Career Minor Leaguers Take Over MLB Pitching Staffs

[Brad K]

First of all, credit to luckymann for identifying this issue. He first posted about it in the historical forum. See his post HERE.

My studies have confirmed his reports of excessive career minor league pitchers being on MLB pitching staffs. I checked a backup of my America's Team save for June 1 1975. 34% of the MLB SPs were Career Minor Leaguers. A third. And this isn't counting cup of coffee players. Only those who never made it at all to MLB.

The RPs are even higher at 45%! This includes CLs. And again, no cup of coffee players included.

Position players are a much more reasonable 19% but I still think most would consider this a bit high.

luckymann noticed the issue with a full minors save on OOTP24. My data is from OOTP23 since I don't change horses mid stream. Whatever version a save starts on, that's where it stays. However since our data is similar I believe there were no changes on the issue between 23 and 24.

I have another save from 1964 where the percentages are much lower however that may be affected by the age of the sim. It is starting only its fourth season while the America's Team save data above is from its 15th season. In the young save the percentages are SP 14%, RP 13%, Position Players 14%, Overall 14%. Or it could be it's more of a problem in the 70s than the 60s,

I really like the idea of career minor leaguers making MLB. And I like that full minors also boosts the stats of guys who got a chance, but not much of one, in MLB like Connie Johnson and Carlos Bernier. But what luckymann and I have experienced in these saves is unacceptable. It's more fictional than a plausible alternative history.

Text files with data below.

[uruguru]

So there are a couple of things that could be causing something like this

1) The OOTP AI uses 2024 criteria which means it might promote players in a 1975 save that would not have gotten consideration in the real world (such as high-strikeout hitters or pitchers with a good FIP instead of ERA)

2) The OOTP AI evaluates players on their current season ratings, something that is not possible in the real world. So basically the AI has a "cheat code" on which players are going to have a bad year in the MLB and can demote them prematurely. Conversely, the AI will know (based on ratings) that an milb player is going to have a good year and will promote him prematurely.

[Charlie Hough]

I posted in the historical thread about this, but there are a few potential factors to consider.

1. Are you using development or recalc?

2. Were these games created in OOTP23 or OOTP24, and were they created in an older version of OOTP24?

With development, you're easily going to see a lot of this. Also, in earlier versions of OOTP24, there were major issues with incorrect player ratings calculations for players who had outstanding stats in class A. It also affected players in AA. Players who were great hitters or great pitchers at low levels of minors would sometimes get MLB-equivalent ratings, making them instant major leaguers.

If none of these are factors, some of this might still happen due to what Uruguru mentioned in his second point. We would need to check many saved games to see a true ratio, though. Since the AI always places the biggest emphasis on player ratings, there are many players who had very good historical minors stats who will get a chance to play in MLB because their ratings are good. They also get a look due to roster decisions, injuries, trades, free agency, or players ahead of them performing poorly at the MLB level.

In many cases, unlike how they did in real life, these players will get a shot, and, once they post good stats at the MLB level, that solidifies their position on MLB rosters. They not only have good ratings now: they have good MLB stats too, so the AI treats them accordingly. In addition to looking at ratings and stats, it also looks at ages and contracts too, where a younger and cheaper player might potentially be a better investment in the long run. The OOTP AI has no idea whether a player was a career minor leaguer or not, so it couldn't possibly consider that.

How much alternate history you're willing to tolerate varies from person to person. If you're playing with historical minors, and especially if you're using development, then alternate history is going to be much more prominent. Even if you're using recalc, you might still see quite a bit of it. You could use the option to block players from reaching the majors, but obviously you're not looking to do that. You just want less frequency of this.

But, if you start coding the game to have the AI take into account whether a player ever made it to the majors, then you're artificially manipulating things in an unrealistic way. So, as long as we make sure that MLEs and player ratings are being calculated correctly, and the AI is evaluating player ratings and stats correctly, then I would rather leave the AI alone in this regard. Otherwise, it might make dumb decisions and be more likely to hold a player down in the minors when he could potentially help the team.

[luckymann]

Great points, chaps, and thanks.

I encourage some variance, in fact I'll tolerate a moderate amount.

