WEBVTT - Using Computer Vision to See What Coaches Can’t 0:00:08.320 --> 0:00:08.800 Pushkin. 0:00:16.160 --> 0:00:19.079 Because I didn't know that sports biomechanics could be a career. 0:00:20.280 --> 0:00:22.880 I decided to go to grad school and initially start 0:00:22.960 --> 0:00:25.960 working on prosthetic limbs. So that was the first two 0:00:26.000 --> 0:00:30.920 years of grad school. Then about two years in I 0:00:31.040 --> 0:00:36.640 discovered baseball pitching biomechanics research. The funny thing that happened 0:00:36.640 --> 0:00:40.360 there was I gave a PhD committee meeting where I 0:00:40.400 --> 0:00:44.440 spent most of the time talking about prosthetic limbs, and 0:00:44.479 --> 0:00:46.920 then I spent the last few minutes as an aside 0:00:47.360 --> 0:00:50.520 on the baseball pitching research I found. And my committee 0:00:50.600 --> 0:00:53.760 was like, Jimmy, the last few minutes were way better 0:00:53.800 --> 0:00:55.000 than the first forty fives. 0:00:55.400 --> 0:00:57.040 So I was like, all right. 0:00:57.240 --> 0:01:00.800 So then they were like, we'll let you do baseball 0:01:00.840 --> 0:01:04.080 pitching biomechanics as your PhD work, but just so you know, 0:01:04.319 --> 0:01:06.480 it might be really hard to have a career doing that. 0:01:08.040 --> 0:01:09.480 But you know, they were like, if you want to 0:01:09.520 --> 0:01:10.760 go for it, you can go for it. And so 0:01:10.840 --> 0:01:13.319 I was like, you know what, Sure, I'll go for it. 0:01:19.200 --> 0:01:22.200 I'm Jacob Goldstein, and this is what's your problem. Today 0:01:22.280 --> 0:01:25.639 we have the second episode in our series about people 0:01:25.880 --> 0:01:29.039 who are working at the frontiers of technology to help 0:01:29.080 --> 0:01:33.640 a lead athletes perform better. My guest today is Jimmy Buffy, 0:01:33.959 --> 0:01:36.639 and as it happened, the concerns of his grad school 0:01:36.680 --> 0:01:40.800 advisors were unfounded. Jimmy has in fact made a career 0:01:40.920 --> 0:01:44.920 out of the biomechanics of pitching in baseball and sports 0:01:44.920 --> 0:01:48.560 biomechanics more broadly. When he finished grad school, he got 0:01:48.560 --> 0:01:51.120 a job with the Los Angeles Dodgers, and he went 0:01:51.160 --> 0:01:54.160 on to co found a company called Reboot Motion that 0:01:54.280 --> 0:01:57.360 works with teams in Major League Baseball and the NBA. 0:01:57.840 --> 0:02:01.360 Jimmy's problem is this, how do you take massive amounts 0:02:01.360 --> 0:02:05.400 of data about how professional athletes move and turn all 0:02:05.440 --> 0:02:10.600 that data into information that actually helps those athletes perform better. 0:02:12.240 --> 0:02:15.160 You end up doing your dissertation research on the on 0:02:15.240 --> 0:02:19.480 the biomechanics of pitching, Yes, of baseball, pitching and baseball. Yeah, 0:02:19.560 --> 0:02:22.639 and and then you get hired by the Dodgers. 0:02:23.080 --> 0:02:23.800 Yeah. 0:02:23.960 --> 0:02:27.520 Yeah, that was That was awesome because I originally didn't again, 0:02:27.560 --> 0:02:30.320 didn't realize that that could be a thing that could happen. 0:02:30.800 --> 0:02:33.840 Well, and it kind of wasn't right, Like, you're kind 0:02:33.880 --> 0:02:36.640 of just coming into this field as it's becoming a 0:02:36.680 --> 0:02:38.799 field where you can get a job where it's a. 0:02:38.720 --> 0:02:40.280 Field, right exactly. Yeah, there. 0:02:40.360 --> 0:02:42.280 I mean the challenge then was there wasn't a lot 0:02:42.320 --> 0:02:45.400 of options for actually even getting the data that you 0:02:45.560 --> 0:02:47.079 need to analyze. 0:02:47.800 --> 0:02:51.120 So ten years ago, like what is the state of 0:02:51.160 --> 0:02:54.160 play in this sort of nasson field that you're in 0:02:54.400 --> 0:02:55.200 helping to create? 