Ep. 378: Repetition and Representation, feat. Cal Jones

Hey everybody, before we get started this week, I just want to let you know we released a new mindset course featuring Rob Bernaki from Island Top Team and BJJ Concepts. It's called Mindset for Betas. It's an amazing resource that breaks down a new way to build a resilient jiu-jitsu mindset. It's part of BJJ Mental Models Premium. I will spare you the full sales pitch because you can try it for free. Just go to BJJmentalmodels.com/beta. I will give you a free month, you can check out the course and if you decide that it's not worth your money, you can cancel, you won't have to pay a cent. I've already been told by subscribers that this is the most valuable piece of jiu-jitsu content they've ever received, so I hope you like it too. Speaker 1: Hey everybody, welcome to BJJ Mental Models episode 378. I am Steve Kwan and BJJ Mental Models is your guide to a conceptual and intelligent jiu-jitsu approach. And I'm here today with probably one of the most recommended guests in our Discord and in our community. People keep asking me to get you on here, man, so I am happy to have you. I've got Cal Jones on the line. Cal, how's it going? Speaker 2: Oh, wonderful, thank you. It's great to be here. Speaker 1: Happy to have you as well. Now, you are probably best known for your coaching work in the constraints led approach CLA space, but maybe it's best suited coming from you. Why don't you tell everyone a little bit about yourself? Speaker 2: Yeah, so I've grappled for, well, 32 years, uh, predominantly a judo coach, but for the past, what would it be, maybe 15 years, I've done double the amount of sort of BJJ or submission wrestling coaching. So we do a whole bunch of nogi. I've got a master's in, what's, I think it's called advanced coaching practice. But I was fortunate enough to do that in Sheffield Hallam. So it's essentially where Keith Davids is based. So we kind of grew up on a diet of ecological dynamics. It was in the blench food there. So, yeah, that's kind of become my wheelhouse. Speaker 1: Nice, nice. And as you implied there, you're likely best known for the advice that you give in the ecological dynamic space. And this led to a kind of an interesting topic that you wanted to discuss here. I thought it would be pretty fascinating. The idea of repetition versus representation. And we'll start off maybe by breaking that down to the most basic level so that people understand the concepts and then we can build on top of that. But first, maybe why don't you at a very high level, beginner level, explain what the concern is here and how that impacts people training jiu-jitsu. Speaker 2: Yeah, so essentially, the, I'm sure people will be familiar with the, the debate, the eco debate already. But the traditional approach that people utilize in a lot of coaching is decontextualized from the actual performance environment. So you'll end up doing dead rote repetition of a skill, of a move, and hope to build some abstract muscle memory or some built-in conditions in your body that you can just press play on what in the literature is referred to as a general motor program. And you can just do a skill but in context. The literature is overwhelmingly against this. It seems to be the case that that's not really how skill is developed. The debate in academia still exists between whether our brain is doing stuff or if we don't need our brain to do stuff. But both sides of that debate both agree that we don't need to be doing rote technical rehearsal of an idealized mechanical model. We need a more representative environment for us to be able to engage with so that we can start to develop skill rather than just this concept of technique. So yeah, the, I think what would be really fun to talk about is how we can start to go up and down the scale of representativeness, which is an awful word to say a bunch of times on a podcast. But how we can go up and down the scale of representative to make sure the type of tasks that we're making use of in our training has the best retention and transfer to actual competition. We're not just doing a move over and over and over and over and over again. Speaker 1: Got it. Sounds good. Now, to maybe dumb this down a level for the beginners who aren't familiar with the terminology here. I interpret this as a battle between how many reps do you want to get in in terms of repetitions versus the quality of those repetitions. How many of those repetitions accurately reflect how you would use this move in training? And the common example that I've discussed many times on this podcast is some of the things that people do during a, I guess you could call it a traditional jiu-jitsu warm-up. If you train at a traditional jiu-jitsu school, maybe your warm-up entails things like shrimping down the mats where everyone lines up and they get on one side of the mat and they have to shrimp and carry their butt across to the other side of the gym. That's a very common way that people attempt to turn jiu-jitsu into a warm-up exercise and coaches like it because it looks all fancy and disciplined. But the feedback that I have always had about that approach is it gives the appearance of jiu-jitsu, but it's not based on the same fundamental concepts as jiu-jitsu. And in fact, doing those kinds of exercises can sometimes give you the wrong impression about what you need to do. So, as an example here, if you are trying to shrimp down the mats, your goal is to move your butt as far as possible every time you shrimp. You're basically shrimping for distance. And that's slightly different from the goal when you're sparring. If I am in a jiu-jitsu match and I need to shrimp, my goal is usually to make a bit of space so that I can regard or something to that effect, get my legs back in. And that doesn't always mean that I am shrimping for distance. If I try to move my butt as much as possible, I can actually wind up creating big openings that allow the other person on top to take a more dominant position like mount or neon belly or something to that effect. So, if the only thing you know about shrimping comes from that drill where you got to move your butt down the mat, you might think that it's good to make big movements when you shrimp. In reality, often good shrimping is little movements because the more you move, the more space you create. So, I would see this as an example of what you're talking about where the the classic drill is all about getting in the most reps and from a repetition standpoint, so when I say reps here, I mean repetition, right? From a repetition standpoint, it meets that goal, but from a representation standpoint, it's not really achieving the goal of actual shrimping in a sparring contest. Do you agree with that explanation? Am I on the same page or am I totally whiffing here? Speaker 2: No, no, you're right on the money. The important thing to be very aware of though is that there is kind of a lower threshold, the lowest threshold at which point training below that level becomes a little bit redundant. So, it's not that it's always useless to do stuff where your opponent's passive with their brain off or there's no opponent. Sometimes there's utility there. But how we make use of those sorts of practice tasks can be wildly different to if the assumption is we're developing a move, storing it in our brain and being able to press play on it when the time comes. So, an analogy I I like to use is if we imagine we have a bowl full of jelly and I just pour it on a table, it will just slosh everywhere. It has no shape or form. But