Feedback Requested: Tweaked Settings for Aging Curves

[endgame]

Originally Posted by Doughnuts?

My statement was only in response to Endgame's suggestion that we all should be testing and providing feedback. I wouldn't presume to comment on your behalf.

My "suggestion" was simply that

Originally Posted by endgame

it wouldn't hurt anything for others to do their own examinations and render judgments based on personal experience

meaning that it's easy to advocate change or be critical of a process that "allowed the monster to live", rather than present inquiry about the solution; offer comparisons of how the problem has affected your world, league, and experience with the game.

[Lee]

Question for you Sev, when you tried using a negative number for aging modifiers how high, or low did you try?

[Seviien]

Lee,
I went to about -0.200. I didn't look too closely, but it didn't look materially different than when I used 0.000.

Dougnuts,
No offense intended... that was just my way of trying to be funny.

Jdawgsenior,
I think the reason the shift feels more pronounced is that the production of the older players is higher than it has been in the past. Hence, they are more noticable. I didn't look too extensively at how production varried with age. All I did was look at the career leaderbaords to make sure they weren't out of whack and examine players who won the MVP or the Cy Young to see when their careers dropped off. Things seemed reasonable, so I didn't think to do anything else.

Markmeister,
I don't think the aging modifiers would explain the varriance you saw in your 1920's stats.

Everyone:
Having played the game some more, the phenom problem is indeed there, and it's annoying. I don't know what can be done about it, though. The only thing that comes to mind is lowering the MLEs for the lowest level of the minors. I don't think that would do the trick, though. Of course, having phenoms isn't really that big a deal for me, so I'm content to live with it.

[Carplos]

What's the "phenom problem"?

[Seviien]

I'm seeing too many great players who are like 20 years old.

[fantom1979]

Originally Posted by Seviien

I'm seeing too many great players who are like 20 years old.

agreed, I am playing around with the settings myself using a international type universe (MLB,AAA,AA,A,and 3 international leagues).

I keep running into several of the same problems.

- Batters: Having a hard time finding a happy medium for career stats. Often in my tests I find batters playing too long or not long enough. One test league I ran had a guy starting his career at 19 and playing until he was 42, averaging 50 homeruns per season for his entire career. I really do not want to see players with 1100 career homeruns.

- Pitchers: They seem to blossom early with their best year usually within a couple of years of hitting the majors. Getting a pitcher to 250 wins with current day MLB pitching settings seems to be quite a challange.

- Salaries: I am trying to keep the talent in the States, and use the international leagues as a feeder system into the MLB. I want the stars of the international leagues to jump ship, while only the middle of the road players in the MLB should go overseas. Setting a very low salary cap in the international leagues and a high salary cap in the MLB seems to help this. Part of the problem is that players that sign with international leagues from the MLB seem to ignore the recommended salaries that I have set in the Game Setup.

- Draft: I want to keep the draft as few picks as possible to avoid the influx of young players into the MLB. The major problem here is that the international leagues horde reserve roster players (I have seen up to 60 on one team), leaving the lower level minors in the MLB with not enough players to field a team. I havent been able to figure out a way to limit how many players can be on a reserve roster.

[Doughnuts?]

Originally Posted by Seviien

I'm seeing too many great players who are like 20 years old.

Yes, this is a significant issue in my league that I have just gotten about 15 seasons into. And what this game does to pitchers is obscene. Every single talented, cy young quality starter has a significant drop in ratings by 27 or 28, even with the aging modifier at .001. I am tweaking like crazy to find a solution, it's driving me nuts.

[TNCubsFan]

I am 30 years simmed into a league using the setting on the first post. ML and full minors, along with coaches and hidden players. There is no salary cap or DH. The BABIP was set to recalculate each year.

There are about 30 or so players aged 34+ who are super studs. The rest are 20-26 and decent. The league ERA last year was 2.35... league ERA has declined each year. I am using the 2-8 system. A good hitter in this league is only develpoing to around a 5, and by good I mean around a .260 average for a career. The super studs are hitting 6 and one or two 7's in contact. Power is non existant except in a few isolated players who hit 50+ HR on a regular basis. The next level seems to be a few 30+ hitters, and after that it is an oppps down the line HR for the vast majority.

