Surprising OOTP5 numbers...

[Ben E Lou]

There's been much debate about player development curves and the age and speed at which OOTP5 players decline. It though it would be worth posting some of my findings here.

OK. I just took a look at some of the top hitters in the history of the game. I looked at batters with over 3,500 hits and/or over 500 homers who are retired. I looked at their careers and (subjectively) determined what was the first year that they were "in decline," meaning after that year they never put up numbers that were at or near "star-quality." (As I said, I know this is somewhat subjective, but in most cases the year was pretty obvious...) Here's what I found:

AVERAGE AGE OF FIRST "SEASON OF DECLINE": 37.65
AVERAGE AGE OF RETIREMENT: 39.95

Keep in mind, these are the EXTREMELY ELITE of baseball history. (There were exactly 20 guys in over 100 years of baseball history who fit the criteria I used...very convenient number.) It seems that it would follow naturally that these guys have better physical skills and are able to perform at a major league level longer than the thousands of below-average-to-above-average players. So, we should therefore see the "average OOTP superstar" have his first "bad" season at age 37 or 38, and be out of the game at 40. Take that for what it is worth.

--Ben

I just ran the same experiment with my 50ish season fictional league, adn actually things look a lot better than I would have thought. I realized that since the league was about 1/2 of the lenght of time of real baseball, there should be 10 "elite" players to match the 20 "elite" players of the 100+ years of real baseball. Here are the numbers I found, again for the elite battters:

AVERAGE AGE AT FIRST "DECLINE" SEASON IN REAL LIFE: 37.65
AVERAGE AGE IN OOTP 5.01: 36.8

AVERAGE AGE OF RETIREMENT IN REAL LIFE: 39.95
AVERAGE AGE IN OOTP 5.01: 39.1

Both are about .8 years below the "real" numbers, but the spreads between "decline year" and retirement are identical: 2.3 years.

[Ksyrup]

I'm sure Marcus is thinking, "Where were you with these numbers before I spent all that time on 5.10 trying to correct a problem that might not exist?"

Unlike the SB problem....

[Ben E Lou]

Originally posted by Ksyrup
I'm sure Marcus is thinking, "Where were you with these numbers before I spent all that time on 5.10 trying to correct a problem that might not exist?"

LOL!

[Ben E Lou]

Incidentally, I think the average may have been right with 5.01, but the distribution may have been a problem. Of the ten guys I looked at in the OOTP5 universe, I seem to remember that two really skewed the numbers because they dropped off and played a whole bunch less-than-star-quality-but-productive seasons, while the other 8 quit either the same year as the drop-off year, or the very next season. I'm running a 100-season-with-5.10 sim right now, and I'll post some numbers about that sucker when it is done.

[Ben E Lou]

Originally posted by Ksyrup
I'm sure Marcus is thinking, "Where were you with these numbers before I spent all that time on 5.10 trying to correct a problem that might not exist?"

Unlike the SB problem....

Regarding the problem not existing, I think a big part of the problem is that people remember the exceptions such as Rickey Henderson and Dave Winfield, rather than the scores of guys who don't make it to age 40.

[Anrhydeddu]

It was amazing to see all of nashing of false teeths early on about good players dropping off at 32 yrs.

Ksyrup, I was going to mention what I think the cause of the SB problem - at least for those using Lahman: it does not interpret (or import) SB Ability correctly. The speed rating is accurate from what I have seen and just a simple edit (bump up) of the SB Ability to C or B can get you the 70-90 SB (assuming A speed).

[Ksyrup]

I'm having the same problem with fictional players. The most SBs in any one year in 85 years in my current sim is 65, and there have only been handful to reach 60. With OOTP4, I used to get 70-90 from the top base stealer fairly regularly (not always), and 90+ every once in awhile.

[HornedFrog Purple]

I have had decent success with fictional stolen bases. In 31 years, I have had a leader with 87. The highest stolen base total in the history is 748, which minus the freak that is Rickey Henderson is within reason.

I believe that some of the .8 is the year before the "dropoff year". In a lot of cases with the CPU-run teams, if a vet refuses to go to the minors and is cut, no CPU team will pick him up. The next season he either retires or is picked up as a free agent then his ratings go down the tubes.

The crash that happens that year though is just brutal. It's like they lose any sense of how to play ball overnight. I think it's too drastic in that regard.

