Understanding Minor League Performance in OOTP 26

[RonCo]

Originally Posted by Matt Arnold

It could be something we can tone down in a patch.

It does not need to be toned down. It needs to be removed.

[fhomess]

Originally Posted by kidd_05_u2

Wouldn't career paths like Campbell or Jackson Holliday be modeled better as coming out of the draft with current ratings that are good enough to play at AAA? Good enough to tear through the minors and not good enough to easily succeed in the MLB, while avoiding this bad situation where potential affects stats.

I don't think how the OOTP ratings model interacts with the stat generation algorithm is diverse enough for what you're describing. The jump from AAA to MLB in game isn't as drastic as it is in real life from what I've seen. I believe they're using the potential adjustment for minor league results to simulate that bigger jump.

Originally Posted by RonCo

The idea of using potential ratings to influence current results is a bad model.

I think this is where OOTP is not yet agreeing. I'm hearing a "ends justify the means" response from OOTP on this. However, I think the issue is that the model is bad because it confuses the end user of the game. Stats should be generated according to the same ratings at all levels of play. If we're using potential ratings in conjunction with current ratings for minor leagues but not major leagues then they're playing two different games. The end user doesn't know that, and can't be expected to logically assume that, either. It's entirely counterintuitive.

Perhaps the issue that OOTP is trying to model is the wider variance of expected outcomes that prospects have based on their skill level which OOTP doesn't really have a rating for.

[RonCo]

Originally Posted by whaleheader

Your original complaint was predicated on a test whose results ended up being generated by a bug. (There is clearly a bug for player results in the minor leagues with over 550 potentials.)

The point here is that this is apparently not a bug, it is a change in behavior that the developers did by design.

[RonCo]

Originally Posted by fhomess

Perhaps the issue that OOTP is trying to model is the wider variance of expected outcomes that prospects have based on their skill level which OOTP doesn't really have a rating for.

No offense intended, but OOTP has everything they need in the thing called "current rating." If you want to have players who race through the minors, then you have your AI drop great players into too low of a level--which is what real teams have effectively done forever. When, a month later, it becomes clear a player is destroying their first level, they bump them up.

There is no value to a model that has potential as an input to current ratings, and there is a large degree of intellectual downside to implementing that design.

[Mr. Marlin]

This has been a very interesting discussion. As a stats-only player (while also relying on the written scouting reports) I haven't actually seen a player rating in at least four OOTP versions now. So, I was blissfully unaware there is a disparity between potential/current ratings and on field performance in the current game iteration. From the stats only perspective though players promoted from AAA to MLB continue to succeed and fail at about the same rate they always have in previous versions.

Stats only has always been a bit of a crapshoot when evaluating players. To play this way means willingly closing your eyes to a method of evaluating potential. But generally, players that perform above league average continue to perform at higher levels once promoted a little less than half the time. That strikes me as realistic.

[fhomess]

Originally Posted by RonCo

No offense intended, but OOTP has everything they need in the thing called "current rating." If you want to have players who race through the minors, then you have your AI drop great players into too low of a level--which is what real teams have effectively done forever. When, a month later, it becomes clear a player is destroying their first level, they bump them up.

There is no value to a model that has potential as an input to current ratings, and there is a large degree of intellectual downside to implementing that design.

I'm not sure if you're trying not to offend me or OOTP.

I may not have communicated it clearly, but I agreed with what you said and agree with this as well. I'm speculating as to OOTP's approach and reasoning for it, not saying I agree with it.

[RonCo]

Originally Posted by fhomess

I'm not sure if you're trying not to offend me or OOTP.

I may not have communicated it clearly, but I agreed with what you said and agree with this as well. I'm speculating as to OOTP's approach and reasoning for it, not saying I agree with it.

I could have worded that better.

Yes, mostly no offense to OOTP (not you!). It's a beautiful, amazing game. But sometimes the best intentions go awry. This is a decision that needs total rethinking.

[Bill Veeck]

As someone with a long history with simulations and also with baseball, what Lukas (and Matt's response re tweaking this aspect) writes makes a lot of sense to me. I'm thankful this has been brought up and debated. This is one of those areas (and in baseball there are many) in which baseball's hidden world isn't easy to replicate accurately without a lot of thought. I think a certain amount of respect or at least consideration should be given to Lukas' assessments here. He has a long history with the subtleties of the game of baseball beyond OOTP.

