysis tools that soup the complexity of the problem, rather than studying the problem and trying to jam it into the analysis tool that you know, how to do. And that's again a thing that makes us really stick out academic, right?

And but it and it becomes a problem because we're the only ones using only people, you know, we're doing things like uncontrolled manifold analysis and You know, fractal this and differential that and see what everybody's getting very sort of stressed and that gets back to the vocabulary issue row.

And again part of our job has to be in people to jobs, are trying to do this with, trying to bring people along some of the reason why we're doing this is because but I think you're I think I think trying to front up to the complexity of that is input of what it is. You're trying to do is important. That end and one of the nice things that's been happening over the last 10 years, is that there's a lot of tools out there. There's there's tools for handling data that friends up to

the complexity of data. There is vocabulary to Lego sayings like this. I'm just thinking about some recent stuff from Marcus Allen and Jimmy Vaughn, you know, they're building a conceptual tool box right there. Building ways of talking about these things.

Right? Because some of those ways of talking didn't exist before because they weren't required, is no one was talking about the full complexity and what, they're a big chunk of a lot of that sort of work is trying to do is build a conceptual tool box, but then enables you to have conversations to tell people what it is that you're actually doing. And you can start having conversations about whether or not you think it's working.

Yeah, so one thing that to jump back a little bit the that you touched on was the measurement part 1. Thing that I took from David Snowden. And his can even framework is that it's because of, it's a complex system, it's better to measure the direction of travel than it is to measure. Like, are we getting to XYZ number and, and so like, to me it's in service of whatever your main goal is. And so, for example, in baseball, you have to ask the question.

Well, how like, what are we trying to do here? If the goal is to win a baseball game, you know, because it mean because it's at what level are you breaking? Down the analysis because everybody wants to win the championship, right? Well in order to do that you have to win games but you only have to win enough games, to get you to the dance. You don't have to win all of them. You just have to win enough because like the team that has the best record in the regular

season. At least in baseball, does not is not, it's not indicative that you're going to win the World Series. Yeah and that's I mean Robert. I guess you could, you could speak more to that. How often the team with the most wins actually is the one that wins the World Series that people have been been? Complaining about that actually more recently that like relative to other sports baseball tends, not to have the best teams playing in the final game, which I don't, I don't hate that.

But it does if it's an interesting outcome of the of the selection process, the selection process selected, the best team, and the best team isn't necessarily the ones of all the best players, but it's the one that was able to consistently, do the work on the day and across, An extraordinarily large number of games, right. So there's there's your selection pressure. If you want a different outcome, you have to change your selection pressure. You have to change the structure

of the Sea of the regular see? Yeah, absolutely. I mean, on the second end of it, in terms of just here probability, like, given how baseball is constructed, you have your three games here at wild cards. You have your five game Division, Series seven-game Championships World Series, like, there's that element of Randomness within that probability to say that the

weaker team. A say is more likely to win because if you do it, whereas its, you know, like football where it's just one game more often. Not the better team will win because in just one game. But since the sample is increase, either 3 5 or 7 there, is that element of Randomness in the probability?

Okay, so I want to own a touch on the probably like the randomness probability stuff because To me when you get into your because there's underlying presuppositions because a lot of people take the randomness probability and then they try to translate it across like those large datasets in the sense of okay well when a guy is on a hot streak or whatever well that's just luck. It's like well is it really?

Because if you if you go with that assumption then you're not going to look for any underlying mechanism that is causing that in the question is is to me an ecological approach explains. Why guy is hot? And then the question is, when we're wearing a cold slump or whatever it is like how do we

get you back into that? And then there are other, there are other things to of like, well, he's just hitting it really hard and he's just getting really unlucky because it's going to guys are like, well, maybe we need to train him to hit it, not at guys. Like and if you don't, if you, if you going with this like luck assumption, that hitting is just luck because it's hard, which it is and I don't think maybe you

can hit the ball with precision. Asian to to wherever you want on the field, I think you there's enough evidence to demonstrate that, you can hit it to an area of the field, you know, like you given like I mean and so to me this is where your your underlying assumptions will drive. What you think is possible? Yeah, especially bicycle vanities is, if if there wasn't an element of skill involved it wouldn't be different type of differently.

Skilled. No, batsman, Rye. Like people would all basically on average over a season end up with about the same batting right now ranking. So what there's D? I mean, clearly something going on. We're different. People are better able to do different things. Yeah, for sure. And as a psychologist, as soon as you see any sort of pattern like that, you know, as a scientist, the first question is 0.

Why is it? What's pulling you away from everything being being even there must be something going on with, you know, look this team work trying to figure out what that is, right? Part of the problem is, of course, it's focused on things like it's focused on boring. If like visual Acuity, or reaction time or all those kinds of things where it actually needs to be. What we actually know is, the behavior is behavior emerges from the collective of constraints that are operating

on you, with the moment. Some of those you bring with you, some of those are imposed on you and it's how they play out over time, right? So as you're trying to do the thing that's, that's where Behavior comes from. So, Yeah, the big problem has always been going to look for that enduring feature of the person that you might be able to improve, right? Whereas it's obviously more complicated than that. All right, Baker. You been very silent for for most of this.

I'm curious. If you have anything that you would like to add and you, how you would like to perturb the discussion. Now, I'm going to join me pastor. This has been a bit of great conversation but to me like one thing I've been thinking about a time as you guys talk. And I've been thinking a lot about this is the idea of like a skill, first like action capacity or capability. Whichever way you want to frame that because that's a conversation.

I've been having a lot with a lot of my co-workers is a rebuilding skills which to me that's more of an interaction process. So we're actually interacting with the environment versus just like this capacity or this

capability. We're hitting the ball hard is probably more of that versus skill is actually a live in an environment with Fielder's with the header and actually doing what you want to do which is get hit scan on base score runs and kind of bridging that Gap where I think the capacity or the capability is easy to measure skill.

On the other hand has it been Quite as easy to measure an oven, me being the pitching, space of the same thing where I think the majority of the time we'll spend working on throwing harder, working on throwing different pitches with better metrics based on whatever model. We're chasing verse is actually skill where we put it into context and actually it only lives within the environment. Are we getting out?

Are we throwing pitches that is producing out and preventing runs which what we're trying to do and trying to get two more measurements on that front. That's been a lot of my Headspace not that I'm good with the analytic side. But that's what I've been thinking a lot about. Yeah, the metrics thing. Yeah, it's a really interesting and challenging problem is coming up with metrics metrics

or skill. That's a really nice way to frame it actually and that's definitely honestly, it's like it has back to what I was saying earlier about analytics and thinking about what to do with all those numbers. So it's definitely been on my mind as well because it is. It is hard, right? I can, I can measure this, you know, you can measure a pitch speed and see if that. Number goes up, that's really easy. But you know, measuring whether or not the pitcher, threw a

pitch. That was the kind of pitch that was likely to get. That better out. Is a more complicated question. It's the more interesting question, right? Like when it comes to, you know, council's decision making, you know, what's a good decision or, sometimes it's not the fastball, that gets the player at it's the, it's the slow one. All the curvy model, the spinning one or something, right? So what constitutes a good decision, you're right.

I think is a very sort of a hide its high dimensional problem, right? There's, you have to be a, you have to measure and assess quite a few different things. And part of the equation is not remember it's not just up to the picture. Part of it is up to how the batter is able and their ability to, you know, the picture could just produce an absolute beautiful pitch.

That's everything's perfect. And you know, right better on the right day, puts it over the fence because that's just where they were living in that particular. Right? So and so does that count as a bad pitch? Well, yes. On one dimension but also maybe not on some other dimensions. And so you have to start thinking about these things in the yeah. So yes, right. That, that issue of of developing metrics of skill. I like, I say, I'm optimistic in the sense that I think it's

possible. And I think that the systems that are being produced right now to generate metrics can be adapted. I think the data is there to ask more sensible questions on the really am optimistic possible but it's hard to figure out exactly what the question is and it's especially difficult given that what? Yeah, what counts as skilled. This is a complicated one. So so for example I'm so I'm just been working on a paper.