if I pour that exact same cup full of jelly into a mold, then it will start to take on some form of constant shape, right? So, what we need are some parameters that can actually add context to the movements that we're doing. So, me just spamming a shrimp up a mat isn't actually developing the skill of shrimping. It might be developing the thing that we refer to as effectivities or action capabilities. I might be able to better move my hips if I am a 70-year-old accountant who has just come from the longest day of work that they've ever had and they have the hip mobility of a 70-year-old accountant, then maybe the unit of engagement that is best and is most appropriate for them would be, can you move, can you actually move your hips like this? They can't engage with developing skill if they can't move that way. But the second they have the capacity to move in that way, doing it ad nauseam isn't developing skill. It's not helping them. They need some sort of a mold that they can have around them to start shaping how that movement becomes skill. So, there is definitely a minimum threshold at which we need some kind of engagement from the environment that we're in that adds context to those movements. The problem that we have, or the problem that I seem to see quite frequently is that a lot of people are using tasks where they have enough, we refer to as specifying information in the literature. There's enough information from an opponent that you can start to coordinate what you're doing based on them. There is a mold present that is shaping how their skill is starting to form. But they don't go up and down this scale of representativeness to start testing it under more and more representative environments. They'll have played a practice task where they're looking at the skill and then it's open mat. Can you do it against an opponent? And there's no sort of gradual hand-holding and a general introduction of more and more representative environments, which then comes to your trading off representation for repetitions. If I only entered competitions to work on my double leg, I might go to a contest and do three in a day. The three that I do, I know work in competition because I did them in competition. They have to work in competition because they did, they worked in competition. But I did three of them. Whereas if I play a little practice task where I'm looking to try and get my knee down behind my partner's hips, for example, I can make that action multiple times in a practice task and I can start to understand penetration steps and level changing and all of these constituent parts that become a double leg. But that does not transfer into skill until I start layering up more and more representative environments. Does it work when my opponent has different connections? Does it work when I'm also allowed to do other types of takedowns? Does it work when my opponent can do other types of takedowns? And it seems to be the case that there's not that care and consideration of going up and down the scale of representation from a lot of the stuff that I'm seeing. But again, I'm not the most active person on social media or I might just not be seeing a lot of people that are doing this. Speaker 1: Well, I always want to cut through the noise here and help people understand. And sometimes that means simplifying the mental model a bit so that people kind of get the context regardless of the complexity of the language. It's always a bit risky to do this because I don't want to lose any meaning. But when we talk about this spectrum where on one hand we've got repetitions and on the other hand we've got representation, in a lot of ways this sounds like a quantity versus quality debate where you can focus on trying to maximize repetitions, which would be quantity, or you could focus on making those repetitions as close to a real-world match as possible. And that I guess would be the quality, the representation. Is that a a fair simplified way of looking at this if people are struggling with the language? Speaker 2: Yes, except for the fact that there is a minimum threshold at which we don't really want to go underneath. So, for example, if I'm practicing shrimping with no opponent present, the retention and transfer of that skill for actually developing an effective shrimp in context doesn't exist. The time that I spend on that action becomes pretty redundant. I could do a million shrimps like that and then be put into a position that I have to shrimp against an opponent and I will have about as much of efficacy as a person that has never done it. It's not contextualized. We need to get to a point that there is some form of information that we are moving in relation to. In the literature we call it perception action coupling. I need to be moving in relation to something that informs that movement. Every movement that I make has utility and is a function of something that my opponent is doing or an opening that I can see. This is where the concept of affordances comes in. When I'm looking to try and break somebody's posture, I could just do traditional, I'm going to pull you into me while you stand there and think about what you're going to have for dinner. But that's very, very unlikely to transfer. Whereas if I'm playing a task where I'm just trying to turn you into a banana, then you're trying not to be turned into a banana. I'm actually developing skill. There's me having to understand haptic feedback, how I can use my hands to actually coordinate those levels of control against an opponent that's resisting. So, there's a minimum point at which the types of tasks that we're using start to develop skill and under that, it doesn't really help. What I'm saying is if we plot it on a scale, so just drilling a thing would be a one and competition is a 10, we probably want to be spending some time at maybe a four, some time at a six, some time at an eight and some time at a 10, rather than what seems to be happening is people have recognized that spending time as a one, two and three, where there's almost no information from their opponent for them to start to coordinate structures around. They've recognized that that's pretty bad, but they haven't started to thread together this link between tasks that are somewhere on this scale and add more and more representative tasks in to make sure that the skill sets that we're developing become contextualized within the actual sport itself. Does that make sense? Speaker 1: Yes, it does. I like how you're explaining that there can be value in repetition with lower levels of resistance because when you hear people discuss this, often they break into two camps. One being the that everything must be 100% eco all the time. Everything must be 100% representative. But then the opposite camp is, no, it doesn't. And I've always felt that there is a value beyond even just motor learning of doing repetitions for the sake of repetition's sake. I mean, if you think of something like uchikomi from a lot of other martial arts like judo, there can be benefit in doing a degree of completely dead drilling, mostly at the conceptual level. If you're so new at something that you don't even understand the objective of what you're trying to do, a bit of dead repetition can help. And additionally, it also allows you to build up confidence in that technique. It can be hard for some people, especially non-athletes, to have confidence to do something and they can be a little bit hesitant and gun-shy. And a few dead reps can sometimes help them build that up. But we want to start moving people away from