In my original post over on the historicals, I didn't request changes be made. What I'd like to know, as I posted over there, is whether changing my AI eval settings to reduce the weight of ratings on how the engine deploys these guys will help reduce in any way the predominance of career MiLBers - especially pitchers - in that save.

I have already ticked them down from 52/26/16/6 to 50/30/15/5 but it's too soon to see if this helps or hinders and it is tough to tell anyway unless there's a massive and obvious shift.

If this is the right course of action, I'll do it more aggressively again in the next offseason, I was thinking to 45/35/15/5.

But I'd dearly like to know if this is the right or wrong course of action, or neither, before doing so.

Thanks

G


Oh, and Charlie, the save in question was created in v24 in October last year, so about mid-release. 3-year recalc with dev on.

[Brad K]

Originally Posted by Charlie Hough

I posted in the historical thread about this, but there are a few potential factors to consider.

1. Are you using development or recalc?

2. Were these games created in OOTP23 or OOTP24, and were they created in an older version of OOTP24?

With development, you're easily going to see a lot of this. Also, in earlier versions of OOTP24, there were major issues with incorrect player ratings calculations for players who had outstanding stats in class A. It also affected players in AA. Players who were great hitters or great pitchers at low levels of minors would sometimes get MLB-equivalent ratings, making them instant major leaguers.

If none of these are factors, some of this might still happen due to what Uruguru mentioned in his second point. We would need to check many saved games to see a true ratio, though. Since the AI always places the biggest emphasis on player ratings, there are many players who had very good historical minors stats who will get a chance to play in MLB because their ratings are good. They also get a look due to roster decisions, injuries, trades, free agency, or players ahead of them performing poorly at the MLB level.

In many cases, unlike how they did in real life, these players will get a shot, and, once they post good stats at the MLB level, that solidifies their position on MLB rosters. They not only have good ratings now: they have good MLB stats too, so the AI treats them accordingly. In addition to looking at ratings and stats, it also looks at ages and contracts too, where a younger and cheaper player might potentially be a better investment in the long run. The OOTP AI has no idea whether a player was a career minor leaguer or not, so it couldn't possibly consider that.

How much alternate history you're willing to tolerate varies from person to person. If you're playing with historical minors, and especially if you're using development, then alternate history is going to be much more prominent. Even if you're using recalc, you might still see quite a bit of it. You could use the option to block players from reaching the majors, but obviously you're not looking to do that. You just want less frequency of this.

But, if you start coding the game to have the AI take into account whether a player ever made it to the majors, then you're artificially manipulating things in an unrealistic way. So, as long as we make sure that MLEs and player ratings are being calculated correctly, and the AI is evaluating player ratings and stats correctly, then I would rather leave the AI alone in this regard. Otherwise, it might make dumb decisions and be more likely to hold a player down in the minors when he could potentially help the team.

"Since the AI always places the biggest emphasis on player ratings,"

"once they post good stats at the MLB level, that solidifies their position on MLB rosters."

I don't follow this at all. Especially since there are settings to tell AI how much weight to give ratings and how much weight to give stats.


"as long as we make sure that MLEs and player ratings are being calculated correctly, and the AI is evaluating player ratings and stats correctly, then I would rather leave the AI alone in this regard."

No reason I see to change AI. AI should not reject players based on them being career minor leaguers. It should reject them based on having inadequate ratings and in game stats to play MLB.I see the problem as being there are too many players rated MLB capable based on their real life minor league performance.

Also, I haven't noticed OOTP thinning the herd. Lots of minor league players quit after a year or two despite being successful. But as far as I know OOTP doesn't simulate this. I'm not saying minor leaguers should be retired by OOTP based on them leaving baseball at a young age real life. However some of younger career minor leaguers should leave the game each year and the pool of those subject to random removal should include those who played several years. Since OOTP allows two season career minor leaguers to have decade long MLB careers it should allow career minor leaguers to quit baseball age age 22 even if they played minor league ball until they were 28.

It's possible that minor league players are rated correctly and AI evaluates them correctly but the problem is OOTP leaves too many in the game.

[Rain King]

Can we get some clarification on what settings are being used when seeing this?

Specifically, is this using Recalc or the Development Engine?

[Brad K]

Mine on 1975 are recalc 3 year not weighted development on.

