Best stealers underperforming historical success rates - Page 2

polydamas · 08-30-2020, 02:12 AM

Quote:

Originally Posted by fredbeene

Sure that would be great.
What is the setting.

Not sure setting the sucess rate is equivalent to the actual ratings of the players.
Rating would barely matter???

https://forums.ootpdevelopments.com/...49#post4673749

The second post there has a bunch of pics.

I used 1.050 (for Stolen Base Success %) for that game, but I would change it to 1.025 and test that first.

Stolen base success in my league is a bit too high. I have one guy at about 90% and a handful over 80%

Brad K · 08-30-2020, 11:43 AM

Quote:

Originally Posted by polydamas

Why not change one setting instead of re-rating every player in the league?

Because I've shown the problem is not an overall problem. Its just at the high end. Besides, this isn't a preference thing appropriate for user adjustment. Its a clear defect the developers need to fix.

thehef · 09-19-2020, 03:34 PM

Quote:

Originally Posted by Brad K

Because I've shown the problem is not an overall problem. Its just at the high end. Besides, this isn't a preference thing appropriate for user adjustment. Its a clear defect the developers need to fix.

A recent historical sim reminded me of this issue, where I'm seeing Davey Lopes, Lou Brock, and many of the others you mentioned earlier in this thread with lower-than-historical success rates. Sure seems like this needs to be addressed.

Do you happen to know if it's been brought up in the beta forums? Based upon a very quick search, I don't think it has.

Brad K · 09-19-2020, 04:47 PM

I didn't post about it anywhere but here. I wanted to see if there was a consensus on the issue. It seems like there is. What do you suggest we do?

thehef · 09-19-2020, 05:42 PM

Quote:

Originally Posted by Brad K

I didn't post about it anywhere but here. I wanted to see if there was a consensus on the issue. It seems like there is. What do you suggest we do?

I just posted it in the beta forum... basically just link to this thread plus a bit of background, so the devs can become aware of the issue (if they are not already) and hopefully address it.

thehef · 09-19-2020, 05:50 PM

Not that one-offs are proof, but in real life 1968, Willie Davis stole 36 based and was caught just 10 times. In a test historical, in 1968 he swiped 38 but was caught 35 times.

IRL 1978 & 79, Davey Lopes stole a total of 89 bases while being caught just 8 times (92% success rate). For the same two years in OOTP, he stole 109 bases and was caught 36 times (75%).

In Maury Wills' record-setting year of 1962, he swiped 104 based and was caught just 13 times (89%). In OOTP that same year he was 118 out of 151 (78%).

Again, these are just one-off examples, but they support the totals in the OP.

Brad K · 09-19-2020, 06:03 PM

Quote:

Originally Posted by thehef

I just posted it in the beta forum... basically just link to this thread plus a bit of background, so the devs can become aware of the issue (if they are not already) and hopefully address it.

OK, cool. Where is this forum?

thehef · 09-19-2020, 06:07 PM

Quote:

Originally Posted by Brad K

OK, cool. Where is this forum?

You have to be a member of the beta team in order to view the forum. If you are interested in being a member, you might want to check with Lukas (send him a PM). You seem to know enough about the game to be able to test it and contribute, but I have no idea if they have enough of those folks or what... But Lukas can let you know

68pirate · 09-20-2020, 09:42 AM

I haven't looked into this, but I wonder if historically low success rate stealers aren't running at all. For example, in 1979 Mike Cubbage had 9 attempts and only one successful steal. If the AI isn't running him at all, those caught stealings might be transferred to guys that do run. That way the overall numbers match up in the historical comparison, even though the distribution doesn't match individually.

Brad K · 09-20-2020, 03:02 PM

Quote:

Originally Posted by thehef

You have to be a member of the beta team in order to view the forum. If you are interested in being a member, you might want to check with Lukas (send him a PM). You seem to know enough about the game to be able to test it and contribute, but I have no idea if they have enough of those folks or what... But Lukas can let you know

I'll think about that but my initial impression is that primarily playing late 60s to late 80s by experience isn't broad enough.

Could you possibly let us know in this thread what kind of reception the concept is getting over there?

Brad K · 09-20-2020, 03:06 PM

Quote:

Originally Posted by 68pirate

I haven't looked into this, but I wonder if historically low success rate stealers aren't running at all. For example, in 1979 Mike Cubbage had 9 attempts and only one successful steal. If the AI isn't running him at all, those caught stealings might be transferred to guys that do run. That way the overall numbers match up in the historical comparison, even though the distribution doesn't match individually.