0:02:56.639 --> 0:03:00.000 So the field is, I would say, is like sports biomechanic, 0:03:00.200 --> 0:03:03.600 and what that is is being able to analyze the 0:03:03.760 --> 0:03:07.519 movement of athletes for lots of purposes, help them reduce 0:03:07.560 --> 0:03:10.720 injury risk, help them improve performance. 0:03:10.680 --> 0:03:14.440 And to be clear, like folk, sports biomechanics has been 0:03:14.480 --> 0:03:16.720 around forever, right, that's what coaches do. They stay out 0:03:16.720 --> 0:03:18.480 there and they watch it. Yeah, And so it's kind 0:03:18.480 --> 0:03:21.519 of becoming it's becoming more technical, right, the field is 0:03:21.560 --> 0:03:22.480 becoming more technical. 0:03:22.600 --> 0:03:22.919 Yeah. 0:03:22.960 --> 0:03:26.799 So the state of the art relied on what is 0:03:26.840 --> 0:03:31.760 called marker based motion capture, which is where you literally 0:03:31.800 --> 0:03:36.080 put reflective markers like little balls, you stick them all 0:03:36.120 --> 0:03:39.440 over somebody's body. Usually the person has to like strip 0:03:39.480 --> 0:03:41.760 their clothes off because you want the markers like literally 0:03:41.840 --> 0:03:45.040 like on the skin, on the joints, and then you 0:03:45.120 --> 0:03:48.400 have these special cameras that track those markers. 0:03:48.760 --> 0:03:50.920 So you've got like a picture in his underwear with 0:03:50.960 --> 0:03:54.000 a bunch of little metal balls, takes them and they're like, 0:03:54.160 --> 0:03:55.800 just pitch like you always pitch. 0:03:55.720 --> 0:03:57.000 Right, And that's the challenge. 0:03:57.160 --> 0:03:59.760 That's why that wasn't That's why it wasn't very wide 0:03:59.800 --> 0:04:02.040 SPA is a thing people did because it was so 0:04:02.240 --> 0:04:04.800 hard to collect the data because ultimately, what you would 0:04:04.840 --> 0:04:08.000 need is you need the data on how someone is moving. 0:04:08.360 --> 0:04:09.280 You need to track. 0:04:09.040 --> 0:04:11.000 Where their elbow is, where their wrist is, where their 0:04:11.080 --> 0:04:13.800 knees are so that you can analyze it. And the 0:04:13.840 --> 0:04:15.960 state of the art for tracking it was an awful 0:04:16.000 --> 0:04:18.240 experience for the people you were trying to track. 0:04:18.720 --> 0:04:22.479 That presumably would mean they didn't pitch like they usually exactly, 0:04:22.480 --> 0:04:25.040 because they don't usually stand there in their underwear with 0:04:25.440 --> 0:04:27.200 met Were they really in their underwear by the way, 0:04:27.240 --> 0:04:28.840 I'm saying it because it sounds absurd, but is that 0:04:28.839 --> 0:04:29.680 actually what they were doing. 0:04:29.760 --> 0:04:32.680 That's actually you strip down to your your boxers, your 0:04:32.680 --> 0:04:34.800 boxer briefs, and that's it, that's all you're wearing. 0:04:35.640 --> 0:04:38.080 So it basically didn't work, and it basically wasn't very 0:04:38.080 --> 0:04:39.560 widely used as a result. 0:04:39.360 --> 0:04:43.320 Right exactly, Yeah, you yeah, you look at studies in 0:04:43.360 --> 0:04:47.279 that field, and people would be throwing like several miles 0:04:47.279 --> 0:04:49.600 an hour slower than they would be throwing when they 0:04:49.640 --> 0:04:55.279 weren't wearing all that stuff. So what changes computer vision? 0:04:56.440 --> 0:05:00.360 That was the big That was the big inflection point. Now, 0:05:00.360 --> 0:05:04.480 to be fair, like, even when I was finishing my PhD, 0:05:04.960 --> 0:05:06.440 and I'll give them, I'll give them a shout out, 0:05:06.480 --> 0:05:10.240 there was a company that was already like working really 0:05:10.279 --> 0:05:13.240 hard to solve this problem for baseball teams that I 0:05:13.480 --> 0:05:16.440 was getting that I got to be familiar with, called Kinnetrax. 