Mine on 1964 are recalc 3 year weighted development on.

luckymann has development on but I don't recall his recalc settings.

[Brad K]

Originally Posted by Charlie Hough

How much alternate history you're willing to tolerate varies from person to person. If you're playing with historical minors, and especially if you're using development, then alternate history is going to be much more prominent. Even if you're using recalc, you might still see quite a bit of it. You could use the option to block players from reaching the majors, but obviously you're not looking to do that. You just want less frequency of this.

Do you think 34% is OK? Are you OK with 45%? Is anyone? At what point does it go from alternate history to implausible and fiction?

[Rain King]

So, Recalc is setting the ratings based on how the player performed that year (or years).

However, with development on and most minor league players being of prime development age...I would imagine that most minor leaguers are going to increase their ratings over the course of a season.

That gives you minor leaguers that are "over-rated" for how they performed historically...and so I'm not surprised that combination means that career minor leaguers are over-performing and reaching the major leagues.

I would probably recommend turning the development settings way down since I imagine the main reason you want this feature on is more for the Aging settings.

A good test would be to run one of the same seasons with Development off and seeing how different the results are.

[Brad K]

Do the development rate settings work on historical players? Or is that a fictional player only feature? Like morale and coaching, how can it work with historical players when the game judges itself on how close the output is to historical?

Why doesn't the game work with default settings, which are 3 year reacalc weighted development on?

[Charlie Hough]

Originally Posted by Brad K

"Since the AI always places the biggest emphasis on player ratings,"

"once they post good stats at the MLB level, that solidifies their position on MLB rosters."

I don't follow this at all. Especially since there are settings to tell AI how much weight to give ratings and how much weight to give stats.

That's the point. If you use the default settings or even higher ratings emphasis, the AI puts major emphasis on player ratings, and when it emphasizes stats, it's also going to defer to players who are performing well. I assume that it weighs their simulated MLB stats higher than their simulated AAA stats as well, so that's where player ratings, combined with good MLB stats during their call-ups, might contribute to what you're seeing.

Also, I haven't noticed OOTP thinning the herd. Lots of minor league players quit after a year or two despite being successful. But as far as I know OOTP doesn't simulate this. I'm not saying minor leaguers should be retired by OOTP based on them leaving baseball at a young age real life. However some of younger career minor leaguers should leave the game each year and the pool of those subject to random removal should include those who played several years. It's possible that minor league players are rated correctly and AI evaluates them correctly but the problem is OOTP leaves too many in the game.

If you code the game to have good minor leaguers retire after just one or two seasons in the minors, then a lot of customers are going to be pretty upset. Imagine having a player who looks like a good prospect and is playing and developing well, and then you lose him for no good reason. Yes, it happens in real life on occasion, but it's very rare, and, even if you model it in OOTP, it's not going to significantly reduce number of career minor leaguers that you're seeing reach MLB.

Mine on 1975 are recalc 3 year not weighted development on.

Mine on 1964 are recalc 3 year weighted development on.

luckymann has development on but I don't recall his recalc settings.

So, these facts are crucial. As Rain King correctly pointed out, with development turned on, it's theoretically possible for quite a few minor leaguers to get a development ratings boost during a season. That might lead to call-ups and lead them to become regulars. Don't forget that money, transactions, injuries, and poor player performance by current MLBers can also be factors in them getting promoted or prioritized.

If this happens and they don't have any more real-life stats remaining for the next season's recalc and beyond, then they're going to stay at those increased ratings. Through continued development, they might get even better.

In contrast, if these are players who still have real-life stats remaining for recalc, then it's more likely that the issue is purely that they had really good AAA stats in real life, so they have pretty decent MLB value, and the AI calls them up and uses them. Or, it's a case of them developing and getting better during the season, before the next recalc, and they end up playing in MLB for that season until their ratings get recalculated downward. But the process might randomly repeat itself in the next season.

Do you think 34% is OK? Are you OK with 45%? Is anyone? At what point does it go from alternate history to implausible and fiction?