that and towards more realistic applications of the training. So, I like this idea that you're talking about where we view this almost as a spectrum and not an on-off switch. Again, am I understanding correctly? Speaker 2: Yes, ish. But I guess that the big difference is that it's not this sort of mixed methods approach, right? It's not that drilling is this IP traditional thing and should never be done by eco bros. It's a cardinal sin against the constraints led approach and Rob Gray will come to your house and punch you. That's not really what we're talking about. It's that the utility and function of different types of tasks have different weighting depending on our theoretical understanding. So, for me, the more parsimonious explanation of how motor learning occurs, I believe in direct perception, perception action coupling, the soft assembly of the degrees of freedom to achieve a task goal. So, if I decide that I want my players to do a couple of drills, I'd contest the term, but we'll say drills just for the moment. I want them to do an unopposed version of the movement structure just so they start to understand how they can move their body like that or so they can get a coordinated structure in place that they can then start to explore in context or for safety or to build confidence, then that's absolutely fine. It's not me saying that I'm trying to build muscle memory in people. The like analogy I like to use for this is if I've made up a game that I call front flip tag and all you need to do is tag another person, but before you can tag them, you have to do a front flip. Well, I can't do a front flip. So, me having an opponent running around trying to dodge me does nothing. That's not the engagement level that I need. I need to learn how in the hell do I coordinate this fleshy mass of flab to do a functional front flip. So, if I just do an unopposed, no opponent series of jumpy, twisty, flippy things, that will probably be a better use of my time than me running after somebody and trying to do a front flip while I'm trying to learn how to make interceptive movements, how I can judge their change of direction, all of these things for a sport that I've just made up. Right? The time that I spend just going through so I can get a functional coordination structure probably has some value. Once I can move my body that way, doing a million front flips isn't going to get me better at front flip tag. I can do the thing. I can now do a front flip. Now it's about how do I actually use this sort of coordination structure to achieve a goal. There's now utility to my front flip, right? This is the problem is traditionally, the assumption of doing a drill is to develop a kinematic structure that becomes muscle memory, becomes a thing that somebody possesses that they can then start to apply in context. I've done 100,000 knee cuts, so I'm hopefully going to be able to knee cut from every single situation. That's not really the point. That's not how skill works. But if I can't move my body in such a way that I can achieve a thing, then doing a couple of repetitions against an opponent that's stood and thinking of dinner isn't the end of the world. Berimbolo is the classic in BJJ, right? If I can't berimbolo, I probably don't need my opponent trying to pass my guard. I probably need to learn how in the hell does that work? Like what am I doing to even move like that? So, I could create a very clever little practice task where we're getting them to explore those kind of behaviors in context. Or I could say, you've got two minutes. This is roughly the kind of thing I want you to try and do. Have a go. It looks a bit like this. Your partner's going to stand there as scaffolding. The thing that I need you to learn to couple to isn't what your opponent is doing at the minute. It's coupling to how do you move and negotiate space with all of your limbs and joints to achieve something that looks a little bit like this. But once I can berimbolo, I don't need to do 100,000 of them so that I hope I can do it against an opponent in context. So, there is some value. It's just what the utility of that thing is changes massively based on the assumptions that I have of how motor learning actually works. It's not building a general motor program that is stored in the brain. It's this functional coupling between what I am doing and my opponent in context. Their behaviors and my behaviors need to align and there needs to be this coupling between them. But if I can't move that way, I probably would benefit from learning to move that way. That's an important thing. I appreciate I'm waffling on, so I do apologize for not stopping, but I'm having fun. The other thing is that all coaching isn't skill acquisition, right? We're coaching, skill acquisition is a large part of it, but building athletes that are safe, that aren't afraid of putting their hands down when they're being flipped over their head and breaking their collarbones is a real concern. Having people who, having a gradual introduction to the floor so that they don't get winded is a real concern. Having people that feel confident that they can do the move without being dangerous is a thing. So, if a coach at any point says to me, I choose to do this thing because I think it builds confidence and security in the people that are practicing this move, I'm fine with it. Like I'm not judge, jury and executioner and their rationale isn't that it's building muscle memory or a general motor program or developing skill. If they said they're doing it because they think it's developing skill, then I can have a discussion with them on the context of that as a task. No amount of uchikomi is going to get me to be a good thrower of people. But I might, I never would, but I'll make the argument. I might choose to do uchikomi to develop my shoulder stability. I've come back from a really bad injury. I'm not ready for contact training yet, so I'm just going to do uchikomi. Maybe that would be a decision I'd make. Yeah, hope that makes sense. Speaker 1: Makes a lot of sense. And I think for many people who are trying to wade into this eco debate, they will find this to be quite useful and helpful, especially to hear this from you. It can be a bit challenging sometimes for people to figure out how to apply this stuff. And if you listen to coaches who are really into the ecological approach to jiu-jitsu, they sometimes act like motor learning is the single most important and only important thing in coaching. Whereas so much of coaching is about confidence and relationship building and even just the exchange of ideas. And to act like the only thing that matters is the motor learning aspect of coaching really undermines the entire profession in my mind because so much of what makes a good coach is around the ability to steer an athlete's mental and emotional state. That skill set in many ways is more important than actual motor learning practice. There are coaches out there who don't really know that much about how to actually teach people skill in jiu-jitsu, but they're really good at motivating people. And if they get the right athlete in there who has the right level of self-direction, they can compensate for that lack of knowledge because the strength of the relationship that they build with that athlete is all that the athlete really needed at that level. Now, I think that the best approaches are a broad spectrum approach to coaching, someone who understands all of these disciplines. But my point here and my takeaway is that you will sometimes hear people talk about ecological