[Ksyrup]

Can you give me some specifics on the era settings and such that you use? I can't get anyone to steal bases in OOTP5 - at least not to the levels they should be. The most SBs for a career that I have seen is 968, but that was a player I created, and he never had more than 61 in any season.

[Ben E Lou]

OK.....I just ran a 100-season sim and posted the results over at .400. FWIW, my career SB leader had 796 SB's, and that's with me not touching any settings, and playing all 100 years in "Modern" mode.

Alright boyz. My 100-season 5.10 sim finished during the night. I've looked around a little, and ran the numbers on 20 of the greatest in that universe. A few comments before I give the hard numbers....

1. Definitely there's a more varied development curve. There are guys who drop off the table, (Example: Abram Feemster which definitely happens to some superstars in real life. (Now that I've done a little research, I can show you that some of the game's elite had OOTP5-like declines, and trust me when I say that this was NOT my intuition on the matter at all. Ted Williams is really the only one who really "retired in his prime" numbers-wise. The rest quit because they were done.) There are also guys who slowly decline over several years. Example:Michael "Bleeder" Arline

2. I actually had difficulty determing a "year of decline" for several guys this time, because a handful had "spike seasons" in their late 30's. (In other words, a 40-45 homer guy might have had 2 or 3 seasons in their mid 30's of 25-30 homers, but then hit 42 at age 37 or something like that...) Example: Gil "The Bull" Martinez

3. While the overall averages got even closer to reality, something still seems amiss. I'll explain in my next point.

4. It seems to me that the stars don't know when to quit. Too many 38-42 year old former superstars are playing until they get released, rather than hangin' 'em up after a subpar season or two.

5. Sixfour may have a bit of a point about youngsters having solid seasons. While it is VERY rare for a 19-21-year-old to have a solid year, it does happen once in a blue moon. (Aaron, Mantle, Gooden, Williams, Ruth). Looking through the career leader board, I'm not seeing anyone have a solid season before age 22 or 23. RETRACTION!!! I didn't look hard enough. Check out Fred "Snake Eyes" Acuna. He had a very nice season at age 19. If there is one, I'll guarantee that there are others...

OK. Here are the numbers, for the stars:

REAL LIFE AVERAGE AGE OF FIRST YEAR OF DECLINE: 37.65
OOTP5.01 AVERAGE AGE OF FIRST YEAR OF DECLINE: 36.80
OOTP5.10 AVERAGE AGE OF FIRST YEAR OF DECLINE: 37.45

REAL LIFE AVERAGE AGE AT RETIREMENT: 39.95
OOTP5.01 AVERAGE AGE AT RETIREMENT: 39.10
OOTP5.10 AVERAGE AGE AT RETIREMENT: 39.65

REAL LIFE # OF YEARS OF DECLINE BEFORE RETIREMENT: 2.30
OOTP5.01 # OF YEARS OF DECLINE BEFORE RETIREMENT: 2.30
OOTP5.10 # OF YEARS OF DECLINE BEFORE RETIREMENT: 2.20

The average numbers with 5.10 are obviously INCREDIBLY close to the way things have happened in baseball history. I'm uploading this universe to a web site if anyone wants to take a look. There are a few freaks of nature on the leader board worth looking at. This universe can be found at: www.younglifenorthdekalb.com/dynasty

[Ksyrup]

Very nice analysis.

On the SB issue, your results are similar to mine. Your top SB numbers in either league are 64 and 65. My problem is that my numbers were attained using various eras, not including the modern era. I don't have as much problem with your results as mine. The game needs to factor in SBs as an era setting.

Another issue I'm starting to see is the "bad players" syndrome. I didn't see a place to look at your retired players list, but I've got a number of guys - and I'm focusing on pitchers here - who would never have stayed in the majors as long as they did, given their numbers. I have those teams from time to time that go 10-15 years without winning much, and sucking horribly (especially expansion teams), and some of the guys on those teams played for 5-6 years and complied outrageously bad numbers, like 25-116, 7.00 ERA, etc. I'm not sure that anything can really be done about it, as I assume the AI would promote players with better ratings if they had them, but some of the bottom-tier players are embarrassingly awful and would have, in real life, been sent packing after a couple of really bad years. It's not that big of a deal, but looking through my league's history last night, I was struck by how long the careers of the worst players were.

[Ben E Lou]

Originally posted by Ksyrup
My problem is that my numbers were attained using various eras, not including the modern era. I don't have as much problem with your results as mine. The game needs to factor in SBs as an era setting.