I am going to run another 30 year League using this same set up but with a cap and some changes to pitchers and hitters develpoment. I know that mine dosent match the starting set up of the thread but the more information we have the better it is for all of us.

[RonCo]

This has been studied and discussed heavily in the beta testers area. I'm sure Markus will address it in some fashion soon.

[No Pepper]

I could be inclined to believe the "phenom problem" is a way to keep average gamers interested in using the player development aspect of the game--and thus play the game longer-- because you get your phenoms "early and often". It would seem to me that if you have to lower the modifiers to 0.001 to see a few percentage point differences in the aging curve, something with default is not right, or better put, rather inflexible, without some other change required.

Has anyone adjusted GM/Manager preferences to see if "favoring veterans" adjusts these curves further? The analysis so far has been centered on who is playing games (AB and IP), not necessarily performance (though VORP was recently brought up). At the very least, it would be interesting to see if these settings worked too. Of course, because the model is complex, one is not guaranteed to see drastic changes if just one adjustment is made.

[TNCubsFan]

Originally Posted by RonCo

This has been studied and discussed heavily in the beta testers area. I'm sure Markus will address it in some fashion soon.

That is completely wonderful. *cue birds singing and fawns playing in the woods*

For the rest of us we want to work this problem out on the members boards. Finding the settings that "work" on the beta board does the rest of us no darn good. As it is I feel like a beta tester with this game whether I am on the beta boards or not.

[Seviien]

My settings work pretty well, and they're posted in this thread -- which is on the main board. There has been a lot of *discussion* in the beta forums, but that shouldn't be read by anyone as those on the beta board are hoarding a set of super-secret settings that makes everything perfect. From a gameplay perspective, there isn't anything there that hasn't already been included in this thread.

[arod23]

Thanks for your work on this Sev. I have only run a few test sims with your .001 aging suggestion (before your recent post), but Garlon and I still think their are issues with pitchers. Basically, batters stay forever, pitchers last 10 years or less, even good ones. This is with historical sims, not fictional. The AI "retires" pitchers far earlier than batters. Not sure of all that is driving it, but a combination of aging, the role assigned by the AI (which can vary), initial endurance rating, pitcher use, etc. In sim after sim, there are many batters with 20+ years of playing time, but very few pitchers. In cumulative stats, there are a disproportionate number of hitters with 500+ HRs compared to the few pitchers who reach 300 wins (or even 200 wins). We are working on making adjustments (artificial) to our historical DB as a work around to what seems like a major AI/program issue. For example, just adjusting the ratio of Games Started to Games Pitched can have a dramatic effect on endurance and role when a player is imported. Please let us know if you have suggestions for DB adjustments to help these aging issues.

[sfgiants88]

Originally Posted by RonCo

This has been studied and discussed heavily in the beta testers area. I'm sure Markus will address it in some fashion soon.

But with no firm plans to release another patch, will we have to wait until 2007 for the fix?

[Raidergoo]

Originally Posted by sfgiants88

But with no firm plans to release another patch, will we have to wait until 2007 for the fix?

At this very second, we're trying to document that certain elements need to be addressed in OOTP 2006. I would say this is a prime example of what could stand to be fixed.

These aging curve figures get the game engine a bit closer to observed Tangotiger data.

I just wonder if the Tangotiger data is flawed. Could it be skewed due to the advances of medical science over the last 2 decades? There's a reason it's called Tommy John surgery; anyone who got his problem suffered a CEI? Throw in contact lenses, better diets, and a million other little things, and maybe modern career arcs are less vicious than before.

[TNCubsFan]

Originally Posted by Seviien

My settings work pretty well, and they're posted in this thread -- which is on the main board. There has been a lot of *discussion* in the beta forums, but that shouldn't be read by anyone as those on the beta board are hoarding a set of super-secret settings that makes everything perfect. From a gameplay perspective, there isn't anything there that hasn't already been included in this thread.

My appologies to you Seviien. I was having a crappy 72 hours when I wrote what I wrote and none of it was directed at you. To be honest your thread probably saved me about 10, 20 year sims to get in the area where you are at the start of this thread.

Thank you for organising this and working throught it with the general public. I wish SI would kick your way 5% of the profit off the game since you are heading up a fix for a major problem with the engine that should have been fixed way before it was released.

