Wow, look at the stolen bases/caught stealing

[David Watts]

I thought this was reworked this version. This may be worse than ever.

[Syd Thrift]

1. One guy? You're basing this on 28 steal attempts from one guy?

2. This is definitely not what I've seen with fictional players. Sometimes it feels like elite baserunners in my league now have a free path to second when they feel like stealing (i.e. when the game's close, and when the guy on the mound isn't particularly great at holding runners).

[David Watts]

Originally Posted by Syd Thrift

1. One guy? You're basing this on 28 steal attempts from one guy?

2. This is definitely not what I've seen with fictional players. Sometimes it feels like elite baserunners in my league now have a free path to second when they feel like stealing (i.e. when the game's close, and when the guy on the mound isn't particularly great at holding runners).

He got better, by August 1st, but he was still getting caught at a 50% clip.

[Brad K]

I doubt it's worse than ever however I can believe no changes and as bad as ever. More data would be really nice.

[David Watts]

Finished the season with 52 steals and was caught 37 times. Maybe it was the year(1947) I was playing. The next closest guy only had 20 total on the year. Started a new random in 1979 last night. I screwed a bunch of stuff up in the one starting in 1947, so figured a fresh start was in order. Will be interesting to see how things play out steals wise in this one.

[Brad K]

David, playing in 1947 is not a valid test. Put him in his own era and see what happens. And the other leading stealers of that era.

[David Watts]

Originally Posted by Brad K

David, playing in 1947 is not a valid test. Put him in his own era and see what happens. And the other leading stealers of that era.

This new random I'm doing will take place using 1979(unless I decide to use 1979-1986, which might happen as well) I often get bored using only one year. This should give me a better idea. I was just shocked when I saw that Henderson had been caught so many times. Didn't think about it being a period when no one was really stealing all that much.

[David Watts]

Was just looking at 1947 on baseball reference. Success rates as a whole weren't great in either the NL or the AL, so I think I would erase this thread if I could. Total over reaction on my part.

[kq76]

I see you're playing an MLB 1947 season. Obviously that's not a historical team, but is the world completely fictional or is that just an expansion team? SB% dipped badly in the 50s, but 1947's was still okay at .695. My guess is you just have an off sample.

From my tests (see below), if you're playing a current or historical MLB league you should get pretty decent SB stats. But if you're playing fictional (even if current year/2022), you better import those historical year settings. Even then you might want to tweak the modifiers (notice the SB% is not bad, but the SB/G is not good). I doubt you're doing this (you're probably running a historical), but others should know that just because you start a fictional league in a certain year doesn't mean that it will result in stats similar to MLB's from that year.

[Brad K]

Originally Posted by David Watts

This new random I'm doing will take place using 1979(unless I decide to use 1979-1986, which might happen as well) I often get bored using only one year. This should give me a better idea. I was just shocked when I saw that Henderson had been caught so many times. Didn't think about it being a period when no one was really stealing all that much.

An idea but not really conclusive. Put him in the correct era and you still have players from other eras who will have different SB ratings from the historical players. You need correct era with all historical players.

Let's say you have a group of random players where the aggregate of their stolen base attempt rate is much less than in Henderson's era. The LTM file will bump up the attempts of all players to bring the attempt rate league wide up to what it is in the LTM file. Who knows exactly what it does but since it does something the output is suspect and not useful for analysis against historical.

[Scoman]

Just do a test of 1901-2022 historical default settings, you can see for yourself SB logic is off and stolen bases are higher in this version,it has already been reported with no response.

[Scoman]

Originally Posted by Scoman

Just do a test of 1901-2022 historical default settings, you can see for yourself SB logic is off and stolen bases are higher in this version,it has already been reported with no response.

https://forums.ootpdevelopments.com/...d.php?t=335979

See post 20 and post 69


I play strictly historical and this is the reason I have went back to 22.

Also this is playing with historical minors box unchecked.

