Statcast Integration in OOTP - Page 2

matskralc! · 06-26-2025, 03:56 PM

Quote:

Originally Posted by Eckstein 4 Prez

Yeah, good point. I'd still like to play that physics-based game someday though. Also, I think you can still make it work similarly even with an outcome-based engine, right? Like, you send a guy to work on his launch angle or spin rate or whatever, and he comes back with that outcome having improved. It's not perfect but it's a way to make it feel more like a modern analytics-based approach.

If I'm understanding you, that's basically what the player development lab is right now. You send a guy to go "work on bat speed" and the little write up talks about how his coaches are going to use extra BP time and video review and strength-training and whatever to help him swing the bat a little harder. But he has no "bat speed" attribute that is going to change. He has a "HR power" attribute that changes, so now he'll hit more home runs that the game will backfill with higher exit velocities because we know that, in general, home runs have higher exit velocities.

It's all cosmetic. We could derive a stat called "bat speed" from a player's power ratings, but we're just representing "how likely a player is to hit a home run against average competition in a neutral environment" with a different number. We could then say that we send him to the lab to work on "bat speed" and then change his "bat speed" attribute which in turn changes his "HR power" attribute, but it's still just all cosmetic because "HR power" is what has to drive everything in the engine. It's just some extra steps, and extra steps are an invitation for unexpected bugs.

Bagpipes5 · 06-26-2025, 05:17 PM

Quote:

Originally Posted by jpeters1734

I feel like I’m talking to a chatbot at this point.

There is no workaround. Statcast is physics. OOTP is not. Unless the sim engine becomes physics based, you’re never going to get meaningful Statcast metrics.

I appreciate your insights into the complexities of integrating detailed metrics like Statcast into OOTP. While it's true that OOTP operates differently from a physics-based simulation, your point about the challenges in obtaining meaningful Statcast metrics within the current framework is duly noted.

However, maintaining a respectful tone in our discussions is crucial for fostering a constructive exchange of ideas. If you find that you're unable to contribute positively to the conversation, it might be best to step back rather than resorting to dismissive comments.

In the spirit of open dialogue, let's explore how OOTP could potentially evolve to incorporate elements of physics-based simulations or alternative approaches to enhance the gameplay experience. Your perspective is valuable, and I believe we can continue this conversation in a constructive and engaging manner.

Looking forward to further insights from you on this topic!

Carplos · 06-26-2025, 07:12 PM

This reminds me of a former coworker who started having chat GPT write literally every email.

Brad K · 06-26-2025, 07:19 PM

Quote:

Originally Posted by jpeters1734

I feel like I’m talking to a chatbot at this point.

There is no workaround. Statcast is physics. OOTP is not. Unless the sim engine becomes physics based, you’re never going to get meaningful Statcast metrics.

"Like"

venflaalachi · 06-26-2025, 08:00 PM

Quote:

Originally Posted by Carplos

This reminds me of a former coworker who started having chat GPT write literally every email.

He sounds exactly like an AI, especially the last two responses.

low · 06-26-2025, 08:07 PM

Skynet is real.

Bagpipes5 · 06-27-2025, 09:25 AM

Ok, well I'm gonna biggity bounce out of this convo as the thread is totally derailed and no good conversation is occuring. Instead the thread is supporting high level bullying. Tells me all I need to know. Sorry I must have emasculated all of you. Good luck!

oldfatbaldguy · 06-27-2025, 12:04 PM

The reasons were explained to you, and you didn't accept the explanation. Now you're calling people bullies and claiming to have emasculated them. So much winning!

WhiskyTango · 06-27-2025, 12:14 PM

As Brad K points out, OOTP is not a physics simulation. OOTP generates a play result and then backfills the details with stuff that mostly makes sense.

Coinciding with how it generates statistics and player performance. As discussed in another thread currently active, potential ratings predict performance outcome and current stats. It's called emulation. Very often people posting in these forums demonstrate they are playing a very different game than the one installed on their hard drive.

Brad K · 06-27-2025, 05:57 PM

Quote:

Originally Posted by WhiskyTango

As Brad K points out, OOTP is not a physics simulation. OOTP generates a play result and then backfills the details with stuff that mostly makes sense.

Coinciding with how it generates statistics and player performance. As discussed in another thread currently active, potential ratings predict performance outcome and current stats. It's called emulation. Very often people posting in these forums demonstrate they are playing a very different game than the one installed on their hard drive.

That is a great observation.

Clearly true in things like discussions on how to break players out of a slump when its not clear slumps are anything more than random. (One dev said they exist. Another said they exist only when psychology is on.)

jpeters1734 · 06-27-2025, 06:21 PM

Quote:

Originally Posted by Brad K

That is a great observation.