So it's some data I collect go to while back on targeted, long-distance throwing actually, and I use it too. Holden's perception. Perception of Target affordances. And we're using this analysis, called uncontrolled manifold analysis, which is just it's a mathematical way of decomposing variability and performance. And the basic idea is that. So I asked you, I measured you throwing a tennis ball to hit a Target 20 times.

And I measure, you know, measure various drawings angles and things like that might get you to do, effectively the same thing 20 times in a row But of course, what we know is you don't actually ever quite do the same thing. You it's repetition without repetition as the rule, right? So, you'll produce basically the same through, but it will never be exactly the same throat because it cannot be. That's just, that's just one of the rules.

And also for, you know, for lots of reasons because of the ways, skilled action moves redundancy and all those kinds of things that are going on in the system. There's always variability. But that variability is organized around a kind of a centralite central kind of goal of what you're trying to achieve. So there's some controlled manifold analysis is just a mathematical way of mathematically. Characterizing.

What you think the goal of the skill is trying to achieve And then analyzing the variability in that movement has two components. One components are variability that stops you from it, achieving that goal and variability that doesn't stop you from achieving that and variability that doesn't stop you from achieving. That goal is just kind of its that's kind of considered to be

good variability. Well that's the variability that doesn't get in the way that's just that's just this redundant system coping with minor variations and what it's trying to achieve is a function of, you knew your balance is slightly different. You started from a slightly different position. Etc. That's all healthy. That's good.

That's adaptive, right? And so one of the things I like about this analysis it turns out the details of exactly how you implement everything, interesting Lee complicated. But the basic idea is first of all it's trying to front out to the complexity of the situation. It's an analysis method that was designed to try and fully front up to the problem of what we're trying to study namely that I can take all those don't measurements and computer an average and then compute

deviation from that average. Judge, but that average is meaningless, right? Interesting isn't. What did people do on average? The interesting thing is, how is the variability in the movement, organized? That's quite an interesting question. So what you end up with is if you get, as people get better at producing these throws more and more of the very, there's always

variability. So you have some amount of variability as you get better more and more of the higher and higher proportion of that variability, lives within the uncontrolled mental lives within the space, where it doesn't affect the outcome and variability, that does affect the outcome, you're skilled. Your skill development is around organizing a movement so that that variability doesn't happen

anymore. so, There's the right. So once on a minute you're trying to develop some measures and metrics to get a sense of how skillful was a given movement, right? And the movement is skillful is if the on this account is if the variability in that movement that you can't get away from, if that the variability is organized. So as to make it so that you still achieve the goal, that's purposeful, that's intentional, that's flexibility.

That's adaptability right, that, that variability is good at. So now we need to cope a little moment to moment variation. Stuff. So, what you do is you, if you measure that and you're good, variability is going out, then what you've got potentially as a metric that tells you a little bit of something about skill as you were talking about.

As now, the details of that it gets for a complicated, you know, the max isn't that have, but it's fiddly to measure and there are constraints in terms of how much you can do in terms of making about, and that's hard to care a pro. Like, you know, I've got people throwing tennis balls to a static Target, right? And I can characterize the details of the static Target. Or if I had that Target moving mind control manifold analysis. Mathematically would get

complicated, right? But at least in principle There are ways of trying to take the kind of data and ask interrogate it with skill based questions rather than and skill. Based questions are processed questions. Not did you throw the ball or how fast did you throw the ball? But how did you get there? All right. And that and the the annoying thing about those process based questions is because of redundancy and things like that. There is no single correct answer.

There's no correct form. There are various There's there's techniques that have better than other techniques, right? My baseball throwing techniques probably not that good relative to someone who's been practicing for also, there is the out of all the range of possible things I can do with my arm only some of them are good at pitching a

baseball right? See there are issues around that but within that space there's a lot of wiggle room around what counts as a good patch and part of what counts as a good pictures. What are you trying to achieve with that pitch? And so All these quick rate that quiz basically, the question you were asking about metrics of skill as a really good question. It's a really important

question. It's a thing that's been really much on my mind as I've been trying to use this new CM method as a metric of skill, specifically skilled interaction and put still perception and interaction engagement with the affordances of the target to be head. So it's there. The possibilities. Are there the tools are there? We're getting there.

There's there's ways of going at this and I think that it's worth doing, but it is hard and at the end of the day I couldn't tell you what, constituted a good throw so I'm not I'm not a Point entirely, sure what this data would do to change the coaches behavior for example, and change what the coach of it. Now, I'm not saying that that that doesn't exist. I just don't know what that would look like right now because it's more complicated. You can't.

It's not just about you need your player to throw faster. It's you need your player to show. A different kind of adaptability and that's how you can't verbally. Tell someone how to do that, right? This is again where the constraints based approach, I think really stuff to kick in. Is because that variability what constitutes good variability is is the is the variability as the arm moves through space? Is that successfully navigating the constraints that are

currently on offer. That's effectively the question, right? So if you want to change how that variability is organized, the actually the only way to do it is to alter the constraints and see what happens. So the constraints based methodology is actually the only way to tackle this potential skill based question, can't lick it, it's not At all clear that you could verbally and structural way or you know, to doing that etcetera. So but it's harder. Absolutely.

It's harder. I don't know what that like, it's not like I couldn't turn that into something. I could sell somebody. Right. I guess I'm as you can see it for me a list if we're to go to the Practical. Because I think that's that's been some of the The criticism is like, oh, this is all Theory. The salt area. Well, to me, I think you can take that what you just said and make it practical.

Because I don't know Andrea, if you seen the logo that I use for this podcast, it's it's trying to represent the uncontrolled manifold analysis, but I didn't quite get enough dots off the line. But anyways, my my thought there because it to me, I've seen the potential of this from, from the very beginning. And like, now, that I've been thinking about it a little bit more as you're talking there.

To me the outcome that you would measure it against is out, like, it doesn't move you towards an out State. And this, that that last piece makes it a little bit more complicated like, does it, is it moving you towards and out state? But the way that I conceptually understand it is that it just includes more data points.

So as I understand to like, when you're doing an uncontrolled manifold analysis, you are, you're constrained in terms of what you're looking at. So, you might actually have to do multiple ones. If you want to get a better order to see more of the reality of what's going on, and so, in this scenario, probably, you might have to have a few like, because, like, in baseball as you progress through the count,

things change. And so like, you know, Robert Robert will do different sort of analysis that will show you the changes as the count progresses. And so similarly to me, I think that's that's how you would build this.

But you had, you would have a top-line thing that would give you because really what you're looking for because to me as I'm thinking about this even from an ecological approach like I'm looking for something that is specifying enough to then pull me a layer deeper into more and more specifying but like Because maybe we could get into this idea to of like specifying information. And is there a difference between specifying information and higher-order variables?

Because to me there's there's layers of specifying information and if you can move to a higher form of specifying information which to me is a higher-order variable, you begin to then that higher-order variable then gives you all the other lower specifying stuff and maybe that lower specifying stuff is actually context-specific. And so like the higher-order variable actually points You too. What is specifying in your environment?