dynamics as if the only thing that matters in coaching is motor learning. And I think you're pointing out that these other things, they are tools in our toolbox that may have a time and place beyond just perception action coupling and building those neural pathways that help us actually understand how to move our body. There's more to coaching than just that. Speaker 2: Yeah, we have what we do and how we do it, right? What we do is underpinned by ecological dynamics. Whatever tool I choose to use, the rationale for me has to actually make sense based on my understanding of the literature around motor learning. It doesn't mean that some tools are forbidden. Like the book that introduced the constraints led approach spoke about using demonstrations and informational constraints. Showing people what a move looks like, having them do a couple of them. It's literally in the foundational paper that introduced the constraints led approach to people. It's not like a controversial position to be taken. It's how that is used can be very, very, very, very different. If I show you how to do a move and I'm telling you, like hokey cokey coaching, it's this, you put your left hand here, your right hand there. If I end up doing that as a type of coaching, then there's probably going to be some very, very big problems. We don't have these idealized models, like kinematic models. The amount of variance in how a thing can be achieved is massive. The true invariants that exist for a move are very, very small. Like a lot of the time people are using the term invariant in this erroneous way to mean something that might be advantageous. But that's not what it is. It's something that has to happen. An invariant is a must. So, if I want to pick you up, I could bend my knees and straighten my knees to lift you. Or I could keep my knees completely straight and just hinge at my hips and lift you. The bending of the knees is not an invariant feature. The invariant feature is just that your feet come off the floor. So, we end up with this like tautological thing where the description of what is happening is very often what the invariant feature is. And how we go about achieving those task goals can be so different. So, if I'm saying this is the move, go and do it precisely like this and we're going to do five of them. So, we're replacing drills with here's five rather than 100. Then that would still be a big mistake in my opinion. We're constraining people's search space to the point that they're not finding a functional form for them in context for how to solve the problem with a skill, right? It becomes this prescribed movement structure. The way that I think we should probably be doing this most effectively is by introducing some form of a practice task to people and having them engage with that task. And if they are struggling, if there's some area that they're not able to coordinate lifting their opponent's leg and twisting them to the ground. Well, at that point, I might introduce some sort of kinematic structure to show them, here is a thing that we call a single leg. It can look a bit like this, it can look a bit like this. The important thing is that we're displacing their weight past the point that they don't have a base of support, so that their center of mass has gone past their base of support and they fall to the floor. Try two or three. Have a go. Can you do it to your opponent when they're stood doing nothing? You can, amazing. So, at this point, you doing another 50 of them is pretty redundant. But I've now framed your search space. So, when you go back in and you engage with that task, there's now some intention that's been shaped into what you're trying to do. You become a little bit more cognizant of the ways that you can coordinate movement to try and achieve those goals. So, if a lay person were to look in and see a room full of people that are doing a single leg and their opponent's stood doing nothing, they could make an assumption that it's a drilling class and there's, we're a traditional IP group. But the way that tool is used is wildly different. It's not as if drilling equals traditional. It's how we use the tool that is different. The degree to which we use it for how long and how it's used. Yeah. Speaker 1: Got it. I like it, man. I like it. Now, you talked about invariants here. We've talked about this before on the podcast many times, but it's a really interesting concept for, again, determining what is universal about jiu-jitsu versus what is flavor text, so to speak. Could you maybe quickly explain what is, and I know you already did briefly, but maybe elaborate on that. What exactly do we mean when we talk about invariants and maybe what are some examples that could help people relate to this concept? Speaker 2: Yeah, so essentially, the concept of invariance, as far as I I'm concerned, isn't that important to me. I tend to think that way people have been using them has been really, really loose and liberal and not in accordance with the literature a lot of the time. So, there are concepts that exist, right? If I want to do a heel hook to you, I need to create rotation of your foot to cause your MCL or ACL or LCL to get ripped, basically. If I put both of my knees above your knee and can secure your knee so it can't twist, that would lead to me being more successful more often. But it isn't invariant. I could just grab your foot and twist it as hard and as fast as I can and you would end up having a very, very uncomfortable afternoon. So, as I think for me, invariants really, really often are a tautological thing. It is the thing that gives a technique its name is the invariant feature of it. So, an armlock, their arm needs to get locked. I don't need to have control of their shoulder line. I don't need to have any of those sorts of things. If I just hold your wrist and I punch your elbow forward, your arm will break. There has to be control of the joints either side of the joint you're attacking, but how that looks can be vastly, vastly different. The same thing for double leg. I need to take both of your legs and make you fall to the floor. There are ways that we can look at that kinematic structure and make it more or less efficient, but we end up with these tradeoffs that occur that maybe if I want to do a hip throw on you, maybe me bending my knees will let me recruit a more effective lift mechanic. I can lift more weight that way. But if I don't bend my knees, I keep my legs straight and I just back into you, maybe it means I can attack from a further range or I can use some disguise to do it or I can catch you from a bit of a janky angle. The invariant thing is just that your weight goes up and you're lifted, right? So, for me, the, well, for what invariants are in the literature is a must. It's something that has to, has to happen for a thing to occur. If it can occur without that thing happening, that thing is not invariant. So, in terms of how it is often used in BJJ, it can be very, very loosey-goosey where it's, this is a useful concept for us to understand. Me having control of your shoulder with my knees probably means an armlock is going to be more successful more often. But again, I can create those layers of control without me having to do stuff. Inertia exists. I could just punch your elbow forward and it will probably break. Yeah, it's a, it's a concept that I think is very, very often has been very muddied in the BJJ space. Speaker 1: I want to have some actionable advice here for people of all levels. And maybe a good place to start is with students listening to this. If