Please stop demolishing the Potential of quality young players who've already fulfilled it

[voxpoptart]

My largest issue with OOTP16 -- an otherwise outstanding game, the best yet in the series -- is the same as my largest issue was with OOTP8 or OOTP4. It's also an issue I think I know how to fix:

A good young major leaguer in OOTP is far more likely to have his ratings collapse than a veteran is -- even a veteran in his early 30s.

This is two problems, by the way. (1) Healthy (or mostly-healthy) major leaguers in their early 20s should generally, to model real life, hold steady or improve. The younger they are, the more likely they should be to improve *vastly*; in real life, a list of {solid} 20-year-old MLB regulars is a list half consisting of Hall of Famers. {EDIT: This is literally true. Responding to Marc, I've shown the data, two posts below this one.} In OOTP, a 20-year-old in the majors is more likely to be reduced to a AA-ball player by Random Talent Changes.

But also (2) OOTP major leaguers from age 26 on are too *unlikely* to suddenly improve or decline. The game didn't make late-blooming Jose Bautistas and Randy Johnsons in the early versions; it still doesn't. Nor do I see a lot of Kevin McReynolds, Don Mattingly types who are genuinely superb but basically start free-falling at 30.

The solution, I think, is to abandon pure Random Talent Change in favor of Random Talent Change Triggered By Performance. The better a player is doing, especially at a level that's quite difficult for their age, the more likely (although unpredictable) it should be that their Potential takes a surge -- even if the player is late in their 20s, although the chances decline with age. The worse a player slumps, the more likely (although unpredictable) it should be that their Potential falls, and moreso as they age.

Potential that is purely hypothetical should fall more easily than Potential that a player has already converted into Current Rating. That, actually, can dramatically fix the game even in the Talent Change Randomness model: just have the game reject falls in Potential that tear apart a good young player.

But a performance-based model -- which would be just as random in the sense that performance variations are random -- would also make the later careers of players more varied, realistic, and interesting.

[marc5477]

This is two problems, by the way. (1) Healthy (or mostly-healthy) major leaguers in their early 20s should generally, to model real life, hold steady or improve. The younger they are, the more likely they should be to improve *vastly*; in real life, a list of 20-year-old MLB regulars is a list half consisting of Hall of Famers. In OOTP, a 20-year-old in the majors is more likely to be reduced to a AA-ball player by Random Talent Changes. (Yes, I know about B.J. "Melvin" Upton, who became a regular at 22, still pretty young. He's an extreme case, and even he's still a tolerable major-league bench player. I see worse youthful declines all over OOTP.)

This is not true. 95% of all rookies who have great years end up slumping late in the year and suck completely their sophomore year. This happens because the league learns their weaknesses and exploits them from then on. Most players never recover from that hump and only a few actually make it past year 4 in the majors. i would say 4% become average players, 0.95% go on to be above average stars, and 0.05% go on to become superstars.

No way in hell that half of 20 year olds go on to the HoF. Almost all of them fizzle out in real life but you dont realize it because their names and performance is never publicized and because no one collects baseball cards any more (else you would know that most dont get far at all). They just disappear and no one knows the better.

One should also take rookies of the roid era as a grain of salt. Everything from the late 90s to today is pretty much garbage due to cheating and no one had more access than rookies between 1995 to 2010 (but they were still playing against mostly honest players thus their stats are hugely inflated). The league needs to level the playing field. Either legalize all drugs or perma-ban all players who are caught without exception. I would also penalize their teams and take away any awards they won if they had a cheater on their team (to make teams be more proactive about the issue). Pick one and go with it.

I agree about random player development. It gets frustrating sometimes.

[voxpoptart]

Sorry -- I should have provided the data behind my claim, and now I will. "A list of major league 20-year-old regulars is a list half-consisting of Hall of Famers" wasn't a guess and or an exaggeration, it's the literal truth.

