Is AI in the MLB dugout CHEATING?

[00:00:00]Is everyone cheating again?

[00:00:03]Not exactly. Just the case of iPads ruining everything again. I'll break down the whole situation right now.

[00:00:09]>> on May's plaining >> Rapsodo and Hawkeye all tell you this information. Paul Skenes has added two pitches to his arsenal going into 2025.

[00:00:17]>> No! Don't do it! Don't!

[00:00:21]>> Buckle up. It's time for some May's plaining. [music] >> Like and subscribe so you don't miss any video in the future. And if the mood strikes you, go check out some Mayday merch over at maydaywear.shop. Where else are you going to get a zero-time All-Star mug? Nowhere, that's where.

[00:00:39]Yes, Major League Baseball is restricting iPad usage in dugouts to prevent the tablets from running artificial intelligence to help make strategy decisions. Yes, you heard that correctly. AI dugout decisions? So, the tablets have access to video and league-provided data and also include a custom tab where teams could access documents and more recently other programs. The gist is the custom tab feature on the iPads became [music] inaccessible on Wednesday night when the second half of the season started and they are completely banned moving forward. MLB executive vice president of baseball operations Morgan Sword, yes, that is his real name, said in a June 11th memo to general managers, assistant GMs, and video coordinators, quote, "In many cases, the custom tab had expanded the use of the dugout iPads beyond their originally intended purpose to include recommendations regarding substitutions, pitch calling, [music] and other in-game decisions traditionally made by players and coaches." Former teammate of mine and long-time major leaguer Adam Ottavino added fuel to the fire when he claimed on his Baseball and Coffee YouTube stream that the Mets had used an AI program earlier this season that Steve Cohen paid several hundred thousand dollars for before MLB cracked down. What is worth hundreds of thousands of dollars? They got got.

[00:02:00]We're going to talk about that in a moment. The Mets were asked about what Ottavino said and did not comment.

[00:02:06]Yikes, things are just going really well over there. Now, let's take a step back here and let's get into the the details because I know for many of you that probably sounded like an Apple keynote speech or something like let me get you caught up, okay? Number one, let's start on it.

[00:02:22]What are the iPads actually used for?

[00:02:25]We've had a lot of people commenting on this on Twitter and like I would never look at the iPad. I just knew how to play and I never needed that stuff.

[00:02:34]False. When I played 3 years ago, the iPads were for a few very distinct things. One, video. So, the players would use video. I would throw a game, I would make a pitch somewhere, or make a mistake somewhere, give up a home run and be like I felt like that got there, did it, and I could check really quick to see if I actually hit the spot because we get a little bit of pitcher psychosis when we can't really see the right angles or the catchers are so good at receiving. We just want to make sure the things that we think are happening in our heads are actually happening in the game. That was mostly what I was doing. Or we'd be out in the bullpen, especially in City Field, for example, where the angle to the plate you're in when you're in the bullpen, you can't see where the pitches are. So, you have to like rewatch them on the iPad and we didn't really have a feed of the game either because of the cheating scandal with the Astros and everybody else. So, we didn't have a game feed either. So, we couldn't we didn't know actually what was happening and how pitches were being executed, so we used the iPads for that as well. Then, the custom tab at that time was mostly just documents. So, we would have scouting reports put through by by the analytics department every single series. They'd be like, "Hey, here's your scouting report against this team. You can use it, you can not use it." I had my own report that I made put in there as well that I could pull up and just kind of review. So, I just check the iPad instead of having a physical copy with me. That's what it was for. That is what the iPad was generally for was just like so you didn't have a big pack of paper, you just had your digital version. We read books on iPads now. Everybody does everything on iPads.