Clearly true in things like discussions on how to break players out of a slump when its not clear slumps are anything more than random. (One dev said they exist. Another said they exist only when psychology is on.)

psychology?

Brad K · 06-27-2025, 06:35 PM

Quote:

Originally Posted by jpeters1734

psychology?

Morale, cohesion, etc.

WhiskyTango · 06-27-2025, 08:34 PM

I like many no doubt sat slumping players. Have since learned otherwise. A truly dialed in game closes the gaps between intuition and gameplay . Now, a person may still sit a slumping player and never be the wiser. I think there's the core of us who don't understand the hostility aimed at those who simply point out observations, current threads cases in point. *Shrug* I may still sit slumping players just because it's the right thing to do!

Brad K · 06-27-2025, 08:54 PM

Ai appears not to sit slumping players. However AI removes pitchers who are aren't tired but are being shelled. The first is contrary to real baseball and the second is contrary to the oft stated principle that in OOTP each at bat is a distinct event caused by the interaction of ratings and randomness.

It is not letting out any inside secrets to let us know if OOTP simulates batter slumps or temporary one game pitcher ineffectiveness or if it doesn't. Why should we have to consider decisions for something if it doesn't exist?

It seems slumps or temporary ineffectiveness would have a detrimental effect on the "all leagues must produce historic stats" principle that permeates the thoughts of the devs and consumer assistants. So on that basis I'd say slumps and one game ineffectiveness shouldn't exist, but if it does it wouldn't be the first time a conflict between principle and programming occurred.

NoOne · 06-28-2025, 05:19 PM

Nobody is being mean.

Stats can be calculated from output, like fwar or woba etc, but anything like exit velo and such is simply not part of the game engine. you see something that takes basic outcome of a PA and generates some play-by-play that would represent that result and fairly well-distributed to seem realistic

The visible players on the field are not teh same as players on the field in a game like 'mlb the show'. They are pre-ordained puppets that don't work on physics or ai decisions etc.

it could add some of the statcast stuff.. e.g. a dobule to field position ranges can be given some random exit velo that makes sense for distance and such. But, it's just a superficial layer. *With any odd field configurations, that exit velow might have to be 500mph to make sense, lol. the distance isn't really part of the math for outcomes, either, but you see they incorporate it into the game as a superficial layer.

For the most part the increment of the game is a PA - though maybe error% chance plays out every pitch? Regardless, this is why overlall pitching stuff and pitch selection impact results but doesn't actually consider any single pitch as part of the plate appearance math.

Yes, they probably could overlay some of that stuff into play-by-play generation. It just might not be useful information. it would be randomly generated numbers to fit the outcome that was calculated for the PA. But, neither are the visible players moving around "meaningful" and people still like that.

i think some stuff woudl be easy to calculate but making sure everything is in teh right proportion and jives with every single outcome.. have to look statcast info by statcast info and look for oddities that need to be ironed out. e.g i bet incorporating exit velo and launch angle could be done with a few lines of code because that stuff strongly correlates to result of the contact.

statcast keeps track of crazy stuff like how high a ball was hit.. 4' 7" or something was the highest hit ball that ended up a home run in 2025, for example... i think it's a bit higher but in that range. you'd have to dive into distribution of such things and implement it side-by-side with the loop of code for play-by-play. some things might require a whole boatload of work to display it correctly and none of it would impact how any game plays out. 100% superficial for people to read in play-by-play.

i'd rather see time and effort invested elsewhere, but i'm just 1 data point and i don't bother with much of the superficial stuff. I already turn off 3d and facegen, replays, wpa graphs, game logs etc, lol, but i do keep box scores. As long as it doesn't slow down the game or able to turn it off, cool beans.

----------------

slumps and hot streaks are a top-down inductive thing that probably aren't true. A pitcher may strt to tip pitches or go throug a divorce off field, but otherwise have their probabiltiy of outcomes shift by real cause and effect and not just some whimsical streak.

They've shown that once you recognize or cold or hot streak, the player is already out of it, for the most part.

law of independent results... if there is a cause for a shift, sure, but things don't magically happen without a causal factor or several causal factors.

a 'confidence' rating might be rational. It woud definitely have to be hidden if it is causal to results and not a superficial display. You'd know exactly when to sit certain playrs, lol. Random things liek 'playing going through divorce' or 'child died of cancer' or 'was out drinking and loading up on the booger sugar until 5am' etc.. could impact ratings in a rational and distributed way to represent increasingly detailed effects on results. Even though it isn't currently in the game, the probability of it is already baked into it because it is part of reality and that's the source.