0:05:18.279 --> 0:05:21.840 But yeah, the big inflection point was computer vision, basically 0:05:21.960 --> 0:05:26.960 using artificial intelligence to identify where those joints are in 0:05:26.960 --> 0:05:31.760 a camera image rather than needing to paste those reflective markers. 0:05:32.279 --> 0:05:35.320 So computer vision takes off. You're working at the Dodgers, 0:05:35.560 --> 0:05:39.080 and then eventually in twenty nineteen, right, you leave the 0:05:39.120 --> 0:05:43.840 Dodgers and you start your company a reboot motion, what 0:05:43.880 --> 0:05:44.640 does your company do? 0:05:45.240 --> 0:05:49.360 We do what we call biomechanics as a service, So 0:05:49.680 --> 0:05:52.840 we try to analyze this computer vision data at a 0:05:52.960 --> 0:05:56.800 very large scale to help teams and coaches make use 0:05:56.800 --> 0:06:03.200 of it to help athletes get better. Bas Yeah, that says. 0:06:03.040 --> 0:06:06.720 But bess Yah and and who are your customers? 0:06:06.920 --> 0:06:14.239 Our customers are Major League Baseball teams actually NBA teams, 0:06:14.279 --> 0:06:19.520 so we've gotten into basketball also sort of like league 0:06:19.560 --> 0:06:23.839 wide data providers. So yeah, leagues teams is basically our 0:06:23.880 --> 0:06:24.440 sweet spot. 0:06:24.720 --> 0:06:27.760 So let's talk. Let's talk in like a little more 0:06:27.760 --> 0:06:30.839 detail about about what you actually do, right, tell me 0:06:30.880 --> 0:06:32.360 the story of what you do. 0:06:33.600 --> 0:06:37.520 So, Evan, my co founder, Evan Demchik, he likes to 0:06:37.520 --> 0:06:41.920 call this the biomechanics Trainkay. 0:06:41.080 --> 0:06:43.080 Let's take a ride on the biomechanics trend. 0:06:43.720 --> 0:06:45.800 And we call it that because with the way our 0:06:45.839 --> 0:06:49.320 product works is we let people get on the train 0:06:49.760 --> 0:06:51.800 at whatever stop works for them, and get off the 0:06:51.839 --> 0:06:53.640 train at whatever stop works for them. 0:06:53.720 --> 0:06:55.360 How far are we going to go with this metaphor 0:06:55.600 --> 0:06:57.360 of a little o'b nervous about it? 0:06:57.760 --> 0:06:59.160 That might be as far as we go. 0:06:59.320 --> 0:07:03.960 Okay, good, good, So just tell it to me. Start 0:07:04.000 --> 0:07:06.279 at whatever seems like the beginning of a you know, 0:07:07.040 --> 0:07:10.120 a of an encounter, and so I'd like to understand 0:07:10.160 --> 0:07:12.600 how that works. That's really kind of a way to 0:07:12.600 --> 0:07:15.640 think about it. So let's start with the data, right, 0:07:16.040 --> 0:07:19.080 what is a basic what is the basic thing that's. 0:07:18.920 --> 0:07:22.440 Happening So the very first thing that happens is you 0:07:22.520 --> 0:07:27.920 record videos of the athlete doing the athletic motion. So 0:07:27.920 --> 0:07:30.880 we'll talk about pitching, So you record videos of a 0:07:30.920 --> 0:07:34.840 pitcher pitching. Our product, we actually have implemented our own 0:07:34.840 --> 0:07:38.560 computer vision models, so we can do that if people want. 0:07:38.640 --> 0:07:43.480 But generally speaking, people have systems like Kinetrax is one 0:07:43.520 --> 0:07:47.120 I've mentioned that, Hawkeye is another popular one where there 0:07:47.200 --> 0:07:49.600 is a system in place that has the cameras that 0:07:49.680 --> 0:07:52.960 record the videos and then runs those computer vision models. 0:07:53.240 --> 0:07:56.840 What those computer vision models do is they extract the 0:07:56.920 --> 0:08:00.880 locations in three dimensional space of all of like the 0:08:00.960 --> 0:08:03.640 joint centers. So where's my elbow, where's my wrist, where's 0:08:03.640 --> 0:08:07.040 my knee? In three dimensional space? That's what comes out 0:08:07.040 --> 0:08:08.400 of these computer vision systems. 