[low]

Originally Posted by Mr. Marlin

This has been a very interesting discussion. As a stats-only player (while also relying on the written scouting reports) I haven't actually seen a player rating in at least four OOTP versions now. So, I was blissfully unaware there is a disparity between potential/current ratings and on field performance in the current game iteration. From the stats only perspective though players promoted from AAA to MLB continue to succeed and fail at about the same rate they always have in previous versions.

Stats only has always been a bit of a crapshoot when evaluating players. To play this way means willingly closing your eyes to a method of evaluating potential. But generally, players that perform above league average continue to perform at higher levels once promoted a little less than half the time. That strikes me as realistic.

I was waiting to hear your opinion about all this since I know you've been playing that way for years. Don't you think potential plays too big of a role, though, when actual stats are concerned?

[OutS|der]

Originally Posted by Matt Arnold

Yes, as Lukas and some others have mentioned, this is meant to be closer to reality, where you have a number of prospects like a Kristian Campbell, who can tear up the minors, but still take some time to adjust to the majors. It was part of the adjustments to generate more realistic career paths for players.

We're probably a little too aggressive at the adjustment right now, though, as pointed out by some on here. It could be something we can tone down in a patch.

So what if he's tearing up the minors? Let's say MLB average is 400 for every rating, say 3-350 for AAA. So Kristian Campbell can quickly raise his current ability where he ends up being too good for the level he's at so he tears it up but when he appears to be MLB ready he's only a 375 rating. More then enough to do good in AAA but below the MLB average so he now slows down in development.

[RonCo]

Originally Posted by OutS|der

So what if he's tearing up the minors? Let's say MLB average is 400 for every rating, say 3-350 for AAA. So Kristian Campbell can quickly raise his current ability where he ends up being too good for the level he's at so he tears it up but when he appears to be MLB ready he's only a 375 rating. More then enough to do good in AAA but below the MLB average so he now slows down in development.

Yes.

[jcard]

Originally Posted by jpeters1734

I just discovered that minor league stats in OOTP donÂ’t mean a damn thing and it has completely changed how I view this game.

I ran a test where I manually edited a player; I maxed out his potential ratings at 600 while giving him straight 1s across the board for every current rating. I put him in Triple-A just to see what would happen.

The result? He slashed .303/.380/.529 with 26 HRs and a 2.8 WAR over 139 games. ThatÂ’s an OPS over .900 from a player with literally no current ability.

Why? Because the minor league engine clearly uses potential in combination with current ratings to simulate performance. That completely undermines any evaluation you might try to do using stats alone. If I were a “stats only” player, I’d look at this and assume he’s Major League ready. He’s absolutely not.

This isnÂ’t just flawed, itÂ’s actively misleading. Minor league performance should reflect current ability. Potential should only matter for development and scouting. But instead, you get fantasy output based on what the player might be someday.

Until that changes, donÂ’t trust the numbers. TheyÂ’re junk.


Edit for clarity:
The point of this thread was never to suggest that anyone regularly uses 1/600 players in AAA or to test out extreme situations. That was just a test case to isolate and expose how the sim engine behaves. The larger takeaway is this: potential now has a real and noticeable effect on minor league statistical output, even when current ratings are minimal. ThatÂ’s a shift from how the game has traditionally worked and has real implications for how we evaluate performance in the minors. It affects how we play, develop, and promote players across the board.

If you can take a second to answer, I have a couple of quick questions—on which you may be able to share experience and / or insight:

1) Was this with the (paraphrasing here) “Control Engine for Minors” option selected?

2) Does the same phenomenon—observed role of Potential Ability on current performance—occur if the test is performed with the Player Development System disabled (thus, theoretically making Potential Ability inert with regards to its more orthodox significance as a target for future Current Ability)?

Thank you in advance for any consideration and reply?

[Paul Reuschel's Mustache]

Originally Posted by fhomess

I think the issue is that the model is bad because it confuses the end user of the game.

Stats should be generated according to the same ratings at all levels of play. If we're using potential ratings in conjunction with current ratings for minor leagues but not major leagues then they're playing two different games. The end user doesn't know that, and can't be expected to logically assume that, either. It's entirely counterintuitive.