Actually it's an exaggeration if you count *bad* regulars, but the claim is literally true of 20-year-olds who are average-to-good regulars. There, names like Mel Ott, Johnny Bench, Al Kaline, Mickey Mantle, Ty Cobb, Frank Robinson, Ted Williams, Ken Griffey Jr, Rogers Hornsby, Jimmie Foxx, Mike Trout, Alex Rodriguez, and Roberto Alomar dominate the top twenty seasons at that age (plus too-soon-to-tell guys like Manny Machado, Bryce Harper, and Jason Heyward).

Willie Mays, Arky Vaughn, Joe Torre, Bill Mazeroski, Eddie Mathews, Fred Lindstrom, Ivan Rodriguez, and Travis Jackson are also in the top-44 list, which is the entire set of "average-and-above seasons by a 20-year-old regular". (As are possible future HoF-ers Alan Trammell and Giancarlo Stanton.) The only players in that entire top-44 list who didn't have good careers were Tony Conigliaro (destroyed by injuries) and Bob Horner.

Even among the mediocre 20-year-old regulars are a smattering of Bobby Doerrs, Hank Aarons, Robin Younts, and Miguel Cabreras, and very few careers that weren't pretty decent. In, again, real life -- and not in OOTP.

Here's a list: http://www.fangraphs.com/leaders.asp...ter=&players=0

[marc5477]

Originally Posted by voxpoptart

Sorry -- I should have provided the data behind my claim, and now I will. "A list of major league 20-year-old regulars is a list half-consisting of Hall of Famers" wasn't a guess and or an exaggeration, it's the literal truth.

Actually it's an exaggeration if you count *bad* regulars, but the claim is literally true of 20-year-olds who are average-to-good regulars. There, names like Mel Ott, Johnny Bench, Al Kaline, Mickey Mantle, Ty Cobb, Frank Robinson, Ted Williams, Ken Griffey Jr, Rogers Hornsby, Jimmie Foxx, Mike Trout, Alex Rodriguez, and Roberto Alomar dominate the top twenty seasons at that age (plus too-soon-to-tell guys like Manny Machado, Bryce Harper, and Jason Heyward).

Willie Mays, Arky Vaughn, Joe Torre, Bill Mazeroski, Eddie Mathews, Fred Lindstrom, Ivan Rodriguez, and Travis Jackson are also in the top-44 list, which is the entire set of "average-and-above seasons by a 20-year-old regular". (As are possible future HoF-ers Alan Trammell and Giancarlo Stanton.) The only players in that entire top-44 list who didn't have good careers were Tony Conigliaro (destroyed by injuries) and Bob Horner.

Even among the mediocre 20-year-old regulars are a smattering of Bobby Doerrs, Hank Aarons, Robin Younts, and Miguel Cabreras, and very few careers that weren't pretty decent. In, again, real life -- and not in OOTP.

Here's a list: Major League Leaderboards » 2015 » Batters » Dashboard | FanGraphs Baseball

Exactly 20 year old? I would need to look at my old baseball cards from middle school to confirm but you are probably correct. Not many 20 year old (or younger) starters in history.

[olivertheorem]

Vox, next you need to provide the counter-examples from OOTP to finish documenting your concern. I'm not sure how exactly you'd compile it, but you next have to show that 20-year-olds who meet the same benchmarks as in your fangraphs query fail to sustain their performance. Hard data, not anecdotes.

And it will help your case if you also post your player development modifiers.

[voxpoptart]

Marc: the data for 21-year-olds is less extreme but generally similar. Look at the link I gave, then change the age range from "20 to 20" to "21 to 21". A list of 22-year-old regulars still consists mostly of players who improved from there, often by large amounts, as you can also determine for yourself.

[voxpoptart]

Oliver: I'm out of town for two weeks, and my data is also corrupted by the fact that when I notice the unraveling happening, I take steps to undo it. So you will have to trust that I am not lying to you (why would I, exactly?) when I give you examples.