0:08:08.920 --> 0:08:12.080 Okay, so pretty much everybody has that at this point, 0:08:12.120 --> 0:08:16.280 Like every professional baseball team has that for every pitch 0:08:16.320 --> 0:08:17.480 in every game at this point. 0:08:17.600 --> 0:08:18.320 Yes, exactly. 0:08:18.640 --> 0:08:21.520 Okay, So it's a ton a ton of data. So 0:08:21.560 --> 0:08:24.360 that's another challenge that we've solved, is not only how 0:08:24.360 --> 0:08:25.640 do you do this, but how do you do this 0:08:25.680 --> 0:08:27.040 at a very large scale? 0:08:27.280 --> 0:08:30.920 Okay, so this data, everybody's got it now, and you're not. 0:08:31.360 --> 0:08:33.760 You can sort of process the data, but that's not 0:08:33.880 --> 0:08:36.559 your special sauce. That's not your secret sauce, right, So 0:08:36.600 --> 0:08:39.800 typically they'll send you that that data of like, here's 0:08:39.800 --> 0:08:41.360 all the body points. Here is how they're moving in 0:08:41.520 --> 0:08:44.000 physical space. Then what do you do with it? 0:08:44.360 --> 0:08:46.600 So, yeah, this is where the special sauce comes in. 0:08:47.120 --> 0:08:49.320 So the first step to that is how do you 0:08:49.400 --> 0:08:53.160 turn those key points into a human skeleton. So you've 0:08:53.160 --> 0:08:55.439 got to figure out, like where do the bones connect, 0:08:55.520 --> 0:08:58.200 what sort of degrees of freedom do those bones have, 0:08:58.760 --> 0:09:01.840 So then you can figure out how do those key 0:09:01.880 --> 0:09:04.640 points animate a human skeleton. 0:09:05.120 --> 0:09:07.160 So you're sort of rebuilding it. It's like you start 0:09:07.200 --> 0:09:08.880 with the picture of a person and then you turn 0:09:08.880 --> 0:09:10.839 it into a bunch of data points, and now you've 0:09:10.840 --> 0:09:13.160 got to kind of build the person back up again from. 0:09:13.000 --> 0:09:17.120 The day exactly exactly. So now you have an actual 0:09:17.280 --> 0:09:21.240 human skeleton where the shoulder is rotating, the elbow is flexing, 0:09:21.280 --> 0:09:24.240 the knee is flexing, the hips are rotating. And now 0:09:24.280 --> 0:09:27.120 once you do that, now you can understand that data 0:09:27.160 --> 0:09:30.160 in the context of how the body works. Okay, so 0:09:30.240 --> 0:09:34.000 once we've done that, then we calculate how energy flows 0:09:34.040 --> 0:09:37.240 through the body, we calculate how momentum flows through the body, 0:09:37.520 --> 0:09:40.359 and once we've done that, we can analyze how efficiently 0:09:40.360 --> 0:09:43.840 the athlete is moving. Are they generating energy and momentum 0:09:44.080 --> 0:09:47.000 in the direction that they want to generate in? What 0:09:47.160 --> 0:09:50.199 is that desired direction? So we then we calculate all 0:09:50.240 --> 0:09:54.640 sorts of metrics around movement efficiency and direction. Then once 0:09:54.679 --> 0:09:58.319 we calculate all those metrics, now we can understand how 0:09:58.320 --> 0:10:00.720 those relate to what you're trying to do. Throw the 0:10:00.720 --> 0:10:01.960 ball as hard as possible. 0:10:02.040 --> 0:10:05.200 So so presumably with the picture, what you want to 0:10:05.240 --> 0:10:08.640 optimize for is is having as much of the picture's 0:10:09.520 --> 0:10:13.199 energy of their body go toward making the ball go 0:10:13.280 --> 0:10:15.760 toward home plate. Exactly right, I mean that is that? Yeah, 0:10:15.800 --> 0:10:17.800 they mail optimization problem mail it. 0:10:17.840 --> 0:10:20.120 Yeah, Yeah, that's exactly it. And that's the problem that 0:10:20.160 --> 0:10:22.840 we try to understand. So when we build our sort 0:10:22.840 --> 0:10:25.679 of models regarding like how does a pitcher create efficient 0:10:25.679 --> 0:10:28.640 fastball velocity, one of the most important things that comes 0:10:28.679 --> 0:10:31.600 out of those models is lining up the direction of 0:10:31.600 --> 0:10:33.960 your torsore rotation with the direction of your arm rotation. 