[00:04:02]It is very much in that realm, okay? So, the custom tab was also for coaches to have like positioning, alignment, all these other things that they wanted to be able to review. It's all stuff that was prepared before the game and might have been prepared using I don't want to say AI specifically, but like algorithmically, they have a program that spits it out a similar way every single time. But, when you were in the game, you only had what data and what you prepared before the game to review. That was always my understanding and so that lived in a custom tab. Now, as of late, I guess, from what I hear and according to Eno Sarris's article here as well, is that tab or what went on that tab got laxed a little bit, meaning instead of it just being documents, you then had something you could interact with and pull information up. Let me just kind of take one step back here and talk about how you're building tools to get yourself basically to the results you're trying to find as fast as possible. It's the same kind of innovation that happened with Google search. How can we be more efficient and get rid of the tedium in between? I have a question and I need an answer. How quickly can we get there? This is just a natural, I think, development on that side and now we have AI and everyone talks about AI all the time. What that means in this case is literally just how can we pull out the information that we need to pull out? I myself have built some things that we use here that are not They're not AI tools. They We don't use AI to make decisions. It's just I used AI tools to make the thing that we need to use, but the thing we need to use is just normal programs that always work.

[00:05:42]Ben zero go, "Oh, what if we had a feature that when this thing happened, then it automatically put it over here?"

[00:05:46]That's where the efficiency and the optimization starts to come from, and I think that's what happened here. So, the next step was likely, and this is just a theory of mine, but the next step was likely, "Okay, now we have a search function that takes us somewhere in those documents." So, like the documents are no longer like just a PDF, they're loaded in and there's like a searchable like webpage you can check.

[00:06:09]So, you have like this little database of the things you already prepared, and instead you go, "Okay, uh uh righty, whatever up, I want to Where do we put our outfielders?" And then it pulls up that point, so you don't have to go through all the documents as quickly as you can while the game's going on. That makes sense to me. Speed. Speed, speed, speed. We've already made these decisions before the game, let's just reference what that decision was. But then what happened was the actual real-time information. The the recommendations started to be part of that as well. Like, uh "Okay, this stuff has happened up to this point. Uh all of the data's being pulled from the game automatically anyways, so all you have to do is grab that data, run it through your little algorithm of how you make a pitch decision, it gives you a pitch, and then you tell it to the guy pitching." That is what seems to have happened with this uh custom tab. The anecdotes say that the Marlins, who have been one of the teams who have talked about using it, that is one of the things they were doing. So, it's almost like you're getting pitch selection data and information from somewhere that's not the dugout, and that's where the line got crossed. And instead of trying to chase down and like ban certain apps and ban certain things, they go, "This is just beyond our out of our depth." It's like the government with new technology.

[00:07:19]They're like, "Instead of just like chasing which things are problematic, let's just get rid of it all and then figure it out later." And that is what we're seeing here, which I agree with.

[00:07:28]So, the interesting thing about this ban is it kind of mirrors problems everywhere else. People are trying to be so efficient and so optimized all of the time that you're starting to lose any uh uh quality uh decision-making ability yourself. The decisions that getting made start getting objectively worse.

[00:07:48]Now, is this cheating in baseball? Not technically, but it is lazy. Yeah, I said it. I think it's lazy. The problem with AI, but not just AI, analytics in general. This is a argument made by anti-analytics people, the traditionalists in baseball if you will, is that analytics, I mean, they always say they're made-up numbers. not. Data is everywhere. Whether or not you want to look at something like data, everything is data. Data is everywhere.

[00:08:15]Everything is math. You may have heard that on a movie somewhere. Everything is math. It's just the way that you can buy it. You don't have to think about everything is data, but it is. There's data everything's a data point. You can use it as data point whether or not it's valuable or not is a different different conversation, but it is there. But, the problem with both AI and analytics is the data that those two buckets receive are past data.

[00:08:40]And baseball has things happen that have never happened before all of the time.

[00:08:46]And they also have an issue with sample size. So, outliers are constantly happening. Crazy outliers are constantly happening. Predictive models can't predict those things. So, if you are typing something in and getting an answer and then putting that answer in the game, you are literally recycling information that's already happened and making all your decisions based on only that. There's nothing to do with remembering here.