The best 20-year-old in my fictional league's opening season, worth 4-5 wins, was 2B Alberto Quetzalcoatl. On a scale of 20, his Con/Gap/Pow/Eye were something like 10-15-10-16, with room to grow. Late in season two I noticed he was in a long slump, checked his ratings, and they were more like 9-13-9-14, with no room to grow.

Second-best at that age was CF Himmi Nguvumali, worth 3-4 wins with ratings something like 10-11-12-11 and, again, room to grow. Once I looked at Quetzalcoatl, I looked at Nguvumali and saw, again, general decline (9-11-11-10?) and potential that had fallen equal to current. In one of the two cases, I found on fixing it - I forget which player - the potential had a lot of falling still to go. As in, prepping for a permanent trip to the minor leagues.

LF Hoshi Nagata was 21 in the first season and was good for two years, a 3-win player with ratings similar to Nguvumali. Third year, I realize halfway through that he is hitting .180 with nothing to compensate -- and his ratings translate, according to the editor, to .225 with nothing to compensate. His potential is again below his actual. In fixing his potential I cut his current ratings, because of his bad performance. He remains below .200 halfway into the league's fourth season.

Tom Collins, 21, was a starting CF in the league's first year; hit .300 (mostly singles), stole bases, played good defense. A year later he was hitting .250 in AA, and his potential was still headed downward.

That's four cases you'd have a hard time matching in 100 years of real baseball. None were caused by injury.

[voxpoptart]

I will add this: check your own leagues. Look in their history for young regulars. Find out what happened to them. I've had this happen in every OOTP version I've played, in very different fictional setups. I am surely not getting uniquely defective young hitters.

Oh, and I keep my development modifiers at default.

[olivertheorem]

Vox, I wasn't trying to imply that you were lying at all. Just that you need to document both sides if at all possible (and also possibly provide the league file) so that Markus & Co can figure out why OOTP isn't matching reality in a given manner.

[thehip41]

Look at JD Martinez.

He is the perfect example of a real life Random Talent change

[voxpoptart]

thehip41: No, JD Martinez is a perfect example of talent changes being indicated by surprisingly good performances. A 20th-round draft choice, he was then an all-star at three different minor league stops at age 21, 22, and 23, which together started to indicate him as a genuine prospect. Then at age 26 his major league statistics blossomed, and so did his perceived future value. JD Martinez is exactly what a Random Talent Change Triggered By Performance model would encourage -- as are cases like Jose Bautista or Chris Davis.

That said, while I'd love for OOTP to move away from purely Random Change, my *main* argument here was that young players who achieve their potential should rarely find their ratings going down, except in cases of injury -- and should, more frequently than now, see their potential increasing instead.

Oliver: Unfortunately, between the fact that I play through seasons slowly, and the fact that I normally fix the losses of potential when they happen, I am the worst player to document these.

Assuming I'm not the only person this happens to, however, Markus or you or anyone should be able to find regular instances in their own games where a successful young major leaguer's ratings and potential, for no obvious reason, just plummet until they're no longer useful. In fact, they should find that it happens to young players more often than it happens to players in their late 20s. And that's just not how real baseball careers work.

[olivertheorem]

The one issue with examples like JD Martinez is that, in reality, the player doesn't magically get better because he has a period of above-his-level performance. Rather, he raises his level (random talent change) and then performs to that new higher level.

Ability precedes performance.

[voxpoptart]

Originally Posted by olivertheorem

The one issue with examples like JD Martinez is that, in reality, the player doesn't magically get better because he has a period of above-his-level performance. Rather, he raises his level (random talent change) and then performs to that new higher level. Ability precedes performance.

Certainly! The problem is that the game's Random Talent Change model doesn't simulate that well. I mean, I guess possibly it might if you set Ratings to a very low (0-10%) share of the AI's calculations, but if you're anywhere close to the 30% default -- I have Ratings at 20% -- you get bizarre-shaped careers like that of my game's Bran Weiland, where a 20th-round draft pick fails to perform at Rookie League; is released; is picked up by another organization and immediately is installed at high-A ball; plays well for a month; is promoted to AA; plays well for a month; and suddenly is starring in the major leagues.