Good point.

[JonInMiddleGA]

Originally posted by Ksyrup ... but I've got a number of guys - and I'm focusing on pitchers here - who would never have stayed in the majors as long as they did, given their numbers. ... and some of the guys on those teams played for 5-6 years and complied outrageously bad numbers, like 25-116, 7.00 ERA, etc.

While I haven't spent all that much time with OOTP5, I think I've got a pretty good hunch about at least one of the things that causes these freaks -- I've never seen the AI convert a MR to SP, even if he has an endurance rating for the role.

So, at least in my experience, you don't have to look too hard to find teams running SP's with numbers in the 2-4 range out there every fifth day while they've got one or more MR's sitting in AAA with 6's-7's. The result is a group of SP's who shouldn't have seen a major league game without buying a ticket simply by virtue of their endurance rating.

[Anrhydeddu]

I can attest to very young superstars. On my 1964 Cardinals, I had 19 yr old Tony Conigliaro with a star-like season (.267-28-62-.841 OPS). The rest of my 19 yr olds with good ratings are way down in the minors. Also, I have a 17 yr old pitcher tearing up AAA.

Ksyrup: Lou Brock on my team had 94 official at-bats (plus some walks) and stoled 23 bases. I would hesitate to think how that extrapolates to a full season.

[Ksyrup]

Jon:

That could be. I hope that is something that gets fixed. It appears to have been fixed from OOTP4 on the hitters side. I created two players and put them into my universe. One was a 3B/LF (created as 3B primary) and the other a 1B/RF (1B primary). Both of these guys ended up switching back and forth between primary positions for their first couple of years, before they settled in at 3B and RF, respectively. I was impressed.

[Ksyrup]

You guys who are seeing high SB totals, can you tell me what era settings you are using? I'm wondering if I should cut down the bunt and H&R settings to encourage fast guys to steal instead of get bunted or H&R over to 2nd/3rd. Of course, if I do that, then I might end up with extraordinarily high double play numbers.

[Anrhydeddu]

I always use the default era settings. For Brock, I did set his personal strategy to fairly high in stealing though.

[Ksyrup]

Well that makes sense, then. I guess if Marcus would ratchet up the AI's use of the individual tealing tendencies for A-rated base stealers, I would start to see the kinds of numbers I want. I just want the ability to generate realistic numbers if I decide to devalue the HR in a particular league. Right now, if I do that, I've got to jack up the doubles and triples, or the league quickly becomes pitching-dominated.

[Havok]

dam... nice job Skydog

we should start calling you DataKing!!!

[korme]

Originally posted by HornedFrog Purple
The crash that happens that year though is just brutal. It's like they lose any sense of how to play ball overnight. I think it's too drastic in that regard.

Agreed, that is my only gripe. Players never seem to gradually decline. It's always a huge dropoff.. all-star at the beginning of the year, to AAA scrub at the end.. that NEVER will happen.

[IMetTrentGreen]

skydog- when you were simming, did you have to sit there and click no to the "import real rookies?" qeury every time? i was wondering if its possible to turn that off

[Ben E Lou]

Originally posted by IMetTrentGreen
skydog- when you were simming, did you have to sit there and click no to the "import real rookies?" qeury every time? i was wondering if its possible to turn that off

They've answered that question over at .400. You can only get rid of it by simming forward. They say there's no way to avoid it if you sim in the past.

I didn't have to in this sim. I started in 2003 and went forward. For the 50-year history in my dynasty thread career, I had to click it every time. I handled that by simming it on my laptop while at work on my desktop.

[Ksyrup]

I haven't simmed "into the future" since the game was first released, but that should only be an issue for sim years prior to 2003. Unless something has changed in the last couple of patches/updates.

[Ben E Lou]

Originally posted by Ksyrup
I haven't simmed "into the future" since the game was first released, but that should only be an issue for sim years prior to 2003. Unless something has changed in the last couple of patches/updates.

Beat ya to it!

[Ksyrup]

Originally posted by Shorty3281
Agreed, that is my only gripe. Players never seem to gradually decline. It's always a huge dropoff.. all-star at the beginning of the year, to AAA scrub at the end.. that NEVER will happen.