However, how good you did yesterday and the day before on its own does not impact probabilities. law of independent resutls. flip a coin and even if it shows heads 10x im a row, it's still 50/50 each flip. Doesn't mean baseball palyers aren't complicated and numerous things we don't see go into how well they will play one day to the next.

i think ootp sticks to this well. I don't know of any part that doesn't. Development of prospects doesn't seem to be based on results -- i'll leave open the possibility of being completely overmatched doing something harmful to the process. We could test that by moving all first year mil playrs to AAA and comparing to normal development probabilities. That would tell us if being overmatched mattered, not necessarily if results matter to development.

i recall them saying it has little to no impact somewhere in the forums (too many years to be 100% confident about that memory), but nuance is possible like shoving a 20/80 18year old in AAA, even if they did state such a thing in a broad way.

WhiskyTango · 06-29-2025, 10:13 PM

[snipped for trolling]

Bagpipes5 · 06-30-2025, 09:57 AM

Your disrespectful remarks are unwarranted and unacceptable. Comments like yours may explain why there's a lack of active participation here. Constructive discussions thrive on mutual respect. There doesn't seem to be any of that here at all.

kq76 · 06-30-2025, 11:39 AM

I'm going to close this thread because I think everything positive that could be said about the topic has probably already been said and I don't like where it's going.

Also, and this is to everyone, if we're going to use AI, let's please be upfront about it. All this AI crap is driving people crazy. It's definitely driving me crazy trying to judge what's real or not with every image or video I see. And now we have to worry about in text form too? No thank you!

06-26-2025, 08:07 PM	#26
low Hall Of Famer Join Date: May 2010 Location: Wisconsin, USA Posts: 6,102	Skynet is real. __________________ My fictional team logos and uniforms

06-27-2025, 09:25 AM	#27
Bagpipes5 Major Leagues Join Date: Dec 2001 Location: TX Posts: 458	Ok, well I'm gonna biggity bounce out of this convo as the thread is totally derailed and no good conversation is occuring. Instead the thread is supporting high level bullying. Tells me all I need to know. Sorry I must have emasculated all of you. Good luck! Last edited by Bagpipes5; 06-27-2025 at 09:27 AM.

06-28-2025, 05:19 PM	#35
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 0/1 (3)	Nobody is being mean. Stats can be calculated from output, like fwar or woba etc, but anything like exit velo and such is simply not part of the game engine. you see something that takes basic outcome of a PA and generates some play-by-play that would represent that result and fairly well-distributed to seem realistic The visible players on the field are not teh same as players on the field in a game like 'mlb the show'. They are pre-ordained puppets that don't work on physics or ai decisions etc. it could add some of the statcast stuff.. e.g. a dobule to field position ranges can be given some random exit velo that makes sense for distance and such. But, it's just a superficial layer. With any odd field configurations, that exit velow might have to be 500mph to make sense, lol. the distance isn't really part of the math for outcomes, either, but you see they incorporate it into the game as a superficial layer. For the most part the increment of the game is a PA - though maybe error% chance plays out every pitch? Regardless, this is why overlall pitching stuff and pitch selection impact results but doesn't actually consider any single pitch as part of the plate appearance math. Yes, they probably could overlay some of that stuff into play-by-play generation. It just might not be useful information. it would be randomly generated numbers to fit the outcome that was calculated for the PA. But, neither are the visible players moving around "meaningful" and people still like that. i think some stuff woudl be easy to calculate but making sure everything is in teh right proportion and jives with every single outcome.. have to look statcast info by statcast info and look for oddities that need to be ironed out. e.g i bet incorporating exit velo and launch angle could be done with a few lines of code because that stuff strongly correlates to result of the contact. statcast keeps track of crazy stuff like how high a ball was hit.. 4' 7" or something was the highest hit ball that ended up a home run in 2025, for example... i think it's a bit higher but in that range. you'd have to dive into distribution of such things and implement it side-by-side with the loop of code for play-by-play. some things might require a whole boatload of work to display it correctly and none of it would impact how any game plays out. 100% superficial for people to read in play-by-play. i'd rather see time and effort invested elsewhere, but i'm just 1 data point and i don't bother with much of the superficial stuff. I already turn off 3d and facegen, replays, wpa graphs, game logs etc, lol, but i do keep box scores. As long as it doesn't slow down the game or able to turn it off, cool beans. ---------------- slumps and hot streaks are a top-down inductive thing that probably aren't true. A pitcher may strt to tip pitches or go throug a divorce off field, but otherwise have their probabiltiy of outcomes shift by real cause and effect and not just some whimsical streak. They've shown that once you recognize or cold or hot streak, the player is already out of it, for the most part. law of independent results... if there is a cause for a shift, sure, but things don't magically happen without a causal factor or several causal factors. a 'confidence' rating might be rational. It woud definitely have to be hidden if it is causal to results and not a superficial display. You'd know exactly when to sit certain playrs, lol. Random things liek 'playing going through divorce' or 'child died of cancer' or 'was out drinking and loading up on the booger sugar until 5am' etc.. could impact ratings in a rational and distributed way to represent increasingly detailed effects on results. Even though it isn't currently in the game, the probability of it is already baked into it because it is part of reality and that's the source. However, how good you did yesterday and the day before on its own does not impact probabilities. law of independent resutls. flip a coin and even if it shows heads 10x im a row, it's still 50/50 each flip. Doesn't mean baseball palyers aren't complicated and numerous things we don't see go into how well they will play one day to the next. i think ootp sticks to this well. I don't know of any part that doesn't. Development of prospects doesn't seem to be based on results -- i'll leave open the possibility of being completely overmatched doing something harmful to the process. We could test that by moving all first year mil playrs to AAA and comparing to normal development probabilities. That would tell us if being overmatched mattered, not necessarily if results matter to development. i recall them saying it has little to no impact somewhere in the forums (too many years to be 100% confident about that memory), but nuance is possible like shoving a 20/80 18year old in AAA, even if they did state such a thing in a broad way. Last edited by NoOne; 06-28-2025 at 05:39 PM.*

