caught stealing real-life/import stats are off?

[thehef]

I was doing some historical testing (started with 1938 and am now in mid-40's) and I happened to notice that real life caught-stealing totals were off for some players:

		ACTUAL		OOTP "REAL LIFE"
PLAYER		SB/CS		SB/CS
Lee Handley	68/54		68/13
Pete Reiser	87/39		87/14
Bobby Adams	67/43		67/35
Pee Wee Reese	232/99		232/64
George Myatt	72/29		72/23
Bert Haas	51/33		51/9

I'll note that for many players, their SB/CS totals - RL vs OOTP - were accurate. I'll also note that for the examples listed above, all of their other stats matched (RL vs OOTP).

Anybody seen anything like this before? I'm pretty sure it's not just a cosmetic thing. Reiser, Adams & Haas were my National League SB leaders for 1946, with 61, 53, and 35 steals, when IRL they swiped 34, 16, and 22.

[joefromchicago]

MLB didn't officially start recording CS until 1951. Before then, I think it was occasionally reported, but not on a consistent basis. The pre-1951 CS stats that you see in BBRef are probably the result of work done by researchers going through old boxscores. That's an ongoing process, so it's possible that OOTP's stats haven't been keeping pace with the research.

As with other stats that weren't recorded, I think OOTP estimates them for purposes of creating ratings. As such, the incomplete CS stats might not be as big of a problem as it might appear.

[thehef]

Originally Posted by joefromchicago

As with other stats that weren't recorded, I think OOTP estimates them for purposes of creating ratings.

Thanks for your reply

I think you are suggesting that when OOTP grabbed Reiser's stats from Lahman (?), that Reiser's CS were 14, and that perhaps that number has since been updated due to work by researchers. Correct?

And that, regardless, OOTP would ignore pre-1951 CS stats and instead use estimates of its own?

Originally Posted by joefromchicago

As such, the incomplete CS stats might not be as big of a problem as it might appear.

I'm thinking that my 1946 NL stolen base leaderboard...

Reiser 61
Adams 53
Haas 37

... does not jibe very well with what these three guys actually did in 1946...

Reiser 34
Adams 16
Haas 22

Granted, each of these guys did play a bit more in OOTP than they did in real life. So that partially explains their higher-than-real-life SB totals. But I'm thinking their high (imported, not necessarily real) success rates factored into them attempting more steals, also.

(FYI, my sim is with 1-year recalc, no real lineups, pretty much all defaults beyond that...)

Of course, we already know that OOTP has an issue with top base stealers getting caught A LOT more than in real life. In the case of these guys, since - based on their imported stats - their high imported-but-not-real success rates (probably) caused them to steal more, OOTP still punished them by having them get caught stealing not just more than their imported-but-not-real success rates, but more than their actual success rates.

[Brad K]

Originally Posted by thehef

Of course, we already know that OOTP has an issue with top base stealers getting caught A LOT more than in real life. In the case of these guys, since - based on their imported stats - their high imported-but-not-real success rates (probably) caused them to steal more, OOTP still punished them by having them get caught stealing not just more than their imported-but-not-real success rates, but more than their actual success rates.

I noticed the issue in 21 and reported it. Hope for a fix in 23.

https://forums.ootpdevelopments.com/...light=stealing

[thehef]

Originally Posted by Brad K

I noticed the issue in 21 and reported it. Hope for a fix in 23.

Absolutely! I can almost live with a situation where Davey Lopes steals 68 bases and gets caught 28 times, when IRL he swiped 63 and was only caught 10 times. But the immersion buster is when Willie Davis goes 42/13 IRL, but in OOPT goes 35/38.

My understanding is that there's going to be some effort to fix this in v 23. Let's hope!

[joefromchicago]

Originally Posted by thehef

Thanks for your reply

I think you are suggesting that when OOTP grabbed Reiser's stats from Lahman (?), that Reiser's CS were 14, and that perhaps that number has since been updated due to work by researchers. Correct?