There was a nice thread about this issue a month or so ago at .400. Basically, I think we came to the conclusion that the real problem here is the fact that the computer - and humans - can see ratings. And because we can see ratings, and the AI uses the ratings to determine when a player is no longer worth keeping, the AI is able to anticipate a player's decline before you would ordinarily "see" it in the player's performance.

I think I used Mark McGwire as an example. Despite his injury-plagued 2000 season, he still hit .305 with 32 HRs in just 236 ABs. The next season, he struggled to a .187 average in 299 ABs. In OOTP5, you would actually "see" his ratings decline between 2000 and 2001, and the AI would likely not even put him on the field to show that he could no longer perform. In real life, even if McGwire didn't look physically able to play to start 2001, no one in their right mind would have suggested sending him to AAA or releasing him.

That's really what's missing from the game. I have no problem with guys "falling off a cliff," but they don't get the opportunity to demonstrate how bad they are before they are demoted or cut. Thus, you have guys going from all-star or solid seasons to the free agent dung heap in the span of an off-season. I'm not really sure what the solution is, unless the game was changed to lessen the impact of ratings as the be-all, end-all of determining playing time. That would be great, but I suspect difficult to get right.

[Ben E Lou]

The more I think about it, the more I think it should happen more in the off-season, and no be so visible until Spring Training reports come out.

On a relate note, check this out:

OK. It is time for some of y'all to read the same hurt-your-image-of-the-legends-of-the-game-stats that I've looked at the last couple of days. I was surprised to find out that things averaged out to closely match OOTP5 to reality. I was REALLY surprised to find out that most of these top-tier guys had numbers in their final season that would indicate that their talent had severely declined. Nearly half of them weren't even major-league-caliber any more when they retired. I don't know all of their history, but it sure looks like a whole bunch of them went out there to start the reason, and either they or their manager realized that not only were they no longer stars, but they weren't even major leaguers any more. Here are the most extreme four of the 20 guys I looked at:

Babe Ruth
1934 (age 39) .288-22-84
1935 (age 40) .181-6-12

Mike Schmidt
1987 (age 37): .293-35-113
1988 (age 38): .249-12-62

Mel Ott
1945 (age 36): .308-21-79
1946 (age 37): .074-1-4

Eddie Murray
1996 (age 40): .260-22-79
1997 (age 41): .222-3-18

Yes, these are fairly extreme drops, but look at the final seasons for the rest of them:

Hank Aaron: 271AB, .229-10-35
Willie Mays: 209AB,.211-6-25
Frank Robinson: 67AB, .224-3-10
Mark McGwire: 299AB,.187-29-64
Harmon Killebrew: 312 AB,.199-14-44
Reggie Jackson: 336AB,.220-15-43
Mickey Mantle: 435AB,.237-18-54
Jimmy Foxx: 224AB,.268-7-38
Willie McCovey: 113AB,.204-1-16
Ted Williams: 310AB,.316-29-72
Ernie Banks: 83AB,.193-3-6
Eddie Mathews: 52AB,.212-3-8
Pete Rose: 237AB,.219-0-25
Ty Cobb: 353AB,.323-1-40
Stan Musial: 337AB,.255-12-58
Tris Speaker: 191AB,.267-3-30

Granted, most of these guys dropped a little more slowly than OOTP5 guys, but clearly those four at the top (25% of the small sample) look as bad as anything I've seen in OOTP5. However, the bottom line is this: the great majority of the best hitters the game has ever known did not retire on their own terms. Cobb and Williams are the only ones of the 20 that look like they bowed out with graceful on-field performances.

[HornedFrog Purple]

I guess what I would suggest is a gradual reduction to futility. What I have noticed happening is an aging player takes drastic hits all during their "final" season. If they take a lesser blast at spring training which makes them a serviceable player at least, then kill them off at the end of the season I think it would help some.

What the AI does is when they see the first drastic sign right after spring training.... bam they are released. I have picked up a couple of guys that had pretty good last seasons that were just sitting on the free agent list. One of them even stuck around for another season after the AI teams had basically given up on them.

[primelord]

Looking at SkyDog's career it doesn't look like SB's are THAT far off. In his career 25 guys have over 500 SB's. In real life 35 guys have over 500 career stolen bases. It seem that we are missing some of the top guys I geuss, but overall it seems close.

[Anrhydeddu]

One of them even stuck around for another season after the AI teams had basically given up on them.