you are a relatively new jiu-jitsu practitioner or at least new to ecological dynamics and you want to know how can I apply best practices to my training? I don't have control over the curriculum or what my coach does, but I can control how I train. What would you suggest for most grapplers, especially those new to this, that they do to better accelerate their motor learning in jiu-jitsu? How could they take this idea of repetition versus representation or I guess quantity versus quality and front load that into their own training practice, even if it's not something that the coach at their gym particularly cares about? Speaker 2: Yeah, it could be very difficult, of course, if you're at a gym that is very drill heavy. You are kind of stuck. There's not that much you can do while a drill is occurring. You can kind of maybe whisper to your opponent, don't just sit there passively, can you try and stop me? And maybe they can try and do it, you might get away with it. Whether that's disrespectful to your coach or not is a different question. But essentially, the overwhelming majority of the time that we spend doing any sport on earth, we should be trying to develop the skill sets that we're looking to try and develop in context. You should have an opponent who is opposing you or you should have similar conditions to the conditions that would exist in a competition setting. Knee cutting a person who's lying down on their back thinking of England isn't going to help me. Me trying to knee cut somebody that is adding frames or trying to move their hips or is doing stuff means I have to actually start to add form to the jelly that I'm just slopping on a table. It's now being slopped into a mold. There is intent and functional purpose to the way I'm moving, so there is this perception action coupling occurring. The big thing is, can we start to play certain types of tasks where we're engaging with an opponent that is offering more and more representative behaviors? I don't just want to play a game where we are replicating some facet of a behavior we're looking to experience and then the next I know I can explore that skill is in an open mat situation. We need to start adding more and more and more representative behaviors in. So, my recommendation is that we follow something I've just been fortunate enough actually to have written a chapter in a new ecological dynamics textbook on coaching BJJ using the constraints led approach. And in that chapter I introduce a thing that I've called the PVCT model, which is not a very catchy name, I know, I apologize. Um, but essentially, my my suggestion is that we should probably start to look at families of tasks that start to add more and more context to how we engage. So, the first category of task I've referred to as precursor tasks. So, is there a thing that can occur in this engagement that will lead to success down river? So, let's say for example, we're playing a mount game. One partner is on another opponent and they're in mount. Our first task goal might be something as simple as, can you get your opponent's elbows past their shoulders or a bicep to touch an ear? Classic task, I'm sure everybody has played something similar at some point. We're looking to create isolation and an open elbow position where people are less likely to be able to defend arm entanglements and we can start to attack arm triangles. It's a really, really useful position for us to be able to explore and it can lead to a lot of success downstream. It's not necessary. I could still finish you without having to do that, but it does lead to me potentially having more success downstream. Once we've done that, we can start to look at the V side of things where we start to add in variability. So, it might be that my opponent for the first round is just looking to try and maintain their elbows being close to their body. They're not trying to escape yet. They're not trying to bridge me off. They're not trying to return to guard. They're just trying to keep their elbows down. Well, the problem we have is the way that I solve that problem might only work against an opponent that isn't trying to escape. I can beat the living crap out of people that are just trying to keep their elbows on their bodies. But the second they try and regard, all of the coordinated stretches that I'd put in place to be competent at that task might break down. So, we need to start adding in more and more variability. It might then be the case that I want the bottom player, let's say, we're playing a similar sort of task, but I've noticed we're playing nogi and I've not explicitly said the top player can threaten their neck as a method of creating action in their opponent's elbows. So, they're just forcing the arms down and it's been really defensible. So, I might introduce some variability where the top player can now threaten strangles, which creates these dilemmas where the elbow line might start to raise while they try and protect their neck. I might introduce that the person who is on the bottom can now look to try and regard or trap and roll. They can start to try and escape the position. So, I can now take the skills that we may have developed in that first game and add more representative and more representative environment for them to see. Does this skill actually transfer? We need to do this really consistently, right? We need to add more variability in the behaviors of our opponent and then we need to add more variability in the behavior of the person that is our primary focus of attention. So, while I'm looking to try and get your elbows up, am I in a position that I could transition into some other type of attack sequence? Could I catch you in, could I just move down and look for a leg entanglement? Could I step off you and look to try and take your back? Could I do other things that then start to contextualize the skills that I'm looking at? We have a concept called a nested affordance where me doing a thing can create all of these branching affordances, all these other opportunities for action that if we don't add more and more representative behaviors to the people that are exploring the task, we don't get to look at these branching affordances that only exist when the behaviors of each person becomes more closely aligned to the sport, right? This is the important thing. Speaker 1: What you're talking about here leads me to ask a question about something you mentioned earlier, which is open mats. Open mats are one of the most free-form places where people get to practice jiu-jitsu and really do what they want. And often these are pretty unstructured. Most open mats that I go to, people pick a partner, slap bump, roll. That's the extent of the planning that goes into it. And the conventional thought has always been, well, you get the most realistic feedback from actually training. And I guess that's true to some extent, but if your training is not conscious and you don't have a plan for what you want to learn, once you get more experience, what winds up happening is you kind of fall into pre-existing patterns and you wind up just playing the same game over and over again. So, I would think that for people who have a bit of experience under their belts, just going to an open mat might not be the best tool for skill development. What do you recommend for open mats? Do you think if people go to these that it makes sense to have some degree of planning or constraints that people put into place, some ground rules so they can focus on skill development? Or do you philosophically agree that people should just go