And so, how all this comes back as to actually how we View and utilize data because it's going back to like what you talked about before. Like what does this data mean? Like you always have to answer that question of like okay what does this mean? And what do I then do with it? And so, what you're trying to do with the uncontrolled manifold analysis is, you're actually making your simplifying. It, you're asking a simple question of did this meat what I

needed. Yes or no. And you Can actually create probably a score for like, how close are you to that line? Because like what's interesting to me is just like when you look at it visually it looks a lot like the traditional linear models that you're actually used to looking at. You know, if your to like look at something you actually want things to come towards line and in an uncontrolled manifold analysis, that's actually what

you want to show improve skill. You have a better Fitness to that line because if you have it, tied to the outcome that that actually matters. And so if we're talking about Pitch, The outcome that you're looking for, is, are you getting an out? The challenge with pitching actually is it's actually simpler for hitting because hitting, it's like the answers. Did you hit it? Yes. Or no. I mean, you could make it more nuanced if you want to get

towards more of the skill. Element of like, did you hit it? Well enough to get you a hit, there's a difference between like asking the question, did you hit it? And then you can look at something like the the temporal constraint. Because for me this is where R, what you highlighted with the uncontrolled of manifold analysis. As I understood, it was like, oh cool, you just have to get the answer.

Like if you just pick a random number like 5, how many different ways are there to get to the number 5? Yeah, the answer is infinite. There's an infinite number. Even though that's like so for example, if you're in a time constraint there is a constraint in terms of like the how you do it. But the number actually within a constraint within an area is actually infinite, right?

Because of the That you can use decimal places like you basically go to Infinity. So, even within a small constrained space, you can get infinity. I don't know that. That's probably more philosophical than what we need to get right now. But I find that fascinating that you can get infinity through that small constrained area. Yeah.

I mean, that's the, that's the essence of the, the kind of ecological Insight. That underpins the uncontrolled manifold analysis, is this notion of murder, abundance, right? And rape. Additional that repetition the idea that the your body, the system that you're trying to control, always has more degrees of freedom than it needs. In order to achieve any given task it just it just does, right? And that's just always the case. And given that that's always the

case. That means there's always more than one way to do whatever it is you're trying to do. And technically, there's an infinite number of ways of doing. Although, in practice, right? It's not infinite and, you know, in principle it is. And that in the whole point of the whole point of the uncontrolled manifold analysis is just to take that insight and then also to identify, make sure there's an infinite number.

Ways I can do stuff, but there's a lot of actually there's a lot of organization and structure going on and how you actually go about navigating that space, right? So there's a lot of things that you never try their options if you ever needed them. And one of the things that you see is that, if you, so for example, you do uncontrolled manifold analysis. You get people to do to try and

do the same thing, say 20 times. All right, so I got people to throw to hit a Target, 5, 10 or 15 meters away. Twenty times. And so and the reason you do that is that you need some sense of the spread of people of what the repetition without repetition is they're trying to do the same thing and they don't do it exactly the same the whole time. So you need to map out that variability but then you get into the interesting question. One of the things I've run into as yeah.

Well what actually constitutes the same thing and so throwing pitching a patch. When it's no outs versus one out, versus two outs, might not count as the same thing. In fact, it almost certainly doesn't, right? And this is where you have to start thinking about what it, you know. What's the various? What's the Coalition of constraints that are currently operate in order to shape that particular behavior from which things are emerging in real time?

So it's not the case that a picture, just comes up and delivers the same picture all the time, right, that context, shifts. And sometimes the context is something like something to do with the rules of the game. And sometimes the context Something something moved to the

physical. The the is the better left or right-handed of a tour of a short of a bigger they short you know bigger they little all those kinds of constraints that you can alter as well and those constraints then again though you know those shifting constraints means that the thing you did a minute ago is now no longer, whatever it is, you're going to do. Next is an example of the same thing in quite an important, sort of sense. And this is where this is This is the essence.

This is kind of getting to the nitty-gritty of why the ecological approach things drills was right because drills drills as an attempt to pretend that Sports is doing the same thing over and over again, which it's not fundamentally, right? It's about skillful engagement with the current constraints of the environment. Now, what you end up doing is something that might look very similar over and over again, each pitch is going to look very similar.

It's going to look like a patch and it's not going to look like that Cricket delivery. Right. Because so so it's not random, but what you're doing, it is constrained but it lives within a bigger space than people give

it credit the pitching movement. And so trying to do the same thing over and over again is not necessarily the best way to train something because what you're actually trying to do is you're trying to teach people how to skillfully engage with those constraints so that they can shake what they're doing. To best meet those demands, right?

And again, the essence, you know, for all its limitations, and for all the various kinds of complexity of applying it things like the uncontrolled manifold analysis there, an attempt to front up to that reality. And to quantify and organize it in a way that you can then talk about scientifically. So you know the very least it's an attempt to to engage with the question on its own front which means that the numbers that come out of it Are closer metrics of skill than other things.

Now, the thing that actually kind of motivated, my initial tweet about being grumpy about Sport Science. Just collecting data came from reading some UCM papers, right? So where people were doing you see, em on a task And then just finding some differences in their view CM numbers. And that was it. I did this task of the BET tasks that you see a number was hiring this one than that.

One. Done. Now, that was it like and I like it frustrated me because you've got this tool that actually has the capability to tell you something more than that. And the only thing anybody seemed interested in was collecting some data applying the method and showing a difference. I was like, okay, so what like using the using that as an example? What more could you have drawn out of it like at a conceptual level? Well, the thing I'm trying to do right now is, I'm bashing my

head against that. Don't know if it's working but I'm going to give it a go. Is, I'm trying to connect and control manifold analysis to the perception of affordances specifically, right? So I'm trying to show that the, that the manifold The thing that defines how the variability in the movement is organized, I'm trying to figure out whether or not I can show that that's being generated that that's been structured by the perception of the affordances of the target

yet. So I'm less interested in. Was there a difference in the UCM scores, between when people through 25 meters to 10 meters to 15 meters and I'm more interested in trying to show, does does my formal analysis of the affordances of the target Add me with a good decomposition of the variance. Right. Am I getting somewhere? Am I producing an uncontrolled manifold analysis? That looks like I'm heading at

the water. It is the system is organizing itself with respect to whether or not there's actually going to work or not. I don't know when a draft of a paper attempting to do it. I'm still going to work through it but that's for me. That's what I want to do it again. That's me attempting to use a method of generating metrics. That is at least in theory up to the challenge of fronting up to the actual complexity of the

problem. I'm trying to use that tool to answer a theoretically driven question about the perception of affordances and the role of that and scale vaknin So again the reason why I'm doing it and as far as I can tell, why it's never occurred to anybody else to do this, which it always seemed like the obvious thing to do to me. That's the other thing that's why I was quite surprised that nobody else seems to have done this at least not the way I'm

sort of flying. That question makes complete sense to me. Given my understanding of what skill acquisition and skill skill performance is from their ecological, kind of understanding. But that's where the question came from. And I think one of the reasons why no one else seems to have done it is that they're not coming from there. They're just looking at the numbers and look at the differences, right? And that's the, that's the issue

again. And like I said, I don't know if I got it, right, but at least I've got a sensible for a swing at it, so, I guess that's why I wanted to follow up with at least my Puzzle of what? I least understand, because to me, they're the entry point is to in a way work within the current model of the system of giving you an indicator good or bad. Because I think this is a simple like, it's a simple starting place, right? If you just even boil it back, everybody one Simplicity, you

know, like what do I initially? Because I know I have a little kid like I'm in some ways trying to give him value judgments, this is good, this is bad. Now, I have to help him understand in situation of like you know oftentimes I'll say things like not Now, because I don't want him to think that whatever it is, is categorically bad all the time, but I want him to understand that in this context, this is not good.

And so, in the same way of like, I think, you know, basically, because the first because when you, when you're, when you're overwhelmed with information, you want a simple value judgment to get you moving. And so that's that's where I go back to like when we're when we're designing, if we were to use an uncontrolled manifold analysis, okay? So for for very specifics in the new hot, Guy system.