[Paul Reuschel's Mustache]

Originally Posted by RonCo

No offense intended, but OOTP has everything they need in the thing called "current rating."

There is no value to a model that has potential as an input to current ratings, and there is a large degree of intellectual downside to implementing that design.

[Paul Reuschel's Mustache]

Originally Posted by RonCo

It's a beautiful, amazing game. But sometimes the best intentions go awry. This is a decision that needs total rethinking.

Yes.

[WhiskyTango]

Right. This game routinely has AI teams cutting top 10 prospects - franchise players - and yet it replicates real world scenarios. ho-kay.

[Zooerp]

15+ year OOTP player and former college baseball player, fully with Lukas, Matt, and LansdowneSt here.

I think it's a really interesting and positive change. The explanation of the way raw tools give players advantages at lower levels makes a lot of sense to me. I'm not too concerned about the output for scenarios that won't occur in the game, especially since I've had a much better personal experience this year with player development and managing my minor league system.

There's an interesting discussion to be had that Lukas and Matt are clearly open to engaging with. After all the hours I've put in over the years, I trust them.

[TheSpeakerfortheDead]

Originally Posted by RonCo

No offense intended, but OOTP has everything they need in the thing called "current rating." If you want to have players who race through the minors, then you have your AI drop great players into too low of a level--which is what real teams have effectively done forever. When, a month later, it becomes clear a player is destroying their first level, they bump them up.

There is no value to a model that has potential as an input to current ratings, and there is a large degree of intellectual downside to implementing that design.

I think this is pretty obviously correct for the vast majority of people playing this game.

Why are the ratings called "current ratings" not current ratings? If the guy has tools that make them better than their peers just don't spawn them in the game at 25/25/25/20/20.

[Mr. Marlin]

Originally Posted by low

I was waiting to hear your opinion about all this since I know you've been playing that way for years. Don't you think potential plays too big of a role, though, when actual stats are concerned?

To tell the truth I really don't know what potential any player in my system has. Other than what might be in the scouting report verbiage. When playing stats only you don't actually see any ratings. It's all about "what are you doing for me right now". Players that perform well in the minors tend to perform well in the majors eventually. Not all. Not even half actually but finding out which ones will without knowing their ratings is part of the appeal for me at least.

The real trick though is determining what is statistically "good". Low you've been playing OOTP a long time and you know as realistic as it strives to be it's still a video game. If you see a minor leaguer slashing 280/350/750 you'd think he's pretty good. BUT.... what is the average for the league he's playing in? He might be average or even below. The secret sauce IMO is comparing player performance vs league average. In real world baseball there are objectively good numbers. In OOTP good is only good if it's better than the average.

I suspect the players that post good numbers as they move up had high potential or "tools". Maybe they did but I wouldn't know. All I knew is they performed well. A few seasons back I did a sneak peek and found out my starting LF had below average ratings and only 1.5 stars. But he was a career .290 hitter and had a <1 ZR and a positive WAR.

JPeters sounds like he's onto something here and like Lukas said they may adjust it in the future. But when you stop worrying about potential ratings and just measure performance the game actually gets easier IMO. There is a great line in the Moneyball movie: "If he's such a good hitter why doesn't he hit good?" It's a lot like that.

Actually I think the things that drive OOTP players crazy is bias. When you play with real players, or historical players you have certain expectations to what they can or should be doing. But they are not real people to OOTP. Just numbers in a game engine. Or looking at ratings, especially potential ratings. Sometimes they go up and sometimes they go down. Does performance drive ratings or vice versa. After playing since OOTP8 I still don't know. But I do know that a player with a .600 OPS is objectively good when his league's average is .480 or something.

When you play with all fictional players (and teams) there is no bias and it doesn't break immersion when the Yankees trade Aaron Judge or something. Or Babe Ruth never breaks 20 HR per season. Or when you see a player with 80 contact hitting .190.

Anyway, just my $0.02

[Brad K]

Originally Posted by Mr. Marlin

To tell the truth I really don't know what potential any player in my system has. Other than what might be in the scouting report verbiage. When playing stats only you don't actually see any ratings. It's all about "what are you doing for me right now". Players that perform well in the minors tend to perform well in the majors eventually. Not all. Not even half actually but finding out which ones will without knowing their ratings is part of the appeal for me at least.

Not as much now.