0:10:34.400 --> 0:10:37.439 The pitching bution is crazy complex, right, Like they're they 0:10:37.520 --> 0:10:39.640 kick their leg up and they got their front arms 0:10:39.679 --> 0:10:42.200 doing something and their back arms going back like a 0:10:42.280 --> 0:10:46.040 lot is happening. Yeah, the sort of platonic ideal is 0:10:46.120 --> 0:10:50.160 every little millimeter of every motion is going toward maximizing 0:10:50.160 --> 0:10:52.000 the energy of the ball going toward home plate. 0:10:52.160 --> 0:10:56.000 Yes, and not just not just maximizing like in a vacuum, 0:10:56.160 --> 0:10:59.640 but doing it in the most efficient way, because if 0:10:59.640 --> 0:11:02.560 you just sort of maximize in a vacuum, maybe you're 0:11:02.559 --> 0:11:04.680 transferring that energy in a way that hurts your elbow. 0:11:04.840 --> 0:11:07.040 You're transferring that energy in a way that hurts your shoulder. 0:11:07.520 --> 0:11:09.280 So not only do we figure out how can a 0:11:09.280 --> 0:11:11.880 pitcher maximize it, but we try to figure out how 0:11:11.920 --> 0:11:14.360 they can have that energy and momentum go in a 0:11:14.440 --> 0:11:18.800 direction that doesn't hurt their joints, So try to have 0:11:18.920 --> 0:11:21.960 them throw the ball a little harder while also reducing 0:11:21.960 --> 0:11:23.160 their injury risk a little bit. 0:11:23.559 --> 0:11:30.640 So you're whatever doing the math at reboot HQ, and 0:11:30.679 --> 0:11:33.319 then what are you sending back to the. 0:11:33.320 --> 0:11:38.199 Team, So some teams so it sort of so again, 0:11:38.440 --> 0:11:39.760 all right, I told you that was the end of 0:11:39.760 --> 0:11:41.160 the biomechanics train analogy. 0:11:41.720 --> 0:11:43.320 I'm going to bring it back for a brief second. 0:11:43.840 --> 0:11:44.679 Okay, I'm ready. 0:11:45.800 --> 0:11:48.160 So we go all the way to building a report 0:11:48.240 --> 0:11:51.200 that has a bunch of suggestions. So there's a report 0:11:51.240 --> 0:11:53.719 that's like, this is how efficient you are. You can 0:11:53.840 --> 0:11:55.599 like tilt your torso a little bit to be a 0:11:55.640 --> 0:11:57.439 little bit more efficient. You can tilt your arm a 0:11:57.440 --> 0:11:59.280 little bit to be a little bit more efficient. So 0:11:59.320 --> 0:12:02.440 we go all the way to generating a report that's 0:12:02.480 --> 0:12:03.840 like the final stop on the train. 0:12:04.480 --> 0:12:06.400 And that's a report for for one pitcher. 0:12:06.679 --> 0:12:09.120 Yeah, that's a report on a pitcher for whatever. 0:12:09.480 --> 0:12:13.200 And and in a way that's like the nerdiest it's 0:12:13.280 --> 0:12:15.360 what a pitching coach does, but just in a way 0:12:15.440 --> 0:12:16.120 nerdier way. 0:12:16.360 --> 0:12:16.600 Yeah. 0:12:16.640 --> 0:12:20.120 Well, so we try to sort of like give the 0:12:20.160 --> 0:12:23.320 coaches superpowers, you know, even though that you know, the 0:12:23.360 --> 0:12:25.800 coaches like can look at an athlete and understand a 0:12:25.840 --> 0:12:29.120 lot about the athlete just by looking at them. We 0:12:29.160 --> 0:12:32.920 try to make a report that can really amplify what 0:12:33.000 --> 0:12:36.160 the coach is already doing, maybe help them discover some 0:12:36.240 --> 0:12:39.560 things they weren't thinking about, or measure some things that 0:12:39.600 --> 0:12:41.600 they were thinking about, but now they can track those 0:12:41.640 --> 0:12:44.040 a little bit easier. So that's you know, that's the 0:12:44.160 --> 0:12:47.880 ultimate thing that we produces a report that can sort 0:12:47.880 --> 0:12:49.719 of like we say, give a coach superpowers. 