and roll for the sake of rolling? I guess there's pros and cons to each approach, but I'm wondering from your elevated perspective as a an expert in the space, where do you line up? Speaker 2: I very much agree with you. We we have the phenomenon that everybody that drives will have experienced where you get behind the wheel and you drive to work and you don't know how you got there. You've just gone and done the same drive in the same way that you've done all the time. You aren't better at driving after that. Whereas if I was taking the exact same drive, I was doing the exact same thing, but I had said to myself, my intention for today's drive to work is to try and break as late into each corner as I can and to accelerate as quickly out from every corner. I might be a little bit more dangerous of a driver, but I can guarantee I would be a better driver at the end of that session, that drive to work. Intention in our training is obscenely important. Regardless of what type of training we're doing, we have to have some form of intention that is shaping where we are looking to try and improve. Exactly as you say, if I go to an open mat, the easy thing to do is to just fall back on the type of thing that I always do. You're going to see the slightly stiffer older guy doing deep half guard and doing pressure passing and they'll be as good at those things as they were before they entered. But if they set a challenge for themselves that they're going to try and explore how can I use the things that I'm already good at to add extra tools to my game? Can I use the threat of deep half guard to look to get into leg entanglements? I'm going to work on my leg entanglement game and link it to the things that are already a consolidated part of my game. I think that would be my big recommendation for more experienced players at open mats. It's not to just randomly pick a thing and explore it, but to see how they can add a skill set to what they are already competent at. You might have a really good wrestling game with really good top pressure, but you don't use that top pressure to engage into leg entanglements. Or maybe my challenge for the day is, can I threaten top pressure to the point that I can get really dominant exposure of people's heels? That would probably be a much, much better open mat than if I just turn up and hulk smash people with the same type of thing that I do every single week. Absolutely. Speaker 1: Well, let me flip the coin around then and we can talk about the other side of this. At the end of the day, I feel that the most important thing for someone's jiu-jitsu practice is that it has to be fun. It has to be something that they get enjoyment out of because ultimately that's how you build up the long-term motivation that keeps you consistent at this. If you hate every aspect of your training, it probably doesn't matter how good the quality of that training is because you're unlikely to stay committed for a long time. So, the flip side to the open mat discussion is sometimes having fun is its own reward. I don't think it's necessarily bad to just show up and have a good time even if that means you have no plan. But we also have to be cognizant of the fact that you're probably less likely to get a lot of skill development out of it if you just show up and just do whatever comes to mind at the moment. Do you think there's an argument for having, I guess, just fun days where you don't really think about any sort of skill development, but you just go for the sake of going? Or are there consistency benefits to always having a skill acquisition plan? Speaker 2: It's a great question. And again, the answer often will depend on the seriousness of the competitor. So, if you're just a recreational blue belt that turns up because you want to go and strangle your friends and pretending to play fight is a bit awkward when you're old, then have fun as often as you want. Like if you're not interested in getting as good as you can humanly be, is reach your potential for skill development, then absolutely. We we're doing a sport. I like playing football. I don't want to join a football team. I'll kick a ball around with my friends, but at no point am I going to go and join football training so I can get better at the sport. I just like kicking a ball around. I think in my opinion, I think there should definitely be space for people like that in grappling. They don't really want to get good, but they do like the sport. They want to come along and play about. Perfect. They're more than welcome to do that. If my intention is to become the next Gordon Ryan, but nicer, then probably spend as much time on skill development as we can. But as part of being an athlete at those levels, having a mental health day is really important. Maintaining motivation for a long period of time is really important. Coaching isn't just skill development. So, if as a coach or as an athlete, the decision is made that for today, our concern isn't developing skill in this area. I just want you to have fun. You've had a really stressful couple of weeks. You've had all of these external factors that are making your time on the mat less enjoyable. Just turn up and smash people. That is absolutely a decision people can make. Speaker 1: Nice. Well, we've talked about the student's perspective. What about from the coach's perspective? This discussion of repetition versus representation or quantity versus quality. As someone who helps work with a lot of coaches on optimizing their curriculum for these kinds of concerns around motor learning, what do you find most coaches do, I guess, suboptimally in this regard and how do you find you can help people improve the quality of the training in their room through this concept? Speaker 2: A great question. Yeah, so I think what I see really, really frequently are people are using, people have got really good at designing practice tasks. They're starting to understand the concept of perception action coupling. They're able to identify where specifying information is. They're designing tasks where people are behaving in such a way that people can start to move in response to it and solve the problem. It's not a case of learning moves. They're solving problems against an opponent that is behaving in a way that lets them engage at a unit of analysis where there has to be a coupling of what they are doing to what their opponent's doing, which is kind of the minimum threshold. They've reached the point at which perception action coupling is occurring. This is amazing. It's so much better than things were 20 years ago and sort of nerding up on all this stuff. You'd see a whole bunch of regressive, here's a move, go and do the move 100 times and then maybe if you've eaten all of your carrots, you can have pudding at the end and you can have five minutes where you get to fight each other. So, we're definitely moving in the right way, but what often happens is people have these unconnected games where there's no through pull and progression on a skill set. They'll do, we're going to play a single leg game where we're going to try and do a thing. Then we're going to play a game where we're trying to get past our partner's legs. And then we're going to play a game where we start in some sort of a pin. Each of those tasks will be quite well constructed. They'll be an opponent that is behaving in a representative way. They'll be perception action coupling occurring. They'll be action fidelity, they'll be functionality, they'll be lots of things that you want to see in a task. But the types of tasks they're