Given what I just discussed above, I think you're focusing on a macro-level number too soon, without a full account of all the potential factors that could be causing this. We need to know a LOT more about the players involved, their stats, when their stats ran out, if their ratings changed through development in your game, the ratings and stats for the MLB players they replaced, etc. We need to know these things before we can determine whether there is any issue at all and which percentage is ultimately acceptable. Those percentages might be perfectly reasonable and expected under the right conditions, and those conditions might exist in your game.

[Brad K]

"If you code the game to have good minor leaguers retire after just one or two seasons in the minors, then a lot of customers are going to be pretty upset. Imagine having a player who looks like a good prospect and is playing and developing well, and then you lose him for no good reason. Yes, it happens in real life on occasion, but it's very rare, and, even if you model it in OOTP, it's not going to significantly reduce number of career minor leaguers that you're seeing reach MLB."

Concerning the first bold, why is that any worse that giving Jim Palmer a career ending injury in 1970?

Concerning the second bold, you don't know how often it happens. You're guessing.


"Given what I just discussed above, I think you're focusing on a macro-level number too soon, without a full account of all the potential factors that could be causing this."



C'mon. 34% and 45%? Well, if we don't start with that then there's nothing to discuss.

What we have in the responses in this thread are basically two things.

1) OOTP is working correctly or it was fixed to work correctly after the versions that created the data.

2) Any observed instances of it not working are a PIC problem. (Person In Chair, as in we're using the wrong settings.)

It's said there isn't enough data. 34% and 45% are being dismissed requesting more detail without acknowledgement 34% and 45% show a problem. The way this is going there will never be enough data, so why would luckymann or I be motivated to produce any more to address the speculations of the cause.

They're not our speculations. Those speculating, compile your own data to prove them. Why is it acceptable for others to speculate and ask that it be accepted as fact?

[uruguru]

Originally Posted by Brad K

Concerning the first bold, why is that any worse that giving Jim Palmer a career ending injury in 1970?

In 100% of my recalc sims, I completely turn off Long-Term injuries because I have "Retire According to History" turned on to account for career-ending injuries. To compensate, I do raise the rate of short-term injuries to OOTP Normal. It's made the games a lot more enjoyable, imo.

I thought I was going to break my laptop that time when 17-year-old hot prospect Larry Dierker had a career-ending injury in his first MLB start, lol.

The bottom line is that, if you play with minor leagues, you are going to see a lot more minor-leaguers on the MLB rosters than in the real world. This is in part due to the X-ray vision the AI has with regards to ratings. If you want to eliminate that, you need to remove ratings from the player evaluation.. which I do. I set them to 0/60/30/10.

There's never going to be a perfect sim, though. You just need to get them in a place where you are comfortable that it feels realistic enough. However, the more you know about baseball history and stats, the more difficult this is.

[Charlie Hough]

Originally Posted by Brad K

Concerning the first bold, why is that any worse that giving Jim Palmer a career ending injury in 1970?

Because users have control over whether injuries exist in their game, the frequency, and whether they want that randomness injected into their games. If you code random minor leaguers to retire after a couple of seasons in the minors, then it's unlikely to be an option that users can control. Hence, some users might get upset by it. I wouldn't be one of them, but I know the developers are cautious about balancing user preferences and fun with the level of realism and what users don't get to control. Of course, a user could choose to un-retire the player, but then I think he becomes a free agent, and you may have to use the editor to put him back on the correct team and give him the contract he had. If you play in challenge mode, you obviously can't do this, so it's not necessarily a solution for all users.

Personally, I wouldn't mind an occasional minor league retirement happening with the frequency that it does in real life, but that's still not going to do anything meaningful to address the percentage you're seeing, and I'll show you some data on this shortly.

Concerning the second bold, you don't know how often it happens. You're guessing.

No, I am not guessing. If you have premium access to Baseball Cube, you can search publicly announced minor league retirements from 2012 to the present. This is a small sample of what's currently available for free, from 2023 and 2024:

https://www.thebaseballcube.com/page...nx&view=retire

If you subscribe to premium, you can see more data like this, dating back to 2012. It's nowhere near complete data, and it's very limited historically, especially since most retirements probably aren't announced publicly or sufficiently tracked. But we have enough of this publicly announced retirement data to make educated assessments of the frequency, and we can also take a look at the entire historical minor league database. I haven't conducted a large-scale analysis of it, but I can tell you that I've reviewed some sizable sample sets of players in the past, just taking a look at real-life minor league progressions and stats, and it's basically the same story.