0:12:51.760 --> 0:12:53.960 Where was the trained metaphor doing that? Is there an 0:12:53.960 --> 0:12:57.000 earlier station of disembarkation exactly? 0:12:57.760 --> 0:13:01.600 So a lot of teams now are hiring people with 0:13:01.679 --> 0:13:06.080 biomechanics and analytics backgrounds, so rather than just use our 0:13:06.200 --> 0:13:09.000 reports out of the box, they want to build their 0:13:09.040 --> 0:13:11.440 own reports and their own statistical models and their own 0:13:11.480 --> 0:13:14.559 AI models. So we also get let people get off 0:13:14.600 --> 0:13:17.040 the train a little bit earlier and build whatever they 0:13:17.120 --> 0:13:19.120 want on top of the data that we're generating. 0:13:19.360 --> 0:13:22.880 Are they not going to disintermediate you once those people 0:13:22.880 --> 0:13:25.400 are there? Does that reduce the value you provide to 0:13:25.440 --> 0:13:26.439 the team. 0:13:26.120 --> 0:13:28.880 Now because we still have to process all that data. 0:13:29.520 --> 0:13:32.720 Do you have some like IP or like why can't 0:13:32.920 --> 0:13:35.400 somebody like you who works for a team just do 0:13:35.480 --> 0:13:36.040 that without you? 0:13:36.160 --> 0:13:39.720 That is a great question. We answered this question all 0:13:39.760 --> 0:13:44.920 the time. Is because it's a very complex engineering problem. 0:13:44.960 --> 0:13:47.440 Not only to do all the math that the physics 0:13:47.440 --> 0:13:50.520 based math, to calculate the energy, calculate the momentum, like 0:13:50.600 --> 0:13:54.000 all of that math is really hard, but also to 0:13:54.080 --> 0:13:57.959 do it at a very large scale. So someone like 0:13:58.040 --> 0:14:00.760 me in grad school learned how to do that on 0:14:00.800 --> 0:14:03.880 a sample size. You know, my PhD was actually really 0:14:03.920 --> 0:14:06.439 just one pitch, but lots of people do it on 0:14:06.520 --> 0:14:10.040 like ten pitches or maybe one hundred pitches, So we 0:14:10.200 --> 0:14:16.760 do it on several thousand pitches like and swings like 0:14:16.880 --> 0:14:17.520 every morning. 0:14:18.240 --> 0:14:18.480 You know. 0:14:18.559 --> 0:14:23.600 There's just every team has like seven affiliates, so there's 0:14:23.920 --> 0:14:26.120 you know, one hundred and fifty games every day that 0:14:26.160 --> 0:14:26.560 need to be. 0:14:26.680 --> 0:14:28.640 So you're doing this on farm teams as well. 0:14:28.720 --> 0:14:29.600 Yeah, exactly. 0:14:29.960 --> 0:14:32.000 So not only did we solve the problem of doing 0:14:32.000 --> 0:14:34.960 it for like one pitcher in a way that's really actionable, 0:14:35.320 --> 0:14:38.240 but we solved the problem doing it for every game 0:14:38.560 --> 0:14:41.840 every day, you know, so that you have the data 0:14:41.880 --> 0:14:42.720 when you wake up. 0:14:44.720 --> 0:14:46.920 So you guys are doing it at scale. The answer 0:14:46.920 --> 0:14:48.760 to why it is that that there is in fact 0:14:48.800 --> 0:14:51.680 an economy of scale, a benefit of scale. 0:14:51.480 --> 0:14:53.640 Yes, which you have, Yes, exactly, Yeah. Yeah. 0:14:54.600 --> 0:14:57.960 What's what's a specific example of a thing that a 0:14:58.040 --> 0:15:01.200 coach might tell a pitcher in response to your report 0:15:01.240 --> 0:15:04.040 to try and get them to throw whatever differently? 0:15:05.440 --> 0:15:06.400 A really. 