06-29-2025, 10:13 PM	#36
WhiskyTango Minors (Triple A) Join Date: Mar 2025 Posts: 221	[snipped for trolling] Last edited by kq76; 06-30-2025 at 11:29 AM. Reason: snipped for trolling
	$Received Infraction$

06-30-2025, 11:39 AM	#38
kq76 Global Moderator Join Date: Nov 2002 Posts: 11,758	I'm going to close this thread because I think everything positive that could be said about the topic has probably already been said and I don't like where it's going. Also, and this is to everyone, if we're going to use AI, let's please be upfront about it. All this AI crap is driving people crazy. It's definitely driving me crazy trying to judge what's real or not with every image or video I see. And now we have to worry about in text form too? No thank you! __________________ My OOTP Wishlist \| My FAQ List OOTP Wiki \| Your Recommended Team Nicknames, By City (A Crowdsourced Project)

06-26-2025, 07:12 PM	#23
Carplos Hall Of Famer Join Date: Feb 2002 Location: Idaho Posts: 2,846 Infractions: 1/0 (0)	This reminds me of a former coworker who started having chat GPT write literally every email.

06-27-2025, 12:04 PM	#28
oldfatbaldguy Major Leagues Join Date: Dec 2020 Posts: 391	The reasons were explained to you, and you didn't accept the explanation. Now you're calling people bullies and claiming to have emasculated them. So much winning!

06-27-2025, 12:14 PM	#29
WhiskyTango Minors (Triple A) Join Date: Mar 2025 Posts: 221	As Brad K points out, OOTP is not a physics simulation. OOTP generates a play result and then backfills the details with stuff that mostly makes sense. Coinciding with how it generates statistics and player performance. As discussed in another thread currently active, potential ratings predict performance outcome and current stats. It's called emulation. Very often people posting in these forums demonstrate they are playing a very different game than the one installed on their hard drive.

06-27-2025, 08:34 PM	#33
WhiskyTango Minors (Triple A) Join Date: Mar 2025 Posts: 221	I like many no doubt sat slumping players. Have since learned otherwise. A truly dialed in game closes the gaps between intuition and gameplay . Now, a person may still sit a slumping player and never be the wiser. I think there's the core of us who don't understand the hostility aimed at those who simply point out observations, current threads cases in point. Shrug I may still sit slumping players just because it's the right thing to do!

06-27-2025, 08:54 PM	#34
Brad K Banned Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,693 Infractions: 0/2 (4)	Ai appears not to sit slumping players. However AI removes pitchers who are aren't tired but are being shelled. The first is contrary to real baseball and the second is contrary to the oft stated principle that in OOTP each at bat is a distinct event caused by the interaction of ratings and randomness. It is not letting out any inside secrets to let us know if OOTP simulates batter slumps or temporary one game pitcher ineffectiveness or if it doesn't. Why should we have to consider decisions for something if it doesn't exist? It seems slumps or temporary ineffectiveness would have a detrimental effect on the "all leagues must produce historic stats" principle that permeates the thoughts of the devs and consumer assistants. So on that basis I'd say slumps and one game ineffectiveness shouldn't exist, but if it does it wouldn't be the first time a conflict between principle and programming occurred.

06-30-2025, 09:57 AM	#37
Bagpipes5 Major Leagues Join Date: Dec 2001 Location: TX Posts: 458	Your disrespectful remarks are unwarranted and unacceptable. Comments like yours may explain why there's a lack of active participation here. Constructive discussions thrive on mutual respect. There doesn't seem to be any of that here at all.