doing are all disconnected from each other and they don't start to gradually introduce a more and more representative environment to battle test the skills they're developing. So, my suggestion, and thing that I really try and work with the guys that I work with on, is to try and get people to understand that the more of the game we add in, the more the skill sets that we're starting to test and develop become contextualized and become a part of our game. The higher the rate of retention and transfer is. So, let's say for example, we're playing, so I'll use a foot sweep game because I've used this in my little book chapter as a as a way of highlighting this. It might be the case that I want people to develop the skill of executing just a foot sweep on their opponent. The first introduction to that might be something where we have to try and educate our player's attention to what the variable, the variable is that they need to pick up on to be able to do a foot sweep, right? My opponent's center of mass in relation to their base of support is very often that variable. If I can get my opponent's belly button past their heels and their weight is backwards, it'll mean that moving their foot to make them fall over becomes vastly easier. So, I might play a game like a tug of war where each person starts gripped up and they're just trying to pull their opponent to their side of the mat. By pulling our opponents, we end up in these really exaggerated compromised sort of V-shapes where your opponent's hips are miles behind them and it's really easy for you to start to explore how we can use our feet to displace their feet and get them to fall over. The problem is, I don't fight people that end up in these really exaggerated positions very often and it's very, very rare that we only go forward and back. So, we need to start adding an extra layer of representation that gets them to test the same skill under a more representative condition. So, I might play a game that I call landmines. Again, I have a sick sense of humor apparently, where I throw down a bunch of rubber spots on the floor and now if I can make my partner step on a spot, I score a point. That is a win for me. By doing that, my opponent's going to end up stepping long or putting their foot into the floor and trying to resist a pull. They're going to end up stepping sideways. There's going to be a multi-directional movement that I have to then start learning, how do I foot sweep when they're moving like this? And the amount of time that my opponent's balance is compromised is vastly shorter. So, now I'm still developing the same skill. My attention was educated to what kind of thing to look for in the first task, but now my intention is the thing shaping it. It's, I'm looking to try and do this under all of these variable conditions. The problem we have is my opponent is still not behaving like a person that actually behaves in a grappling sport. They're only doing the same thing back to me. We haven't said that they can also throw me. They could just single leg me, they can hip throw me, they can do whatever. All of the things that exist should then be start to add added back in bit by bit by bit. So, if I then say we're going to play the same task, we're going to play this landmines game again, but now you can do any throw, any takedown. If you get your partner to the floor, you score a point. If you make them step on a spot on the floor, you score a point. But if at any point you can use your foot to displace their foot and foot sweep them, I will buy you a bottle of whiskey. The people in that class will then start to explore the skill set of using their feet to displace their opponent's base of support in the context of all of the other things that are available to them. Can they do it when their opponent is looking to engage onto them and attack? Can they see how the threat of that foot sweep links into hip throws? How it links into guard pulls? Whispering. How all the skills from the sport start to connect to the skill that we are exploring as the main focus of our attention for this session. That, I think, is incredibly important. The other thing that is really, really under under explored are the transitions between siloed phases of play that exist. So, for example, after we have been successful at foot sweeping our opponent, we need to engage onto them. So, I don't want to play a guard passing game and a takedown game. I want to play the thing that I'm looking at, one phase before it and one phase after it. So, after I have been successful or unsuccessful at sweeping you, you will land on the floor and how you land will determine how I pass your guard. I very rarely have fought somebody and they sit down and put their hand out and go, slap hand, fist bump, are you ready to pass my guard, let's go. The guard passing that I do is a consequence of how we end up in that position. If I foot sweep you, the way I look to get past your guard and do things to you is wildly different to if I've just done a shoulder throw and I've thrown you head over heels. The way I engage into that position is very, very different. So, we need to be able to explore the nested affordances that come as a consequence of exploring the skill set that is our primary focus for that session. And the only way of doing that is to add more and more and more representative behaviors for the people that are engaged in that task. It's not just, here's a takedown game, here's a guard passing game, here's a pinning game. It's, we need this extra add-on of more and more and more representative consequences and behaviors for the people engaged in that task. So, I can see, does my skill, does the thing I'm developing work when that person is also allowed to do X? Does my ability to escape somebody having chest to back and hooks in change when they're also allowed to strangle me? Probably does. I probably need to add that layer of representation back in to make sure any of the skill sets that we're exploring are congruent and fit the context of the actual sport, not just the slice of the task that we're taking that has specifying information. There is still enough information for us to start to develop some functional coupling, but it doesn't have as much as adding in more and adding in more and adding in more. We need to start stretching and tempering and testing the skills we have in more representative environments as we progress through a practice session. Speaker 1: Thanks, Cal. That's great. Well, one thing I want to ask you, you've mentioned this term a few times and it has been sometimes since we've brought this up on the podcast, but it's a core concept to this whole area of sports science. Perception action coupling. The challenge is, this is a pretty hard concept for many people to get their heads around. So, I always want to see if I can find someone who can explain it better than me. Could you give sort of the dummy's guide on what perception action coupling is so that everyone understands the idea? Speaker 2: Yeah, I I can try. So, there's two main schools of thought on how motor learning occurs. There are two main camps. One of them is more of a cognitive version of how it works where information is what we call impoverished. It's just binary code, zeros and ones that we take in with our sense organs and our brain has to add value to it. It interprets that information that is out there. It does higher order mathematics, works out the properties of the things in front of us. It'll judge how fast they're traveling, what direction they're going, what the materials