That is working in baseball is similar to what you're talking about with football, like they're able to especially with the pictures. I even think to with the hitters is they're able to get a ton of kinematic data. So now that you have all this kinematic data of how like how the picture is moving in now interacting with the problem and now you have it tied to an outcome because every pitch produces an outcome. So if you can Like in some ways,

right? Your high-level number or analysis is going to actually compress a lot of things together so you're going to actually lose some of the specificity to get a more generalized thing. But it's really when you when you're trying to simplify things to make it more manageable, you know, to make the complexity and all the information manageable and to constrain to afford, you have to go back to words like what's your, what's your overarching goal? Like, where you trying to go?

Go essentially. And so if the goal is to trying to win, like this is why everybody breaks it down into like okay, when this pitch because this is how you move the thing forward and so To me, when I'm designing this, this initial metric is to tell you. Okay, are they better at overall? At this skill of moving you

towards this end goal. And if you're the person who has more ways of doing that in theory, to me as is one more skillful and regardless of the situation is, but he's going to be able to perform over a wider range of situations and partially to you can see that the system is because To like, what you do is you take, you take measurements over time and you look at the direction of travel.

This is the whole thing, like what we're talking about before, like how it practically layers is like, okay, if you can track this of like, okay, what's the current state of the system? How much abundance is it showing and does that abundance? You could say, decrease over whatever time scale you want to look at of course of a game the course of a season. You know, like in the past using his past body of work to predict

the future. I mean everything to me is about like timescales and what you want to look at it because some things information becomes more relevant. You could say this is kind of like weather forecasting, right? The prediction is a lot better. The closer you are to the time of the event but the further out you go the more the more the less accurate your forecasting is.

I mean because that's that's part of what the analytics teams are doing because if you're looking at scouting and all this sort of stuff and how much you're going to pay a player all, This sort of stuff, your ability to get a good because to me this is the whole in decision-making, right? We go with the best. The first best available. We don't go with the the most optimal, the most logical whatever. It's the first thing that shows

up is usually what we go with. And so to me that that shows that getting a close approximation to get you going is actually how we do a lot of things. Like we don't we don't need to be exact like is this a to me? Like if you were to get an A Perspective control. That's exactly what perspective control is. Doing is getting you moving towards. Your your own goal and you're just simply saying, like, Okay, it is if I keep doing what I'm doing, am I going to end up at

my end goal? And if not, what do I need to do to get back to trying to meet my own goal? And so I mean to me it's if you're using an analogy it's like a heat-seeking missile, right? Like it just tracks wherever it's going, There's a few different layers there. Start. In terms of precision and how much you actually need to know. So the baseball Scout it might be the case that some of these metrics that's officially correlate with something.

I sufficiently informative that it will enable them to select one person rather than another, or Focus attention on that person rather than that person. And the the number that they're using, is sufficiently informative. That, that that, that using it to guide, that decision works well enough. To make it worthwhile doing at that scale. But that's that's that's one kind of decision, making context, right?

Then there's the decision, making context on, you know, the athlete on the field on the day, right? And there's questions around how that works and how best to coach there. And then there's the other question, you know, from my question as a scientist, as I'm trying to figure out and shows clearly as possible and test between different possible mechanisms for how that behavior

happened to come out. And so really matters to me together as right as possible because I have to go I'm kind of trying to zero in on that detail but just because I'm trying to zero in on that detail in order to achieve what I'm trying to achieve doesn't mean that zeroing in on that details the thing they need to do and it might not be the thing scout needs to do. So there are there, if there is an issue of these metrics being being puffed, you know, being

sometimes been good enough as fine. That's part of it. The other thing that's lurking in there is this is a matter is a very important notion of task specificity, right? Is that when I'm trying to figure out how things actually work, one of the big limitation, one of the big challenges and of the reality of the situation is that behavior is deeply task-specific, right?

What we do in this context, it doesn't take very much of a shift of context to get you behaving quite differently, and that's one of the big kind of, it's a In pain quite frankly because it makes doing the science of gets back to that question about, you know, when I'm doing science I need to ask people to do try and do the same thing over and over again, at least a few times because looking at that variability in those you know their Tendencies and all that sort of stuff is

how I go and figure out what's going on. But it turns out it's really hard to ask that question, which is fine. As a mover, as a person trying to move and do the things right. You know. It doesn't you know having that room to maneuver doesn't isn't isn't so much of a disaster but it's a pain for me is the scientist and it may, I get ya anyway, so these and then on the issue of perspective, control versus these things. Oh, that's the other thing from the point of view of the Scout.

I've been what they're trying to do is they're trying to explicitly, they're trying to predict they're actually, they're trying to predict future performance. They're in a situation where something like perspective control, simply can't happen because they're creating a Situation where they're trying to predict something in the future, that's disconnected from the vents a little bit going on.

Now, whereas Prospect of controllers, what happens when an organism tries to interact with its current environment in a way in which some of the things that needs to interact with their own the future. And it turns out that from the organism point of view. Well, ecologically, we argue that prediction isn't an option. It actually, the only option you've got his perspective control, the only thing you can do is interact with the environment.

Now, in ways that then enable you to achieve things in the future. And one of the things that we've revealed is that it's possible to do that, right? It's possible to interact with what's currently going on. In a way the substance that you end up reliably producing some future outcome without ever having to explicitly know what

their future outcomes. Exactly going to look like but again There's this difference of trying to make of an organism trying to make a decision in the moment based so that it can control, its perception action systems versus the very distinct from that scenario of trying to take some numbers and use those to explicitly. Make some prediction about something, it's a very different set up, right?

And again, that's the other thing is that, then we're not the same, which means that the requirements they have for the data that they use are going to be different too. To jump back just a little bit because I have this bad habit of throwing too much out there like to ask you a question. Then I'll carry on under something else.

What is your thought though, on? Too close to tie up that, that last question of okay using an uncontrolled manifold analysis, to give you a score, you could say, right? Because like if if you have a tighter fit to that line you know because because that sounds to me, as I was listening, you criticize the That one paper, that's kind of what they were doing, right? Is that a fair interpretation of what you have? Like, they're just looking at the fitness between the uncontrolled?

Manifolds and, and so to me though, it's like if I wanted to measure skill, that would be one way of measuring it based upon whatever it is that I tie it to. And so whether I tie it to like the batted better ball outcome. Like, did you hit the ball? Yes. Or no. And then and, or did I Did this move me towards an out?

I think that's what you have to do with pitching and Mason much more complicated, but you could just simply say in such because there's just too many pitches because in which the the potential of an out is limited because it requires a batter to swing and or you have to be an account that would basically meaning to strikes that allows that that next pitch, has the possibility to result in and out with no swing and so, Like that's weird to me, it's more of a question of like you could

probably capture, does this move you towards getting getting an out because if you throw balls into the strike zone, that's moving you towards getting it out.

And and so like it sufficient to to move the thing forward because if guys throw more strikes in a general sense, Robert, you you're going to have to speak to this more, but if you throw more strikes in a general sense, you probably have a, it's a close enough correlation to seeing you move towards being a Better pitcher because you're going to get more outs like you're going to have a lower era, you're going to have a lower Whip and these different these different traditional

metrics because where I go at the end. Okay, so I need to stop before I do that thing again. So what are your thoughts on using the uncontrolled a manifold analysis in that way? Is that an appropriate way to use it? Or does that violate some some fundamental principles? I have to think some more about

this. I don't, I think the underlying sort of motivations for the uncontrolled manifold analysis that the The essential problem that that analysis is trying to tackle name of this issue of motive redundancy motor abundance. The idea of was more than one way to do any given movement and that you're trying to understand what, you know how people go about organizing those movements and coping with that possibility. I think, first of all, I think you could probably throw.