Meanwhile, a different player (and this is very common) can have a great year at class A, not get promoted, have another great year at class A, not get promoted, have another great year at class A, not get promoted ... because the game has pre-judged the player as not having the potential to do better. I'm not claiming that should *never* happen; it does happen in reality! But in real life, surprisingly good performance at class A (or any level) also has a solid chance of indicating that the player could become much better than anyone thought -- and even if I set Ratings at 0%, that wouldn't be nearly as true in OOTP.

Random Talent Change Triggered By Performance would not be the true method by which real life works, I agree. But it would produce results that look much more like real careers than OOTP currently does. And in a way that would be unpredictable over the course of a career, and lots of fun.

[Rain King]

Which is why I think the most realistic method would be to have no hard-coded potential ratings at all. You have the current ratings, a baseline development engine factoring things like body type and position and then outside factors that can affect how "normally" the player moves along that baseline (i.e. personality ratings, coaching, injuries and of course some randomness). You still have scouts, but now those guys are truly projecting based on weighing certain qualities more/less than the other scouts based on preference.

[SirMichaelJordan]

Originally Posted by voxpoptart

Meanwhile, a different player (and this is very common) can have a great year at class A, not get promoted, have another great year at class A, not get promoted, have another great year at class A, not get promoted ... because the game has pre-judged the player as not having the potential to do better. I'm not claiming that should *never* happen; it does happen in reality! But in real life, surprisingly good performance at class A (or any level) also has a solid chance of indicating that the player could become much better than anyone thought -- and even if I set Ratings at 0%, that wouldn't be nearly as true in OOTP.

Random Talent Change Triggered By Performance would not be the true method by which real life works, I agree. But it would produce results that look much more like real careers than OOTP currently does. And in a way that would be unpredictable over the course of a career, and lots of fun.

And that is the reason the rule 5 was created.

[voxpoptart]

Originally Posted by SirMichaelJordan

And that is the reason the rule 5 was created.

Except that in OOTP, the decision never to promote him will be correct, because no amount of good performance is going to change the game's hardwired Potential rating. That would still be true sometimes under Random Talent Change Triggered by Performance; there do in fact seem to exist players who maul class A and never can get the hang of AA, although it's hard to know for sure how much that's affected by sample size. But in OOTP, good performance is as likely to be rewarded by ratings going *down*. And that, in my opinion, just shouldn't (with rare exceptions) be part of the game.

And yes, I'm talking about batters, not pitchers.

RainKing's more radical argument to "have no hard-coded potential ratings at all" is interesting, and I'm dubious about it but not utterly opposed. However, mine is less radical and should be much easier to implement.

[Lukas Berger]

Originally Posted by voxpoptart

Except that in OOTP, the decision never to promote him will be correct, because no amount of good performance is going to change the game's hardwired Potential rating. That would still be true sometimes under Random Talent Change Triggered by Performance; there do in fact seem to exist players who maul class A and never can get the hang of AA, although it's hard to know for sure how much that's affected by sample size. But in OOTP, good performance is as likely to be rewarded by ratings going *down*. And that, in my opinion, just shouldn't (with rare exceptions) be part of the game.

And yes, I'm talking about batters, not pitchers.

RainKing's more radical argument to "have no hard-coded potential ratings at all" is interesting, and I'm dubious about it but not utterly opposed. However, mine is less radical and should be much easier to implement.

I don't meant to seem aggressive, hope that's not the case, but I think you're worrying too much about the potential rating here and also not taking advantage of options already provided by the game.

The potential ratings are a guide, not a guarantee. They don't actually matter all that much. They do matter I guess, but guys with high potentials will bust, and guys with low potentials will outperform expectations, just as in real life.

I've had guys that have been 1* or 2* potentials their whole career be absolute strong players for years if they start performing while young and then just don't stop. Their potential rating might not ever catch up to their actual performance, but they're still darn good players in spite of that.