You brought something fundamental, I believe. The AI can only be realistically program to act the same for each of the team (or opponents). Despite hype to the contrary, I have not see a single AI that act very different from another AI (esp. in the much-vaulted Civ3). Therefore, if an AI gives up on a player too soon (we see this quite a bit in real life sports), all of the AI teams feel the same way - except the human team which can always evaluate talent better and pay accordingly.

[HornedFrog Purple]

Agreed. Maybe in future versions as the "owner personality" aspect is improved on, AI owners/GM's can have that tweakable such as "willing to evaluate older talent", "willing to rebuild with youth" etc etc etc

[Malificent]

Has anyone tested these numbers using an imported v4 league?

[FBPro]

Any league that you import will then develop on the same basis as one created in v5.

[oykib]

Originally posted by primelord
Looking at SkyDog's career it doesn't look like SB's are THAT far off. In his career 25 guys have over 500 SB's. In real life 35 guys have over 500 career stolen bases. It seem that we are missing some of the top guys I geuss, but overall it seems close.

The problem is, though, that no single player puts up incredible stolen base numbers. That means that you don't get any players whose primary value is speed being particularly valuable.

You can't get a Vince Coleman or Maury Wills, a guy that's all speed and defense. An ( Avg/Fair/Fair ) with a/a steal/speed, and a 'b' range at his position should be a borderline all-star if he's utilized properly.

The best you can hope for right now is Roger Cedeno, Otis Nixon, or Tom Goodwin. That speed can still turn an otherwise awful player into a serviceable player. But it won't push him past that until he gets over 50 SBs. He has to get to about 70 before he becomes genuinely good.

[FBPro]

In an online league that I play in we've had a couple of players go over 100+ sb for a season already and we are only in year 12 or so of the league so high numbers are possible.

[Ksyrup]

I think there are two possible solutions:

1) Tweak the AI so that it adjusts the individual settings for players A-rated in SB.

and/or

2) Make SB an era setting.

I'd prefer #1 over #2, as I don't necessarily want all players to attempt more steals just because, but I do want the best base stealers to take advantage of that skill more than they are currently.

[SFL Cat]

Very nice analysis.

On the SB issue, your results are similar to mine. Your top SB numbers in either league are 64 and 65. My problem is that my numbers were attained using various eras, not including the modern era. I don't have as much problem with your results as mine. The game needs to factor in SBs as an era setting.

Another issue I'm starting to see is the "bad players" syndrome. I didn't see a place to look at your retired players list, but I've got a number of guys - and I'm focusing on pitchers here - who would never have stayed in the majors as long as they did, given their numbers. I have those teams from time to time that go 10-15 years without winning much, and sucking horribly (especially expansion teams), and some of the guys on those teams played for 5-6 years and complied outrageously bad numbers, like 25-116, 7.00 ERA, etc. I'm not sure that anything can really be done about it, as I assume the AI would promote players with better ratings if they had them, but some of the bottom-tier players are embarrassingly awful and would have, in real life, been sent packing after a couple of really bad years. It's not that big of a deal, but looking through my league's history last night, I was struck by how long the careers of the worst players were.

It seems to me that a lot of managers on crappy teams may hold on to not so stellar players just because that player may have a quality the manager likes, might be a manager's project that never pans out, and so on. It seems like a majority of the time you see sweeping player changes on rosters is when a new manager (especially one who has a different philosophy on how to play the game) takes over and rebuilds the team to fit his vision. There are a lot of consistently mediocre teams that keep consistently mediocre players simply because they is no one better down on the farm to bring in to replace them.

[Ksyrup]

I'm simming a career right now that I started in 1896 and am up to 1994.

Here are some of the dishonorable mentions from my league:

Matthew Spencer (1901-1907) 13-104, 8.24 ERA
Sixto Ortiz (1958-1963) 15-87, 8.19 ERA
Jeffrey Sanroman (1949-1961) 52-201, 7.81 ERA
John Larmon (1952-1964) 61-161, 6.35 ERA

There are a bunch more, but these were some of the worst.

I'm not so concerned about them at the beginning of their careers, because of the way I handled expansion - just signed all free agents to the expansion teams and released all minor leaguers age 30 or over and assigned them to the expansion teams (although not all of these guys played with expansion teams). It's the fact that these guys lasted so long that is a bit unrealistic.

[Malificent]

FBPro, are you absolutely sure about the players already in existence from OOTP4 developing the same way as players created directly in OOTP5?