You see him at this kind of data as long as you weren't careful. One of the things I've learned and trying to use the uncontrolled manifold analysis, and one of the original purposes of it was to test control hypotheses. So what you do is you take the same movement data and you decompose it with respect to different potential outcomes,

right? So you do and to try and figure out which one produces, the better decomposition of the variance right now idea, is that so, you know, so you're if I've got joint angles for a throw for, People. I could decompose that with respect to the release velocity where I can decompose that with respect to the position of the hand at various points in the trajectory. And the idea is that you're you decompose the variance according to different potential, candidate outcomes, to see which

one it looks like. The variance is being organized, with respect to maintaining right. And so so sometimes those are what you'll find is that sometimes you'll try and decompose the variance relative to some variable that you think the system, Trolling and you'll end up with nothing. Just doesn't look like the system, cares about that thing at all. And then sometimes you'll find that it is in fact looking for

something like that. So one of the things that I found like the original motivation in the original thoughts and show no paper was explicitly about testing hypotheses about what you think is the thing. The system is trying to organize itself with respect to it, what's it trying to achieve and you can look to see the different portions of how good the decomposition of that's the thing that's kind of trailed away.

In the using, which two people don't use it to test hypotheses anymore, they just apply the method and Report the found and then the other thing I found went there, my data is that it gets complicated because actually I've seen, you know, doing the decomposition against different potential performance variables. Multiple variables. Sometimes look like they're working and it's hard to tell which one's the system is actually working towards and

wine. So it's so the moral of the story is I think the underlying logic of the of the ECM is the right underlying logic of it really forces you to ask the right kinds of questions and concerns. We be used to guide them but then the other question like whether or not you could actually just actually do that with a numbers. I don't know if spend a lot of time thinking about it. But the the end of line. The UCM, embodies the motor, abundance hypothesis, right?

It's an analysis technique that literally builds into itself. The assumption that there is repetition without repetition because of the way, the movement systems are organized relative towards what they're trying to control. And so that understanding that, like, going at the question with bat theoretical weyer, is how you get the better questions and go to your data. Just thinking about the other thing as well. It gets back to the question of what counts as a good pitch, right?

So sure you could just pitch strikes all the time. You could just try and do everything you can to make the pitch. Go through the strike zone every single time. You could do that and it would probably work out, okay? But that's a fairly rigid solution to the problem of becoming a good picture, right? So I always think about the kind of sports player that the players that we always really look at and go that person's mm, just a master of their sport. That person is an amazing

athlete. There are the ones, they're not, the ones, who can reliably produce something. They're the ones who are the most adaptable. They're the ones who never seem to get phased, right? They're the ones that you can't. Throw off because whatever you throw at them, they have a solution, right? Those are the ones we look at and go. That's a person that really understands their sport, right?

That's an amazing athlete. That's an incredible effort and there's all kinds of examples of, you know, there's players over the years of lots of different sports, those kinds of numbers. And I think if you like we all kind of know what those people look like. We all value, those kind of people, right?

You could look at the picture who only ever tries to throw to the exact same place in the strike zone every single time and they might on Courage do. Okay, in terms of else just because of the way, baseball works. But you might not have a look at them and go, man, that's an amazing picture. Might go there are good reliable picture but they're not the one to pull out the magical play or to or to just create something, right? And those are the ones and that gets back to the question of

skill. Right? Skill is not being able to do the same thing over and over again, skill is being able to do the right thing at the right time. And what that is and that means being skillfully. Couplet your environment. You have to be flexibly and adaptively couple to your environment. So that what you do is constrained by the task demands, and by your capabilities, but constrained in that kind of More and that flexible and adaptive kind of when it's capturing that flexibility, right as hard.

But there are numbers, I mean if you really want to go there is their numbers for characterizing, the behaviors of dynamical systems that tell you things about coupling strength and the tell you things about you know as your the stability of your system versus the you know there's you know there's various numbers you can produce it tell you things like that that are diagnostic bit more diagnostic about about the about the kind of system that you have or if you have a rigid system,

that's Not very flexible. Would you have one that's really just writing that edge or if you have one that's too unstable and it's trying too many wacky, right? You can you can, you know, you can you can quantify these things, you just have to ask questions of the day. So how would you tell me more about this coupling idea because like this being able to quantify that?

Because for example, I know, I know there are some people who are trying to figure out how to better Quantify, an ecological approach. And to me, that is one thing. Because, you know, it gets thrown. This is where I recognize my limitations in terms of my understanding. Like, I know what I haven't interacted yet. I know that I've touched the tip of an iceberg, right? Like, I know certain things exist but I don't know the details of how they exist and how they work. Right.

Like in and I think for other people in the Baseball World, they don't even know these things exist. Like I think that's that's where You know, I know I saw some organizations like looking to hire people and it's like okay that's a start. But if you don't even have a clue for how to even utilize this person, they're just going to sit in a corner generating ideas that basically almost go to no one and like it's it's going to be based upon human psychology on which one's actually take.

And like, to me sometimes, what actually takes is actually a it gets twisted sometimes like I'll just use the example of like the constraint. Let approach. Somebody in the baseball industry, popularize, the constraint, let approach but they had a misunderstanding, like they saw the potential of it and then used it to fit their their their technical model into it. So like you can strain to force what you want and then you open up your take constraints off later on.

And it's like, that's not how like that's that's not, that's a fundamental misunderstanding of the constraint. Let approach like we constrain to afford, not constrained to get what I want. And then I take the The constraints off and then you maintain that because that that's just a traditionalist of approach. Yes, there are always constraints, right? And if they're on constraints, then you don't get any structure

and your behavior, right? So I'm giving you know, this is this is where, you know, look if you're going looking for numbers, right? People have been studying, nonlinear dynamical systems physical systems for a long time and figuring out ways of quantifying their behavior, right? So you know, so for example, there's a classic chemical reaction. I can't remember the name of it. And after a couple of Russian chemist that figured it out and that's this weird.

Also, lighting chemical reaction basically where you mix up these various elements and they chemically react and turn into one thing and then they chemically then they that reverses and the exact time it just it does this by itself effectively, right? And it oscillates and so there's a bunch of different you know there's the this and and that that happens and the details of how they happen for example. Emerge under the constraints of the various tasks.

So the size of the container matters, right? If you put it in, put these liquids into a really big container where there's no, you know, with it with the liquid doesn't brush up against the edges. For example, then you don't get these kinds of reactions, the temperature of it, obviously matters how much you put in with all these for the kind of physical constraints that you

can vary. But basically you just get this kind of interesting nonlinear and and you get this nonlinear kind of behavior of it. There's many ways of mathematically, characterizing, the behavior of those. And there's numbers that tell you things about, you know, how these things are, how the various components of couple to one another and, and rates of reaction, and all kinds of things that are just kind of more Dynamic than numbers that

enable you to assess these. So, one of the things that the ecological approach does is because we're because we're going the ecological approach Gibson's, big Insight was to spend a large amount of time. Characterizing the problem faced by Theory of perception and action, right? And that's that's one of his major contributions is that it didn't just jump in. Trying to figure out how it worked. He spent a great deal of time.

Trying to figure out how to succeed correctly, characterize the nature of the problem that we're trying to study because we do that. If you start with a particular understanding of the problem instead of like I say, just gets back to the same old thing instead of just taking your data and jamming it through the analysis technique.

But you know, ecological psychologists have been Been at the Forefront of going and looking for numbers that are appropriate for studying this kind of system, once you know what kind of system it is. And it's like well and the immediate thing is well you know, there's no point in running an over on it because it's just assuming some sort of linearity. What I actually want to know is something that needs. I need a different kind of number and so we go. I wonder if anybody else has

ever done anything like this. And then you go well, who studies dynamical systems? Well, there's a bunch of physicists and chemists chemists kinases chemists have been doing these various things. Let's go see if they've got any numbers. And so we are looking for other other numbers from other fields and we're quite happy to go grab

and there's tons of them. And you know, it's one of the things I would quite like, actually the field to get a little bit more organized around is producing like this, this expertise are distributed throughout the field. I think and different people are different expertise, and I would be quite nice to kind of consolidate at have a bit more about to provide people, like yourself for example, with a bit

of a pass. Both of we went looking for a number that enabled us to ask and answer this kind of question. And we found this one. This one and this one so that you can just do when you come along and have that kind of question. You've got a slightly shorter syrup, I think that would certainly be something that would be useful. But the moral of the story is, when you do is you, this is you have to do what Gibson? Did you have to start correctly?