It's definitely relatively rare for a player to only play only a season or two in the minors and then no longer have stats, especially at younger ages in the modern era, when they're typically in the minors. Usually players stay in the minors for at least several seasons before they retire, injured, or they get released and no one picks them up.

Those instances of just playing a year or two are a pretty rare occurrence, especially relative to the thousands upon thousands of other players who play at least several seasons or more. You can look at the historical minors data and see it for yourself. Few players are willing to give up on their MLB dream so easily, and organizations typically aren't in a hurry to give up on them either. But, in a lot of cases, they don't necessarily retire. They're obviously forced into leaving the minors because they never developed enough and presumably didn't get another contract. Again, this is an entirely separate topic, as it's not really going to make a huge difference in what you're concerned about anyway.

Ultimately, if you take all of this up with the developers, they are likely going to ask similar questions about your saved game and seek similar data and information on the players, historical stats, and ratings in your game. There's a limit to how much might be needed or might be feasible to review, but I highly doubt that they are just going to look at a macro-level number and immediately conclude that something is broken.

The number might look odd at first, but you have to do the due diligence, narrow down what's actually happening, why it might be happening, and work with more complete data. Then you can determine if something isn't right, isn't ideal, and needs to be tweaked or fixed. That's always been the general process around here, so you can disagree with it if you prefer, but I don't think that will get you very far in getting this reviewed or tweaked. Sometimes it's frustrating when people ask for more data or information, or they challenge initial conclusions or raise potential counter-points. I've been through it countless times, but it's more than reasonable. So, I'm not going to get into any unnecessary arguments about it.

[Brad K]

Originally Posted by Charlie Hough

Ultimately, if you take all of this up with the developers, they are likely going to ask similar questions about your saved game and seek similar data and information on the players, historical stats, and ratings in your game.

Thank you for preparing me for the interrogation!

Originally Posted by Charlie Hough

There's a limit to how much might be needed or might be feasible to review, but I highly doubt that they are just going to look at a macro-level number and immediately conclude that something is broken.

34% and 45% equal "He's dead Jim."

[Rain King]

I mean, if you don't want any kind of discussion regarding potential solutions to the potential issue utilizing the settings available in the game then what are we doing here?

Go make a post in the bug forum.

[Brad K]

How can there be a solution offered when the major position is to deny there's a problem?

[Rain King]

Nevermind

[Brad K]

Originally Posted by Rain King

Nevermind

Last edited by Rain King; Today at 05:30 PM. Reason: Not knowing when to leave someone blocked.

Thank you.

[fredbeene]

Originally Posted by Brad K

First of all, credit to luckymann for identifying this issue. He first posted about it in the historical forum. See his post HERE.

My studies have confirmed his reports of excessive career minor league pitchers being on MLB pitching staffs. I checked a backup of my America's Team save for June 1 1975. 34% of the MLB SPs were Career Minor Leaguers. A third. And this isn't counting cup of coffee players. Only those who never made it at all to MLB.

The RPs are even higher at 45%! This includes CLs. And again, no cup of coffee players included.

Position players are a much more reasonable 19% but I still think most would consider this a bit high.

luckymann noticed the issue with a full minors save on OOTP24. My data is from OOTP23 since I don't change horses mid stream. Whatever version a save starts on, that's where it stays. However since our data is similar I believe there were no changes on the issue between 23 and 24.

I have another save from 1964 where the percentages are much lower however that may be affected by the age of the sim. It is starting only its fourth season while the America's Team save data above is from its 15th season. In the young save the percentages are SP 14%, RP 13%, Position Players 14%, Overall 14%. Or it could be it's more of a problem in the 70s than the 60s,

I really like the idea of career minor leaguers making MLB. And I like that full minors also boosts the stats of guys who got a chance, but not much of one, in MLB like Connie Johnson and Carlos Bernier. But what luckymann and I have experienced in these saves is unacceptable. It's more fictional than a plausible alternative history.

Text files with data below.

is there still no option to prevent players who never made it to the majors, not allowed. like ghost players or fictional