Additionally, both ratings, though particularly the overall rating, will take performance more into account and will even fluctuate to an extent based on performance if you enable the option 'Overall rating based on AI evaluation, not pure ratings' and also make sure the ai evaluation is set to place a good deal of emphasis on stats. You'll end up with guys with overall ratings higher than their potentials, but that will just reflect their actual performance to a greater degree.

Doing that will very likely provide you with an experience much closer to what you're looking for.

[voxpoptart]

Lukasberger: this post is about too-frequently watching young, successful major league batters get worse and worse, sometimes falling out of the major leagues for good. This is a thing that *occasionally* happens in the real major leagues, but -- as I've documented at length earlier in this thread -- can be much rarer, depending *how* young the player is, than "successful young major leaguer turns into Hall of Fame level superstar".

This is about actual declines. (Remember, my star ratings are in fact 80% based on stats.) This is also about players in their late 20s whose ratings and performance rarely decline very far at all.

* Of the hitters who started my OOTP league in their late 20s, and were drafted in the first seven rounds, all are basically as good as when they started, four years later, with only two exceptions (CF Carmo Faquinha went from backup All-Star to part-time player, SS Glen Guerin has gone from very-good to average regular).

* Of the younger players drafted in the first seven rounds, I've already discussed the significant declines that started happening to 2B Alberto Quetzalcoatl, CF Himmi Nguvumali, LF Hoshi Nagata, and LF Tom Collins. There were also LFs Hyeon-uk Park and Nathan Gallagher (both of whom began at 23, and started large declines at respectively 25 and 24). And RF Mark Murray (All-Star in year one at 22, platoon mediocrity since).

Not even mentioning smaller cases. I'm more aware of, but much shyer about fixing, players on my team. My own 10th-round pick, RF Shawn Dempsey, was my cleanup hitter at 24; he was in AAA at 26, and I didn't even begin to rescue his ratings until after I'd traded him. I had a platoon outfielder, Jules Scallini, who was worth a WAR at 21; he was on the bench at AA by 23. I had a major-league utility player, Blair Delzoppo, who wasn't especially good at 22, but wasn't awful; his Contact had dropped to 6 (out of 20) by the end of the next year, and his results matched.

That didn't happen to my late-20s players. It didn't happen to anyone's. That's messed up.

[voxpoptart]

And I hate having to repeat this, but Lukasberger and anyone here: STOP CONDESCENDING, and START CHECKING YOUR OWN LEAGUES.

This has been an issue with every OOTP edition I've ever bought, in every game configuration I've tried. I can't, surely?, be the only person it's happening to. I suspect that -- because I don't sim dozens of seasons at a time and then make a vague attempt to catch up and learn the names of the stars -- I'm just noticing it more.

[SirMichaelJordan]

Originally Posted by voxpoptart

Except that in OOTP, the decision never to promote him will be correct, because no amount of good performance is going to change the game's hardwired Potential rating. That would still be true sometimes under Random Talent Change Triggered by Performance; there do in fact seem to exist players who maul class A and never can get the hang of AA, although it's hard to know for sure how much that's affected by sample size. But in OOTP, good performance is as likely to be rewarded by ratings going *down*. And that, in my opinion, just shouldn't (with rare exceptions) be part of the game.

And yes, I'm talking about batters, not pitchers.

RainKing's more radical argument to "have no hard-coded potential ratings at all" is interesting, and I'm dubious about it but not utterly opposed. However, mine is less radical and should be much easier to implement.

I never had a problem with players who are performing not getting promoted unless they were obviously block which is not a frequent case because fictional players pretty much play 2 or more positions at a high level (which is an issue)

I play with ratings at 25% and TCR @ 200 and with those extreme settings, I've never witness a guy suck his entire career then perform at a high level out of nowhere or guy who are outperforming everyone in the minors and still has a low potential rating. Guys I see that are selected in the Rule V are guys who are too good for the level they are at (usually A AA) but not quite ready for the next level as far as their individual ratings which the game also looks at.