Characterizing the problem and fronting up to the full complexity of it. And then you ask questions about how you go about measuring it. In order to figure out how it works. Why don't you the long as you do it in that order? Then you're asking your least

going on? I mean, cuz I think that's where the the challenges like, I mean, we're a minute or hour 40 into this and like you just hit on it like a core concept because I've been thinking about like what are the Core Concepts that that really Drive in ecological approach? And I think a lot of people like One misconception.

I've seen somebody put out there for for what an ecological approach is is oh you just make it more game like yeah or and I'm just like well it's not actually what an you'll approach is like it's about the performer. Environment relationship mediated by information like this information or energy exchange like and it's mutually reciprocal. Yes. And it's that then motivates oh some game like stuff might be a good idea, right? That's the trick, right?

I wasn't let's go game like and we dragged everything else along with it, it was going more. Game-like. Seems like the kind of seems like it might be a better way of engaging with this kind of system. Well especially if the whole purpose is perception action coupling and the fact of Attunement like to get that perception action, coupling you need to be attuned while and then you also have to add in color.

Like this is where it's like I struggle immensely to be able to deliver to somebody in this short simplistic way, all these things because I can't give you like, if I just simply told you that first line organism environment relationship, mediated by information recipient like that, you're leaving out. Other huge element that doesn't get you from there to more representative practice design.

You need the Attunement piece and then you also need this calibration, like all these things are layered in there and this is where for me it's like I somehow need to be able to help people understand. Like the appeal of it. So to back up into maybe I don't know what we want to close on because I want to be respectful of your time as we could go on for for quite a while here. Plus the plus, these other guys have may want to jump in with something.

I mean, I do want to potentially hit on the whole like Because we got, we started touching on it and I took a different affordance. It went backwards in time instead of going in the direction where we were at of your, you know, we're talking about like predictive models, you know, for scouting and Etc. So, if these things exist and we use them, you know, to some extent. Like there's there's utility and using them. Why is it inappropriate to then? Take that notion that?

Because we use something like a predictive model and whatever for Be able to do things in life and to make decisions and to base our actions on. Why is that a poor use? When it comes to how we, you know, begin to interact with like when it comes to sports and like, you know, because people have talked about, like the brain is a prediction machine,

like it's designed to predict. And I've even in the 4E space, I've heard a little bit of like, well, the brain actually predicts something like its internal State, whatever. And so like it. And so I'm like, okay, I can see a little bit. Where you're going, but this is where I guess again I'm trying to as soon as you give an opening, somebody who is for a predictive model is going to

stay there. And so that's where I'm like, how do you build a good case for perspective control versus here's here's my take on that because it's a good question. My take is that? Yeah, so you can do prediction under certain circumstances but the things you need to know and the things you need to be able to do in order to be able to do Fiction and not the kinds of things that proceeding acting organisms are able to do when they're acting, right?

So in order to be able to make an, excuse me, not to be able to make predictions. As your baseball Scout, you have to take a bunch of measures off of a person and then relate them to a bunch, of course, toward set of knowledge about a bunch of other things and how those things relate to Future performance. And so the evidence shows that actually, that's not what organisms do the evidence actually shows, the evidence is very strong Wrong for Prospect

control, at least on that. And, you know, ongoing actual perception, action Loops you don't see any you whe, When you go comparing these things, you see you don't see any evidence of broken. Why? Because the things you need to be able to do to be able, to predict organisms, can't do, don't hate you. Don't have access to all that information. You don't have. And so, can you give a concrete example for the nice ears? So, right, well, there's a right. There's a concrete example.

It's my favorite concrete. The fundamental sort of idea of prospect of control. Is that right? So say the classic experiment for this from 20 years ago. So is you've got your hand, you're on a slider and removing it from side to side and your job is to intercept an incoming Target, right? And so what you do is you have a look at the pattern of the movement as the as the targets coming in and you compared to two possible ways of doing this.

The prediction model is that you can see it coming. And so you predict where the thing is going to Go and move to there. Because why would you do anything else? If you successfully predicted, whether things are going to go and you want to intercept it? Why would you do anything else, right?

So what you should do if you're predicting a simple move to where you predict the thing is going to go the prospect of controllers about coupling yourself to an information variable and if you have a look you can you can if you can take a candidate information variable and you can plot out the value of that information variable over time. And what you can find is that that You know, as it changes its

value over time. As the thing approaches you that variable tells you that you should be doing something. And so, you get these two predictions, either you go somewhere on and get ready to get something or you move and response to the value of this variable.

And one of the weird things about these, nice interception tasks is that, if you are doing Prospect of control the information, the way the information variables change over time, you often end up getting what's called a movement reversal, required by the information variable. Where are you /? Sure Shoot where you need to be and you have to come back and that's dictated just by the value of the information variable movement reversals are

a dumb idea. If you're if you're doing prediction, why would you why would you predict to the wrong place? The why would you make a prediction, go to the wrong place and then have to come back to it, right? Takes too much time and one of the one of the signatures of process of the existence of prospect of controls. The fact that movement reversal show up all the time and interception text, right?

You can be doing a simple task. This in the lab you can be a goalkeeper so Kathy Craig does a lot of really cool work around goalkeepers and soccer where if you watch a goalkeeper they'll often go that way and then come back that way. But good goalkeepers don't start moving until later so that they wait until the information doesn't end until the

information is gone. Past the point where it's going to require a reversal because it means that they can then move more quickly, but then they have to move more quickly. Lee because they've got less time to respond. So there's all kinds of interesting things. So the base of all the evidence showed basically you get a movement reversal if you are assembling your behavior in real time, as a function of the constraints that you have

available to you, right? And if you are if you are assembling your behavior out of a series of constraints and as those constraints change and evolve over time, then that projects a very particular way of getting to where you're going to go and the evidence you know. Liability and regularly shows that that's what you're doing, right? We're as if you're predicting, then you would do something

different. You get all kinds of examples about this is the classic is the outfielder problem, right? How do you go about tracking a and catching a Fly ball, right? And you know, the evidence is all in favor of some sort of information coupling because the shit this is the difference again. Actually this is the other big difference a predictive account tells you an outcome.

You will end up here. The prospect of account tells, you sure you'll end up there, but you'll get there via this particular pattern, this particular path because you have coupled yourself to an information variable that's changing in a certain way. So it's a it's an out, it's an outcome versus process of production and the trick is that for a long time, people only measured outcomes but other people arrived, at the same time they must be predicting.

It's the only way to get to the Future State, then they started looking at process. How did they get to that particular point and the The devil is in the detail. The devil was in the how. So, if you want to convince people about these things, you have to get people out of the mindset of. What did they do? What to do? You know, where did the person end up and the evidence in favor of prospects of control, is in the, how did they get to that particular place?

Right. And then exactly how they got there. That gives you information or that tells you about which information variable and so a big chunk of this is effectively what that Tells me is that perspective, control people do prospective control because that's all they can do all the time. The only thing we are able to do is to is to organize a various components, under the operation, of all the various constraints that are operating on the Sony. Some of those constraints are

physical things like gravity. Some of them are dynamical. Properties of are, some of them are perceptual and informational, right? But you bring all of those things together. Behavior emerges out of That and there's lots of reasons to think that that's the behavior is emerging out of its coalition's of constraints and the predictor versus perspective controls the jest.