[Zeyes]

Originally Posted by Doughnuts?

Yes, this is a significant issue in my league that I have just gotten about 15 seasons into. And what this game does to pitchers is obscene. Every single talented, cy young quality starter has a significant drop in ratings by 27 or 28, even with the aging modifier at .001. I am tweaking like crazy to find a solution, it's driving me nuts.

I've just tried a league with the developmental modifiers jacked up to 2.000 (aging left at the defaults) and I ended up with tons of talented players (both pitchers and hitters) falling off the cliff already around age 24/25...it's mostly anecdotal, but this seems to imply one of two things:

A) The developmental modifiers have an effect across the entire career, not just during the rise towards the player's max ratings, and later in the career their effect is merely counter-acted by the aging modifiers rather than replaced, OR

B) the aging modifier isn't tied to age at all, but rather to the timing of each player's achieving his maximum ratings. If that's the case, the main problem seems to be that most players simply don't plateau at/near their top level long enough (with the default aging setting), instead starting to fall off nearly as soon as their development cycle peaks.

I suspect it's the latter.

Edit: I should say that what's cliff-diving at age 24/25 are the players' potentials...some of them have their ratings and performance follow along right away, while others manage to hang on for another couple of years with ratings in excess of potentials, but even they're done completely by age 28 or so.

[Doughnuts?]

Originally Posted by arod23

Thanks for your work on this Sev. I have only run a few test sims with your .001 aging suggestion (before your recent post), but Garlon and I still think their are issues with pitchers. Basically, batters stay forever, pitchers last 10 years or less, even good ones. This is with historical sims, not fictional. The AI "retires" pitchers far earlier than batters. Not sure of all that is driving it, but a combination of aging, the role assigned by the AI (which can vary), initial endurance rating, pitcher use, etc. In sim after sim, there are many batters with 20+ years of playing time, but very few pitchers. In cumulative stats, there are a disproportionate number of hitters with 500+ HRs compared to the few pitchers who reach 300 wins (or even 200 wins). We are working on making adjustments (artificial) to our historical DB as a work around to what seems like a major AI/program issue. For example, just adjusting the ratio of Games Started to Games Pitched can have a dramatic effect on endurance and role when a player is imported. Please let us know if you have suggestions for DB adjustments to help these aging issues.

This is exactly what I am finding in my fictional universe as well. Batters producing at All-Star levels at 43 -44 and pitchers basically finished by 30.

[Charlie Hough]

Originally Posted by arod23

Basically, batters stay forever, pitchers last 10 years or less, even good ones. This is with historical sims, not fictional. The AI "retires" pitchers far earlier than batters. Not sure of all that is driving it, but a combination of aging, the role assigned by the AI (which can vary), initial endurance rating, pitcher use, etc. In sim after sim, there are many batters with 20+ years of playing time, but very few pitchers.

Based on my relatively few experiments, it seems to me that the entire problem is either randomization or the game retiring players based on length of service. I have modified pitcher aging dramatically, to slow down the process of ratings deterioration and perhaps keep pitchers active and successful into their late 30's and early 40's, but many are still retiring for no apparent reason.

Pitchers will come off a winning season with a good contract, no injuries, and ratings that have not dropped off at all, but they will suddenly retire. In these case, the pitchers are usually in their mid to late 30's, which is why I tend to think that it's either random or based on years of service. If they've been in the major leagues for 15 years or more, then maybe the game makes them much more likely to retire, despite their success and strong ability.

While there should be a chance that a player will retire at the top of his game, it should be quite low. In real life, it is very rare for a player to retire when he is still effective, is showing no signs of decline, and can still make a lot of money. But the game does not seem to recognize this at times. So either it's randomizing retirement or basing it on years in the majors.

The other problem I find is that players give up on their careers too easily. I've seen pitchers at age 30 decide to retire just a month after becoming free agents. In several cases, they spent the last season on the reserve roster and did not appear in a major league game. But they should not retire so quickly and easily, especially when their ratings have remained unchanged and they now have the freedom to sign elsewhere. But it happens after every season in my historical sims. A player just needs about one year out of the majors, and he's ready to retire, even though he might sign on with a new club before the next season begins.

[akw4572]

So, IS there a setting that will curb the problem of pitchers nosediving/retiring in their early to mid thirties???