they're made from are, all of the things that those things that we experience are, has to be interpreted by our brain. That is what they call indirect perception. In indirect perception, we don't have to worry about perception action coupling as a concept. What we're doing is we're having these large generative things where we have representations of moves and representations of properties that will have a prediction that is made and we press play on some form of a movement or some kind of a consequence to that and magically things occur. The explanation for how that happens is kind of non-existent. It's one of the things that's really fun in the literature is they just say, and then things happen. The flip side of that is where we have an understanding or an assumption that perception is direct, that the information that is out there is itself specifying. That when I look at an object, I don't need my brain to have to interpret ones and zeros. I move directly in relation to that information. So, if for example, somebody is throwing an object at my head, I don't have to have my brain work out the trajectory, velocity, angles, all of the things that that has. It doesn't have to predict where it will be at a certain time and press play on me ducking or moving my hand to catch it. What would happen is there is a variable that we can move in relation to. For that, for anybody that has read Rob's stuff, there is a thing called tau, which is the rate of change of the size of an object. So, all I need to do is move my hand at a rate at which it will intercept this thing as it gets closer to me. That variable I am picking up directly with my sense organs and I'm coordinating my action directly based on that information. There is this perception that is occurring and an action that is being coupled to it so that we can end up having this really close link between the things that we are experiencing with our sense organs and the movements that we are making. We also have this flow back in that the actions we take lead to perception as well. So, as I am moving, there are things that I start to perceive because I am moving. So, we have this reciprocity between movement and perception and movement and movement and perception, if you see what I mean. Hopefully that's clear. Speaker 1: Well, you know what? I will leave it to the listeners to tell me. Feel free to write in and ask if you need further clarification because I always find this to be an area that people get hung up on and the more diverse perspectives we can provide on how to explain it, the better. But Cal, as we get close to the hour here, I want to give you a chance to talk about your work. If people want to follow you or get remote advice from you because one of the things that you do is provide coaching to other coaches who want to implement CLA into their gyms. How can people find you and follow you and what kind of services do you provide if people wanted to work with you to some degree? Speaker 2: Yeah, so my Instagram, I think I'm Cal W Jones. If you're interested, if you have any questions, feel free to add me, follow me. I don't know what they say for Instagram. I'm a bit of a luddite with technology. But feel free. I'm always overjoyed and delighted to engage with people on these subjects. I'm a sad little human being that actually enjoys discussing this stuff. If you like to, you're more than welcome to. As for my consultancy stuff, I have a bit of a group of people that I work with. I think we're up to 49 now where I essentially offer a consultancy service. So, how people make use of that can be wildly different. We have some people that have asked for help just interpreting the literature and seeing how that has any implications for coaching BJJ because it's not intuitive and the way that it's written is often impenetrable. If you've ever read a Kelso paper, you know what I'm talking about. For some people, I will assist with how they can design practice tasks better, how they can ensure that there is information that they can start to get this perception action coupling. For some, it's how they can design sessions so that there isn't this disjointed, here's a game, here's a game, here's a game, here's a game, but there is a through pull to make sure that skill is being developed across a session. For some people, it's developing their school's curriculum. For some people, it is upskilling the assistant coaches and co-coaches that work with them. I've described myself as a really boring genie in a lamp that can only really do coaching stuff. But yeah, if there's any kind of engagement in a coaching sphere that you might need some assistance or an impartial other person to run your ideas by, I'm I exist and I offer that as a service. As part of that, we have a sort of community of practice that I've put together from the guys that work with me that I've called the grappling with CLA community. So, we have monthly meetups where I sometimes get a guest lecturer, somebody that's written papers in the space or somebody who works using these sorts of things or we'll just have a general topic of discussion that we'll all meet up and discuss together. We have a WhatsApp group for short fire questions and we have a subreddit where my intention is in five years time, somebody might think, oh, I wonder what meta stability is and they can just type it into the Reddit and somebody will have had that discussion and they can go back and check it. Whereas in like Discords or WhatsApps or short form things, if you're not there at the moment that something is being discussed, you kind of miss the conversation. So, yeah, that's my hope. I know people will ask or if they are interested, will be concerned about the pricing. I run on what I call an anarchist with a mortgage model where essentially I recommend people pay roughly what they charge somebody for a month at their gym. So, if you coach like people in a in an impoverished neighborhood, you would end up paying less than somebody who exclusively coaches a Silicon Valley billionaire, who again, I'm hoping I will meet one day. Speaker 1: Amazing. Well, again, I will put links to your stuff in the show notes and if you are a Silicon Valley billionaire and you want to dramatically overpay Cal for his services, that's how you can find him. If you would like to dramatically pay the standard fees for my services, you can find everything that we make at BJJ Mental Models.com. It is of course the the podcast that we're listening to here plus mini episodes which are just more quick hit concepts that get you value in ideally under 10 minutes. We release both of those once a week and of course there's our newsletter as well. All of that is completely free. If people want to level up with us, I would suggest that they check out BJJ Mental Models Premium. It is the world's largest library of structured jiu-jitsu audio content. Think of Audible or Masterclass but focusing around jiu-jitsu. If that's your preferred vehicle for learning and you want to get deep into jiu-jitsu strategy, tactics, mindset, concepts, philosophy even, that's how you do it. The first week is free so you can get all of that at BJJ Mental Models.com. I will put a link to my stuff and to your stuff, Cal, in the show notes. But thank you so much for doing this. This was a long time coming and I'm really glad we were able to connect. Speaker 2: Oh, absolutely. It's been joyous. It's been really fun. Speaker 1: Amazing. Thank you to you and thank you to the listeners as well. We truly appreciate you and we will talk to you next time. See you then.