But it can't possibly be anything else because Movement rehearsals are a costly right perspective control makes you have to move further, which means you might not get there in time for example. So if you could predict maybe it would be better and the fact that people don't do it as a bit of a head that maybe they can't. Maybe people don't have access to all the things you need to be able to do in order to do a

prediction. Maybe always got us access to a bunch of things that are the kinds of things that enabled you

to do prospected for. All right, so yeah, I tend to take all of those that kind of experimental evidence as fairly strong evidence that we've got no choice but doing Prospect of control, because we don't have access to the required, to, I think you answered this, but I think people might have glossed over it. The one of the reasons that people Are attracted to the predictive model is because of the fact of like processing delays.

Like how long it takes because, right, it's this because there's a couple of things on that of like, okay, you have your motor visual delay, but also, there's also this assumption that the brain is the thing that causes everything to happen, right? So like you send like, it has to be triggered first from the brain and I don't think that's actually how it works. And I haven't, you know, like I've understood certain things.

Maybe like an, I haven't found the the research, but I thought there was something talking about how oftentimes the system can can respond locally first. And then I also heard like David Stone talked about this often times to the brain, will simply come in to check whether or not what, what, what actually did like was what you was towards your end. And this is, I guess the other element to of like even by your

response. Ask the question of like what do you think the role of the brain is especially from an ecological perspective of like, if there's no. Representations yet, we have this ability to imagine things like all these things is. This is why I think it's really hard for people to swallow the ecological perspective on this because of because it doesn't square with some of their own

personal experience. And I think some of it has to do with language though and it's the subtlety of language. Because even as you were talking about predictive versus prospective, even in talking about the research, you use the word predictive if I understood correctly or prediction to even talk about the research in terms of the research, Findings and that I think even confuses that even confuses people further.

Like this is where like languages to what you talked about before earlier of like languages. This very complicated thing to work, through to communicate ideas to that, then shape Behavior because the whole reason to even because people are probably like, why does this even matter? Because if you don't like it, why why it's so important to parse this out is because this

actually shapes your behavior. What you do practically once you do practically a Shape by your understanding by your beliefs. Like and if and that's why it's me it's so important to have this conversation into like really Ask people to pay attention to what is being said? Yes. Which was July, which question is sorry. The addressing the The issue of the the motor visual delay and why people want to grab onto a predictive model to to deal with

that problem? Yeah, so the basic idea there comes from the, you know, the notion of just studying reaction times, right? So I can show you a really simple setup where I facial life and get you press a button and there's a delay that's right. And that seems to be the case. There is tight takes time for information to enter the eye and get to the brain and then have that.

Turned into into our response. And so part of the answer to that is well, that's us really boring tasks that you're asking the system to do, what happens if you ask it to do something more interesting. Like try and catch a fly ball, right? Or you know, try and catch a line drive. It's coming straight at you,

right? And so, one of the things that's really important to note, this is one of the things that the ecological approach likes to shift into this, to stop looking at failures of system and start looking at successes because the successes are actually quite interesting and informative. And actually the successes are the rule, not the exception, which is lucky because if they weren't, we would have died, right? So one of the really important things about this visual motor delay.

Question is that if all of our experiences lagging, a couple hundred milliseconds behind reality. Then we're all dead. All the time, instant right. Cannot you cannot run a system like that, right? So two possible explanations one. Is that there isn't a delay, another is that there is a delay, but we're operating on some sort of Reproductive system, right? So the first of all, there's your motivation for having a productive system as of the

existence of these delays. And the fact that we're not dead, seems to suggest that we need to be doing at least some sort of. So that's one of the motivations. Then the problem with predictions. And one of the key problems with predictions is that they are probabilistic and can be wrong. And prediction is actually a very unstable way of controlling a system, right? It's the kind of any of reach, if you start trying to implement predictive controllers into

robots and things like that. And various kinds of controllers, you can do it under very constrained circumstances but you get quite unstable Behavior out of that system, right? And it's really easy to, to like, you have to build like I say, prediction requires Access to quite a lot of information, right? This is Dan Dennett calls this a loan of intelligence to every Refreshments. All right, these kinds of systems need to have a lot of intelligence and capability loan

given to them somehow. And part of the problem is, where did that one come from, right? And it turns out that a lot of those loans don't have a good explanation, it's not clear how you give the system access to the information that needs in order to be able to make anything like that. So that's part of it.

Then the other part of it is well just because my reaction time in this one boring task is a certain that simple tap measuring reaction time in a really simple task, doesn't mean you've measured true Reaction Time by stripping out all the other stuff. Remember what I said? Behavior is task-specific what you've done. Is you've measured reaction time in a very boring, very simple setup, but there isn't any possibility of prospector control right? Because there is no time varying

information, right? There's just a light comes on Draw an idol is nothing but your tracking. So your reaction, time to intercept. Something like a line drive is actually a completely different task, right? Completely different task. Your coupling yourself to information about the motion of a ball that is continuously available, and removing different lens and the inertial properties of those rules.

So, the other big thing is that just because you can give visual motor delays, and some tasks doesn't mean that that's the rule, that just means that that's the context in that task. And if you made the task, so boring that A delay was inevitable that's on you the experimenter not necessarily on the organism, right? So the question is, what happens if you provide people with access to the ability to do

perspective control? Like so what happens when you track people and figure out how they catch fly balls? Well, a couple themselves information variables so that their movements are organized with respect to some future state. So the present value of the unfolding of these things over time. So you know there's other things going on here, right? You mentioned sort of local solutions to things.

So one of the ways in which I move my arm has to do with just the local dynamics of the arm, right? Not my nervous system, doesn't have to tell my arm to do everything. All right. There's a whole model of say rhythmic movement and oscillating, limb movement, or the equilibrium point hypothesis. And there's a, bolo physiological evidence. This is how the nervous system is organized there, where the brain doesn't have to say move here. Move here, move here, move here.

Move the movie as I say, you know, wrote about a motor program, Yeah, no. You just have to go get you. All you have to do, is my arm right now is set at a particular balance point. My brain just goes change the balance point and the art and the local Dynamics.

Take care of the rest, right? So, there's a big chunk of of our behavior on the straps of the detailed structure of their behavior that is taken care of at the local level, and have just has to do with the dynamic organization at that local level. So, that's another thing. Another thing there is when you go looking at the brain and considering it in terms of its Network, Structure. There's all kinds of interesting Network structures.

Things called motifs, right? So I Motif as a particular as a particular Network organization, designed to achieve a certain functional outcome for a brand Network and you can build motifs, which are network organizations that couple two distinct spatially distinct parts of the network and the couple's them together with zero timely, right? So there is neural solutions to these act textual Solutions, Network Solutions to these problems.

And so on and so forth. So the problem for the issue is that yeah, sure. Right. You have went to front up and ask these questions about but at all times the notion of a transmission delay in a visual motor delay. Being a problem is premised on the idea that behavior is a linear as the result of a lineal causal chain of events, right? Where light comes into the eye, then has to do something else, then has to be something else, then has to be something else than has to come out as an

output. That's not what's happening. What's actually going on? Is that you are a nonlinear dynamical system that gets up and running and then tunes into everything else that you're going on. That's going on in your environment. And that's happening at multiple time, scales supported by multiple mechanisms. And so all of a sudden the problem becomes to write that processing delay is just becomes an artifact of the tasks of an artifact of the organism.

So effectively, what you have to do is you just have to start asking more. Yeah, just ask different questions. Again the moral. So you you focus much more on process than outcome, right? You have to ask questions not about when did the thing react, but how did it go about reacting on what was available to it to react? And what will the constraints that were operating on it? While it was act. So, it's all right, kind of bring this to sort of a close. It's kind of a, I'm a Minnesotan.

So there's a thing called a Minnesota, goodbye. So, I don't know if you're familiar with that, but I'm going to kind of it. Basically you say goodbye and then you talk for 15 minutes and then you say good-bye again and you talk and so we're going to its kind of what's Happening Here. I want to throw this over to both Robert and and Baker to see like okay what thoughts? Because I've been kind of dominating the conversation.