0:15:08.080 --> 0:15:12.720 Common low hanging fruit type of piece of feedback that 0:15:12.840 --> 0:15:16.120 often comes out of the reports is how a pitcher 0:15:16.400 --> 0:15:21.040 is using their lead arm in a typical pitching motion, 0:15:21.320 --> 0:15:23.720 the pitcher will reach forward with their lead arm. 0:15:24.160 --> 0:15:25.600 So, just to be clear, the lead arm is the 0:15:25.680 --> 0:15:27.520 arm that is not holding the ball. 0:15:27.400 --> 0:15:29.080 Right right, yeah, the arm in front of you. 0:15:29.200 --> 0:15:29.480 Yeah. 0:15:29.680 --> 0:15:30.440 Yeah. 0:15:30.480 --> 0:15:32.720 A pitcher will reach forward with that lead arm while 0:15:32.760 --> 0:15:35.600 the rear arm is holding the baseball, and they'll rotate 0:15:35.800 --> 0:15:38.960 that lead arm really hard, and that's the thing that 0:15:39.040 --> 0:15:43.640 kind of initiates the torso rotation. So a very common 0:15:43.720 --> 0:15:47.040 flaw that we see is if a pitcher has a 0:15:47.200 --> 0:15:51.400 very vertical pitching arm, they're pitching the ball way over 0:15:51.440 --> 0:15:54.520 the top of their head, but their lead arm when 0:15:54.560 --> 0:15:56.760 they pull it through, when they swing it through, they 0:15:56.800 --> 0:15:58.400 swing it in a very flat. 0:15:58.160 --> 0:16:02.040 Plane as horizontal. 0:16:01.480 --> 0:16:05.520 Horizontal, right, yeah, That is not a very efficient plane 0:16:05.560 --> 0:16:07.320 to use when you're throwing the ball on a very 0:16:07.400 --> 0:16:10.560 vertical plane. So a very common low hang piece of 0:16:10.560 --> 0:16:15.280 fruit feedback that comes out of the reports is having 0:16:15.320 --> 0:16:18.960 pitchers just try to rotate their lead arm pull with 0:16:19.000 --> 0:16:21.840 their lead arm in a more vertical plane to better 0:16:21.920 --> 0:16:24.200 match what their torso is doing, a better match with 0:16:24.280 --> 0:16:25.240 their pitching arm is doing. 0:16:26.320 --> 0:16:28.160 That's a good one. I feel like that one's so 0:16:28.240 --> 0:16:30.360 simple that you don't want it to get out that 0:16:30.480 --> 0:16:31.840 everybody ell to start looking at it. 0:16:31.920 --> 0:16:34.440 No, I mean really, I mean like it's this has 0:16:34.480 --> 0:16:37.080 happened when I went to talk to a team and 0:16:37.120 --> 0:16:40.080 we talked about some pieces, you know, some low hanging fruit, 0:16:40.120 --> 0:16:43.120 and they're like, okay, great, we'll take the lead arm thing, 0:16:43.160 --> 0:16:45.800 will implement it everywhere. And I'm like, well, what about 0:16:45.840 --> 0:16:48.160 like the other ten pages of the report. 0:16:49.800 --> 0:16:50.760 Good with the lead arm thing. 0:16:52.200 --> 0:16:56.240 Thanks by so for you. The end of the train 0:16:56.600 --> 0:16:59.840 is the report. But that report goes to the coach, right, 0:17:00.000 --> 0:17:03.680 and so presumably the meaningful change hasn't happened yet, right, 0:17:03.680 --> 0:17:06.040 It has to somehow get from the coach to the picture. 0:17:06.720 --> 0:17:08.840 And like, I know that piece of it is not 0:17:09.640 --> 0:17:11.679 your business now, but it was kind of your business 0:17:11.680 --> 0:17:13.959 when you were at the Dodgers. Presumably you're familiar with 0:17:14.000 --> 0:17:16.240 it now, Like, how does that piece of it work? 0:17:16.320 --> 0:17:18.120 Is it like the pitching coach is like reading from 0:17:18.160 --> 0:17:20.320 the report to the picture. I imagine not, but I 0:17:20.359 --> 0:17:20.679 don't know. 0:17:20.960 --> 0:17:27.480 No, definitely, definite, definitely not. Even at the Dodgers, my 0:17:27.680 --> 0:17:32.000 role was not being the one to coach the players. 0:17:32.200 --> 0:17:34.800 It's like, whatever you do, don't talk to the pictures. Man, 0:17:34.960 --> 0:17:35.879 go back to your computer. 