Yea, someone has to be shorted on CS. It would be interesting to find the inflection point between having CS added and CS subtracted. Will think about an efficient way to recover the data. Pretty tough looking at everyone instead of just the top stealers.

thehef · 09-20-2020, 03:45 PM

Quote:

Originally Posted by Brad K

I'll think about that but my initial impression is that primarily playing late 60s to late 80s by experience isn't broad enough.

Not necessarily. I'm exclusively historical, mostly 50's to 80's (though I start alot of my sims earlier just to have the history). And I don't actually do a lot of testing, unlike many over there. Rather, my contribs are historical minor league ballpark names, and historical minor league franchise histories (both works in progress, not fully rolled out in OOTP yet)... If it interests you, hit up Lukas to see if you might fill a need. If not, no worries.

Quote:

Originally Posted by Brad K

Could you possibly let us know in this thread what kind of reception the concept is getting over there?

Absolutely. So, far: 13 views and no comments. Not unusual for this period in cycle of OOTP. Someone will have something to say eventually. Else I will bump it and ask Lukas to bring it to the devs' attention. I'm fairly certain they'll look at it. I have no idea on this, but if a fix or adjustment is involved enough, then it might be something that will wait for OOTP 22. But again, I have no clue.

Brad K · 09-20-2020, 03:48 PM

Thank you.

thehef · 09-20-2020, 03:53 PM

Quote:

Originally Posted by 68pirate

I haven't looked into this, but I wonder if historically low success rate stealers aren't running at all. For example, in 1979 Mike Cubbage had 9 attempts and only one successful steal. If the AI isn't running him at all, those caught stealings might be transferred to guys that do run. That way the overall numbers match up in the historical comparison, even though the distribution doesn't match individually.

Quote:

Originally Posted by Brad K

Yea, someone has to be shorted on CS. It would be interesting to find the inflection point between having CS added and CS subtracted. Will think about an efficient way to recover the data. Pretty tough looking at everyone instead of just the top stealers.

A super-small sample size plucked nearly randomly (just Cubbage and some other slow guys that popped into my head). At any rate:

Player, Real Life Steals & Caught, 1st Sim Steals & Caught, 2nd Sim Steals & Caught

HTML Code:

Cubbage		RL: 6-15	Sim 1: 12-6	Sim 2: 7-7
Yeager		RL: 14-18	Sim 1: 7-9	Sim 2: 10-5
Cey		RL: 24-29	Sim 1: 14-7	Sim 2: 11-10
McCovey		RL: 26-22	Sim 1: 20-16	Sim 2: 21-28
Stargell	RL: 17-16	Sim 1: 20-10	Sim 2: 12-11
Munson		RL: 48-50	Sim 1: 29-19	Sim 2: 39-28
Fielder		RL: 2-6		Sim 1: 0-8	Sim 2: 0-2

Based upon the above, I'd say the guys who run infrequently are running somewhat less in OOTP, but finding a bit more success.

Brad K · 09-20-2020, 06:26 PM

Another thing is those guys SB/CS may be occurring on broken hit and run plays. So there's a question of whether running in that situation functions the same way as a steal attempt.

Brad K · 09-20-2020, 06:28 PM

I'm going to look into my long term sim but I'm mostly posting during breaks between yard work. Between a lot of rain and a lot of heat I've gotten way behind. Plus I have some equipment maintenance to do. I'll get some data.

Brad K · 09-20-2020, 06:31 PM

I see Stargell is 17 SB 16 CS. Here is one of the CS. Looks like Berra missed a hit and run sign or couldn't pull the trigger.

https://www.youtube.com/watch?v=nZ194opBr8Y

Lukas Berger · 09-21-2020, 06:52 PM

Quote:

Originally Posted by thehef

Absolutely. So, far: 13 views and no comments. Not unusual for this period in cycle of OOTP. Someone will have something to say eventually. Else I will bump it and ask Lukas to bring it to the devs' attention. I'm fairly certain they'll look at it. I have no idea on this, but if a fix or adjustment is involved enough, then it might be something that will wait for OOTP 22. But again, I have no clue.

We've been following this thread.

It's very likely not going to be anything we'll be able to get in this year but I'm quite sure this will get looked at during the dev cycle for OOTP22.

Brad K · 09-22-2020, 05:17 PM

Concerning the SB success rate of those with a low rate of attempts, the data can't be sorted within the game into what we need and some - maybe a lot - of manual sorting will be needed.