Just wondering how much testing was done on long term sims in historical because this issue sticks out like a sore thumb and still no one has addressed it.

[thehef]

The issue with stolen bases - prior to v 23 (I don't know if it's been fixed; pretty sure it has at least been tweaked) - was that the best base stealers would get caught way too often. League totals would be good because lots of guys who had very few attempts would do better in OOTP. Individually, those wouldn't stand out, but cumulatively, that's how the league totals would look ok.

IOW, and I'm just making this up but you'll get the idea:

- The top 15 or 20 base stealers in any give season would get caught significantly more often. So guys then had, say, 50 SB and 18 CS in real life would go 55 & 35 in OOTP (and my extreme example would be Willie Davis, who always seemed to have a few OOTP seasons where he'd steal 32 and get caught 38 times! Definitely an immersion-buster!)

- Dozens - perhaps more - guys who IRL stole 3-out-of-6, 2-out-of-5, 4-out-of-10, etc... would in OOTP look like 5 out of 6, 3 out of 5, 7 out of ten... So those appear to be insignificant differences from real life, so - taken individually - we don't notice anything out of the ordinary. But when a bunch of low-attempt guys have somewhat higher success rates, that's how the best stealers can be way off while league totals look good.

Best way to test this is to run a long historical replay and then look at the career OOTP SB/CS numbers for guys like Brock, Lopes, Henderson, Morgan, Aparicio, W Davis...

[Brad K]

I have an old post about this with plenty of data and acknowledged by a dev,

Better graphics got the priority.

[Brad K]

I didn't think it would be so easy to find

https://forums.ootpdevelopments.com/...=base+stealers

How long do you think this problem has existed? I have OOTP11 on my old computer. Is it worth checking? LOL.

[Syd Thrift]

Originally Posted by Brad K

I have an old post about this with plenty of data and acknowledged by a dev,

Better graphics got the priority.

Was it prior to v23? Because graphics or not (and for the 10 millionth time, that’s a different team; you may as well be complaining that they didn’t do it because Rogue Legacy 2 got released), they revamped steals this year.

[Brad K]

Originally Posted by Syd Thrift

Was it prior to v23? Because graphics or not (and for the 10 millionth time, that’s a different team; you may as well be complaining that they didn’t do it because Rogue Legacy 2 got released), they revamped steals this year.

Don't try this different team BS. Fact is they've assigned people to work on graphics and enhancements and haven't assigned people to work on the basic functions of the game.

There are indications they're even working on stuff like AI warming up relievers when all that does is satisfy a few who can't deal with things in the abstract and which no matter how well done will give yet another advantage to the human player. What the game really needs is a Help AI switch.

So what's next for this thread? Is Garlon going to show up and give one of his explanations how its not a problem at all, just max out setting A and cut setting B to the minimum and when you do that all the SB stuff will be right (without mentioning that doing so screws up five other things).

[Brad K]

Originally Posted by Syd Thrift

they revamped steals this year.

Revamped but you didn't say fixed. Intentional?

See the link in post 12. Not fixed.

[kq76]

Originally Posted by Scoman

Just do a test of 1901-2022 historical default settings, you can see for yourself SB logic is off and stolen bases are higher in this version,it has already been reported with no response.

The single season and career leaderboards do seem a bit too high, but overall they don't look too bad to me. The Rock's 157 single season record is certainly impressive, but it doesn't irritate me in any way. I understand others would always prefer results to be closer, but if we were to err on the side of a touch too low or high, I'd much rather go too high with SB.

I only simmed 1920-2021, with all the defaults, as I figured that would be good enough to see how things were. Is there some other setting for historicals that likely would have made the career records more in line with reality? I checked and I did use 3 year recalc as that seems to be a setting many mention.

[kq76]

A comparison between OOTP and real for individual players below. I expect to see variation with individual players so I don't think too much of this (I'm more in awe of Raines's real life .847 SB% and Jose Reyes's OOTP run), but I thought others might.