0:17:36.960 --> 0:17:39.760 Yeah no, no, I mean thankfully I got to be 0:17:40.400 --> 0:17:43.280 in the in the room as you know, dark Walld, 0:17:43.320 --> 0:17:47.439 the interaction is happening. But I think that is the 0:17:47.440 --> 0:17:52.439 the art of coaching that is so important is understanding 0:17:52.960 --> 0:17:55.960 the picture and how the pitcher thinks about themselves and 0:17:56.080 --> 0:18:00.080 giving the right feedback to have the picture do the 0:18:00.119 --> 0:18:02.240 thing that you that you want them to do. 0:18:02.800 --> 0:18:06.000 Uh huh. Knowing how to talk to a player in 0:18:06.040 --> 0:18:09.360 a way that is not generic. Presumably different pitchers need 0:18:09.400 --> 0:18:12.680 to hear different things, even if the outcome is the same, right. 0:18:12.760 --> 0:18:15.399 I mean, there's a there's a classic like debate in 0:18:15.520 --> 0:18:19.200 baseball of like do you swing down on the ball 0:18:19.320 --> 0:18:22.280 or do you swing with an uppercut? And in reality, 0:18:22.440 --> 0:18:25.760 like the batpath is an arc, the path goes down 0:18:25.800 --> 0:18:28.440 and then the path goes up. But some coach, some 0:18:28.480 --> 0:18:32.120 players respond better when a coach will say swing down 0:18:32.160 --> 0:18:34.960 on the ball, and some players spawn better when a 0:18:35.000 --> 0:18:37.040 coach will say, you know, have get a little bit 0:18:37.040 --> 0:18:39.640 more uppercut, do your swing. But in reality, you know, 0:18:39.960 --> 0:18:41.760 the bat goes down and the bat goes up. So 0:18:41.800 --> 0:18:44.639 it's like really understanding what is the player what helps 0:18:44.640 --> 0:18:45.520 the player the most? 0:18:45.960 --> 0:18:48.320 So I heard you use this phrase in another interview 0:18:48.680 --> 0:18:50.600 that I think is kind of what you're talking about here, 0:18:50.600 --> 0:18:54.159 And it's feel versus real? What is that? What is 0:18:54.160 --> 0:18:55.080 feel versus real? 0:18:55.320 --> 0:18:59.480 So it's exactly what we're talking about. Is sometimes with 0:18:59.560 --> 0:19:03.080 the apt let feels like they're doing is not. 0:19:03.359 --> 0:19:05.000 What's actually happening. 0:19:05.200 --> 0:19:08.480 But if you understand what the athlete feels like they're doing, 0:19:08.920 --> 0:19:12.720 you can give feedback that interacts with how they're feeling. 0:19:12.960 --> 0:19:16.000 So if they feel like they're swinging down on the 0:19:16.040 --> 0:19:18.880 ball and you give them feedback related to that, even 0:19:18.880 --> 0:19:21.560 if they're actually swinging with an uppercut, you know, the 0:19:21.640 --> 0:19:24.119 important thing is like how do they feel and how 0:19:24.160 --> 0:19:26.520 do you give them feedback related to how they feel, 0:19:26.760 --> 0:19:30.400 which then impacts is real. But it's understanding the interplay 0:19:30.400 --> 0:19:31.600 between the feel and the real. 0:19:32.920 --> 0:19:36.840 Yeah, it's wild that, like there is there is a 0:19:36.920 --> 0:19:39.800 human being doing a thing, and then there's this huge 0:19:39.920 --> 0:19:44.560 industrial machinery of your company and the team and all 0:19:44.600 --> 0:19:47.639 of these scientists and all these cameras and computers that 0:19:47.680 --> 0:19:50.640 are trying to get this human being to change their 0:19:50.640 --> 0:19:52.320 behavior in a very subtle way. 0:19:52.440 --> 0:19:52.960 Yeah, exactly. 0:19:53.000 --> 0:19:57.560 It's a behavior that is in many ways intuitive right, 0:19:57.600 --> 0:20:00.520 it's sort of partly conscious but partly intuitive him like 0:20:00.560 --> 0:20:05.480 there's a really interesting human being at the center. 0:20:05.200 --> 0:20:05.840 Of all of this. 0:20:06.160 --> 0:20:09.080 Yeah, And I think that's also why it's so important 0:20:09.600 --> 0:20:12.720 to have the coach in the loop, because they understand 0:20:13.160 --> 0:20:15.800 the human being even beyond just like the feel versus 0:20:15.800 --> 0:20:19.240 real aspect of like did the athlete not get a 0:20:19.240 --> 0:20:21.560 good night's sleep last night, then maybe today is not 0:20:21.760 --> 0:20:24.000