[00:09:09]There's no skill up here needed. You don't need to develop your ability to see an opportunity that is is maybe a data point that your brain sees that the model doesn't see, which will always be happening. If you're updating this model all the time, you're trying to search for these things and it makes sure it's getting all the inputs, but it will never get all of the inputs. Now, I don't know exactly how these decisions were being made or how they were being put into the game, but we have to find a way to preserve the filter of the human brain. I love technology. I love analytics. I love numbers. I love all this stuff. I'm building a thousand things myself cuz it's fun. But the moment I have a problem where I'm like, this is my question, and then I just ask and then take the answer wholesale, the moment that happens, that is an example of it not working. I think technology should be seen like this. If it can get rid of the 80% of the tedium to get you to the last 20% of being creative, you get to retain the the humanity of it, the creative part, while removing the tedium that may be a barrier to entry to doing the thing at all. I never learned to code, always wanted to learn to code, but guess what? Learning to code was tedious as hell. It's find the comma that you put in the wrong place in these 100 lines of code. No, thank you. I'm out on that. Now, being able to know how the code is working isn't as important as having a really great great question and having a great system of how you uh systematically work through to get to the outcome that you want. If you know what your question is, and you know what you the outcome you want is, that barrier to get to that point has been reduced or eliminated completely for those that want to do those things. And that's why I enjoy it. That's why I think it's it's uh really valuable, and that's where I think the value for this stuff is. That can't exist in baseball, and that would have drove me crazy if it did. If I had to go out and I go, "Oh, it doesn't matter who I am. I'm just a husk human being with physical ability, and that's all I'm here for." We need to avoid that like the plague. I think it's a great move because not even from a someone's going to get an advantage over the other people, which there will be, but it what it will do is it will if create such an efficient and optimized game that nothing unpredictable ever happens again, and then the baseball gods die. And we cannot let the baseball gods die. So, how do we maintain the humanity of the game of baseball, and how do we keep the baseball gods alive? One, I think it's just once the game begins to the moment the game ends, your preparation is what you have to work on.

[00:11:39]You can have the PDFs, you can have things that you wrote, you can have your notes you wrote, you can physically write notes. Your brain is the thing processing it, but the thing you use is your brain and your body to make baseball happen. I think that that is not that big of a an ask. It's very easy to implement, and I think that banning the custom tab in general moves us in that direction. Can you not remember what you what to do? Can you not remember the pitches that you decided you throw, the way you want to approach a guy? That's on you. There's no one there going to be there to to back you up, but I think the motivation now, especially for coaching staffs, everyone wants to keep their job, and everyone wants to prove that they know better, or they've come up with the better solution, or that everyone's trying to prove that they have the better system, and they want to they want to separate themselves. Also, inefficiency isn't good for money efficiency. And so, there's a also push to the teams that don't want to spend as much money, and they don't want to have to spend money on the actual talent and the people who can do these things, and then trust that that's going to happen. They're going to do it organizationally, so it doesn't really matter what players are on your team, you're going to get a lot of the same results. And for me, as a player, and honestly as a fan, that will hurt the game more than it saves the money, and I don't think that's being thought about as much. So, I hope that is something that we pay attention to moving forward, and that is something that doesn't get lost. Anyone who's who's participating in that, or just wants the success, it's like cheating in a video game. Like, yeah, your bot aims for you. When does that stop becoming fun? Pretty dang quick. As always, I need to know what you think about this whole thing in the comments. Are you surprised AI made it into the dugout so quickly? Are you for or against it in baseball?

[00:13:24]Don't get it wrong. Let me know down in the comments. This has been [music] May's planning. I like to thank producer David Gordon, assistant producer Henry Neeman, and also [music] assistant producer Caleb Bartholomew. And thank you for making this video happen and as always [music] drink your water, watch your baseball, and I'll see you next time on your mother's, sister's favorite [music] baseball show about AI Mayday. See you later.

[00:13:52]>> [music] [music] [music] [music]