And that, regardless, OOTP would ignore pre-1951 CS stats and instead use estimates of its own?

Correct

Originally Posted by thehef

I'm thinking that my 1946 NL stolen base leaderboard...

Reiser 61
Adams 53
Haas 37

... does not jibe very well with what these three guys actually did in 1946...

Reiser 34
Adams 16
Haas 22

Granted, each of these guys did play a bit more in OOTP than they did in real life. So that partially explains their higher-than-real-life SB totals. But I'm thinking their high (imported, not necessarily real) success rates factored into them attempting more steals, also.

Stolen bases are so situational that it's difficult to tell what's going on when the OOTP stats don't match up to real life. I agree that those guys shouldn't have that many steals, but then they also shouldn't have that many attempts. So they're stealing more bases, but they're also running way more than they should. That might be a problem with the individual ratings or it might be a problem with the strategy settings or it might be a problem with the AI or it might just be a statistical burp. Hard to say.

I'll just note that all but one of the guys that you mentioned were National Leaguers and, as I mentioned previously, the NL didn't keep CS stats before 1951. Evidently, the AL kept CS stats starting in 1920, so maybe this really is an estimated-stats problem. Perhaps the NL players get a boost because they look a lot faster than their AL counterparts who had all of their CS recorded. That might be worth a look. What, for instance, did George Case do in your league?

If this is a known issue, then maybe the best we can hope for is that the development team will fix it eventually.

[thehef]

Originally Posted by joefromchicago

What, for instance, did George Case do in your league?

First, Case imports with the correct SB/CS stats: 349/109. His issue is the other one referenced above - the immersion-buster where the top base stealers get caught waaaaaay too often. In my sim, he went 330/297.

[David Watts]

Originally Posted by thehef

First, Case imports with the correct SB/CS stats: 349/109. His issue is the other one referenced above - the immersion-buster where the top base stealers get caught waaaaaay too often. In my sim, he went 330/297.

Something I've always planned on paying more attention to, but never end up doing so, is whether or not OOTP managers are having guys attempt to steal 3rd base way too often?

[thehef]

Originally Posted by David Watts

Something I've always planned on paying more attention to, but never end up doing so, is whether or not OOTP managers are having guys attempt to steal 3rd base way too often?

Interesting... One thing that's been determined is that the top base stealers get caught way more often than they should because a bunch of low-attempt guys end up with a better steal % than they should. For example, while Steve Garvey had 5 SB and 4 CS in 1974, in OOTP he might go for 7 SB and 3 CS. That's not noteworthy to the gamer because 5/4 and 7/3 both seem like reasonable random "rolls of the dice." But when you multiply that scenario over dozens, maybe hundreds, of players it - for lack of a better term & understanding - clashes with League Totals, and therefore the high-rate guys get punished in order for OOTP stat totals to jibe with League Totals.

My understanding is that the developers think this may have to do with real-life resulting in more busted hit-and-run plays - for the low-attempt guys - where the runner gets caught, than OOTP produces... Not really sure on this, though. Just what I thought I saw...

[fredbeene]

Originally Posted by thehef

Interesting... One thing that's been determined is that the top base stealers get caught way more often than they should because a bunch of low-attempt guys end up with a better steal % than they should. For example, while Steve Garvey had 5 SB and 4 CS in 1974, in OOTP he might go for 7 SB and 3 CS. That's not noteworthy to the gamer because 5/4 and 7/3 both seem like reasonable random "rolls of the dice." But when you multiply that scenario over dozens, maybe hundreds, of players it - for lack of a better term & understanding - clashes with League Totals, and therefore the high-rate guys get punished in order for OOTP stat totals to jibe with League Totals.

My understanding is that the developers think this may have to do with real-life resulting in more busted hit-and-run plays - for the low-attempt guys - where the runner gets caught, than OOTP produces... Not really sure on this, though. Just what I thought I saw...

Is there a way to turn off the OOTPB stat total to align with the League Totals? Doesn't it take away the fun of it?
Things don't seem random when it skews all results to hit a predeterimed outcome