For example, if we say 25 to 50 career attempts, well, doing that in a ten year career isn't the same as doing it in a two year career. Additionally the game doesn't list attempts separately so the initial sort will need to be on SB or CS. Then some kind of qualifier for opportunities... maybe PAs as a rough measure? (The game doesn't show singles separately.)

I don't know... I want to do this as I have 30+ years of sim data but am struggling with how to proceed.

thehef · 09-23-2020, 01:09 AM

Maybe try the Export Player Stats feature (on the League Settings > Functions panel, right side under Import / Export). Then you can pretty easily add SB's & CS's to get attempts, and also I'm thinking you're on the right path by looking at a ratio of SB Attempts to Plate Appearances to differentiate between high-attempt guys vs low-attempt guys, and those in the middle... And of course all of this will need to ultimately be compared to real-life stats.

In the beta forum, Lukas has expressed the outlines of some testing that should happen for this at some point. The details of what he means is a bit beyond me

but I wouldn't be surprised if he reaches out to you, given your interest & willingness to help out.

09-19-2020, 04:47 PM	#24
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	I didn't post about it anywhere but here. I wanted to see if there was a consensus on the issue. It seems like there is. What do you suggest we do?

09-19-2020, 05:50 PM	#26
thehef Hall Of Famer Join Date: Jun 2006 Posts: 5,345	Not that one-offs are proof, but in real life 1968, Willie Davis stole 36 based and was caught just 10 times. In a test historical, in 1968 he swiped 38 but was caught 35 times. IRL 1978 & 79, Davey Lopes stole a total of 89 bases while being caught just 8 times (92% success rate). For the same two years in OOTP, he stole 109 bases and was caught 36 times (75%). In Maury Wills' record-setting year of 1962, he swiped 104 based and was caught just 13 times (89%). In OOTP that same year he was 118 out of 151 (78%). Again, these are just one-off examples, but they support the totals in the OP.

09-20-2020, 09:42 AM	#29
68pirate Minors (Single A) Join Date: Apr 2008 Posts: 84	I haven't looked into this, but I wonder if historically low success rate stealers aren't running at all. For example, in 1979 Mike Cubbage had 9 attempts and only one successful steal. If the AI isn't running him at all, those caught stealings might be transferred to guys that do run. That way the overall numbers match up in the historical comparison, even though the distribution doesn't match individually.

09-20-2020, 03:48 PM	#33
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	Thank you.

09-20-2020, 06:26 PM	#35
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	Another thing is those guys SB/CS may be occurring on broken hit and run plays. So there's a question of whether running in that situation functions the same way as a steal attempt.

09-20-2020, 06:28 PM	#36
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	I'm going to look into my long term sim but I'm mostly posting during breaks between yard work. Between a lot of rain and a lot of heat I've gotten way behind. Plus I have some equipment maintenance to do. I'll get some data.

09-20-2020, 06:31 PM	#37
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	I see Stargell is 17 SB 16 CS. Here is one of the CS. Looks like Berra missed a hit and run sign or couldn't pull the trigger. https://www.youtube.com/watch?v=nZ194opBr8Y

09-22-2020, 05:17 PM	#39
Brad K Hall Of Famer Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,657 Infractions: 0/2 (3)	Concerning the SB success rate of those with a low rate of attempts, the data can't be sorted within the game into what we need and some - maybe a lot - of manual sorting will be needed. For example, if we say 25 to 50 career attempts, well, doing that in a ten year career isn't the same as doing it in a two year career. Additionally the game doesn't list attempts separately so the initial sort will need to be on SB or CS. Then some kind of qualifier for opportunities... maybe PAs as a rough measure? (The game doesn't show singles separately.) I don't know... I want to do this as I have 30+ years of sim data but am struggling with how to proceed.

09-23-2020, 01:09 AM	#40
thehef Hall Of Famer Join Date: Jun 2006 Posts: 5,345	Maybe try the Export Player Stats feature (on the League Settings > Functions panel, right side under Import / Export). Then you can pretty easily add SB's & CS's to get attempts, and also I'm thinking you're on the right path by looking at a ratio of SB Attempts to Plate Appearances to differentiate between high-attempt guys vs low-attempt guys, and those in the middle... And of course all of this will need to ultimately be compared to real-life stats. In the beta forum, Lukas has expressed the outlines of some testing that should happen for this at some point. The details of what he means is a bit beyond me but I wouldn't be surprised if he reaches out to you, given your interest & willingness to help out.