So I want to see if you guys have anything that you want to chime in on. Oh, wait. Okay, you sure I'll keep it short and sweet. I do not have anything bad. all right, so to kind of then Wrap up. What are some good resources that you would recommend people? Like because obviously we can't cover everything. Like what are what are some

things maybe to like that? You would like to say to like the community and then maybe it's specifically a baseball Community but probably just the community at large and and then like you know, next steps for people. So to the community and I just want to say that I think I think the ecological approach has a lot to offer in terms of helping you ask and answer the You want answers to, I really do believe

that. I really believe that the people who are working, you coaching and athletes and, you know, scientists and data analysts and all those kind of things. Everybody's, everybody's working in good faith, everybody's trying to do well by the people that they're working for and with and they're all trying to find ways to do to ask good questions and come up with good answers. That enable people to have a good practice.

I think that's true. My own pitch here is that I think that the ecological Roach, broadly speaking has a huge amount of offer that entire process in terms of figuring out what what constitutes a good question, because our, what constitutes, a good measure of a behavior that can actually tell you something. I think that there's loads of loads of value there to be had. And there would be, you know, part of part of my job is just trying to generally raise

awareness. And make people understand that there are options out there when you're trying to figure out how to think about Behavior. There's more than one way to think about it and the ecological approach is a is a substantive player on that field, right?

When other friends little group were never friends, we've been out an atom and going for a long time and we have a lot of reasons to think that the things that were saying in the things that were advocating for our MD, how these systems, these Behavioral Systems work. So I think we, I think we have a lot to offer and it's I've seen that a lot lately where people are wandering around. They've got really good questions, but just don't know how to go about asking.

In terms of resources frankly if you're if you're if you're in the sporting world and you're interested in finding out anything about the ecological approach I can only recommend Rob Graves podcast and book, right? His podcast is a remarkable resource that he's been producing over the last five or six years. And it's it's an incredible contribution to the field. And it's accessible, right? It's really accessible.

He robbed spends. A lot of his time and there's lots of different ways of going in. And so he goes in the interview and he talks to people scientists petitioners Like that. So you can go and listen to people who are trying to make this stuff work and listen to how they talk about and what they're trying to do and how they approach of them. But he also spends time working his way through kind of

theoretical issues. In theoretical topics particular questions what is calibration and what is direct perception? What is direct learning Etc, where he's he's trying to walk people through the basics of that and again like you're going to have to do some work. If you're interested in this and you think it might be have, you're going to have to give it some time and you're going to have to come to it, prepared to put a bit of effort in because there's no other way to go about

doing that. But if you're if you are then we can try and you know they would that conversation can be productive with and so like in terms of resources in terms of first steps, I always recommend rods podcast is just a really good resource. I mean yeah, I think those are the best things really to recommend right now and then once you've got that foot in the door, come talk to me and I'll tell you the next. Well, I mean, I think that's

that that actually right there. That last piece is the is the thing that I've learned, and I think maybe you can speak to this. I have this fear of You know, talking to people reaching out because like from I'm only where I'm at today because of the fact that I've had such close interactions with with Sean and Tyler like it like I wouldn't because if I were just reading the literature by myself I guess they help me attuned to what's more specifying.

Yeah. And like otherwise like because I look at for example Sean where he's at right now then when you do it on your own like it takes X number of years of like Acting with the information, not that Sean hasn't been talking to people and whatever, but like he has done a lot of it on his own trial and error and like we get the benefit of his in some ways. Like, for me personally, I get the benefit of his mistakes. Yeah. And like, I'm able to shortcut, like, for me, I got to where I'm at.

That's I'm not saying I'm aware of Sean's at, but I'm a lot closer to where Sean like I much further on that path than like, where Sean was potentially is you know, And because it because of the fact of the interaction that I've gotten a shortcut, these things and that I think is like you don't make the mistake that that one organization did of butchering, the constraint. Let approach if you actually talk to people.

Yeah. Who are in the industry, like who have developed these ideas and Concepts that gets corrected. Sorry. You're going to say, well, I'm just, I'm just and as an educator right out of my jobs, being a scientist, but part of my job is being an educator. And as an educator, you're exactly right. Like, I've benefited from that as well, you know, so, my PhD supervisor supervisor, Jeff Bingham student of turvey, right site.

So, terribly worked with Gibson and learned and figured out a whole bunch of stuff, and then he told it to Jeff. And there was, there was indeed that shortcutting, and that's the whole point, right? And then, I got to say, I got to learn from Jeff and then, I've learned over the years of talking and interacting and working with people, as well.

That's the whole point, right? So in terms of reaching out, right, reach out to us because You know, I'll just sit here and chat with you guys for a couple of hours. It was fun. And I've learned, I've told, you know, I've learned a bunch of things myself, right? I get it, you know, I'm trying to get better at articulating. These things, you guys have said, some stuff that's made me stop and go. Oh, that's interesting. I can put that up with that, right?

This is, this is, this is part of the job. This is part of trying to figure out how to get this to work, is in this communication. And yeah, it, like the whole point is that it's not up to you to reinvent ecological psychology every single time, right? You get to, you get to stand on the shoulders of those. That's the whole point of having that existing literature but it's an existing literature, right? It's a bunch of papers. It's a bunch of books.

It's got to be brought to life somehow, right? And so one way to bring it, you know, I try to bring it to life in classes and Rob tries to bring it to life and his classes, but also through his podcast and other people. Other coaches are trying to bring it to life through the papers that they publish, but also through their practice and their activity at their, their clubs, right?

So there's lots of different people trying to bring this stuff to life and lots of Wax and different people responding resonate and they're interested and engaged with different ways of bringing those things life. So and we're all we all kind of give a damn. We all think this is a good idea and it's not just that it's a good idea but it's an idea that has value. We think this is a we think this is a really good way of going about doing and it's going to

for me as a scientist. It's going to teach me things about actual mechanisms of perception action systems and for a coach, you know, people get into it because they think it's actually going to win. Able them to do what they actually want to do. Just to help support their athletes, become the best, they can possibly be and so on and so forth. So yeah, we're all in it and we all want to be talking about it and we all want to be engaged.

So, you know, that's the other thing just reach out. And, you know, I'm on Twitter and you can drop me an e-mail. I'm always happy to have this chance, because it's fun. It's part of moving this forward and keeping it alive.

That's the other thing, right? Is that, it's, it's important that people can Like yourselves, keep coming in and keep coming in with, with fresh ideas and fresh understandings, and fresh questions, and problems that we're trying to address so that we're spreading out and trying to tackle more more, and more complicated and different things so that we're actually building some, right? I think that's really important and we're all world came to make sure that happens.

So to kind of land a plane where can people find you reach out? Like, what's your Twitter handle potentially email. Like, where's the easiest way for people to kind of reach out to you? What is a pretty reliable place to find me these dice on acts like scientists on Twitter and I have a Blog. It's called notes from to Scientific and psychologists which has a long history for various reasons, but that's what it's called. There's a boatload of stuff that I've been developing are over

the years and writing and yeah. So I'm a reader in Psychology at Leeds, Beckett. University and Leeds United Kingdom. And so, my email here is a DOT Wilson at Leeds, Beckett, but ac.uk and I have a lot of conversations and with people and I'm always kind of happy to try and find some time. You know, he's a busy teaching of started above things going on as well, but I'm usually happy to try and find time to talk. It's part of my job.

I consider it to be part of my job but also it's fun and it's exciting and Well, thank you so much for coming on and talking with us for for a good length of time. Thank you so much. Andrew filament meant a lot. Learned a lot here in this conversation. Yeah, appreciate it. No, this is good. This is a really fun fun chatting part of