Authentic gameplay: Frequency of PH & PR

SteveP · 02-05-2009, 01:20 PM

There is no need to rehash the problems with excessive PH/PR – and related fielding substitution problems -- in OOTP. However, my study of real 1960 games vs. my 1960 league did produce some useful insights related to this subject.

Some time ago, I adopted a practice of adjusting the League strategy setting for PH for position players to “Very rarely,” and the related setting in all team strategy pages near to the “Never” end of the slider. It’s clear now that these settings do produce reasonably authentic gameplay in OOTP. There is the side benefit that fielding substitution problems are greatly reduced.

I hit on a good metric to use in calibrating PH frequency. IRL, pinch-hitting for two position players in the same inning occurred seldom or very seldom (post-season play being a possible exception, as well as the rare manager who liked to pinch hit a lot for L/R matchup reasons). Pinch-hitting for three position players was about as rare as triple plays. I don’t believe this is era specific, but so far I’ve only studied 1960. My conclusion is that if you are seeing pinch-hitting for two or more position players in an inning happen fairly often for several teams, you should consider reducing the team strategy settings. With one exception, mentioned below, there’s not much you can do in OOTP to fine-tune this.

Because I like some variety in how teams in the league make gameplay decisions, I have many team strategy sliders set to one notch above “Never,” with some set to “Never” (which actually gives pretty authentic behavior compared to many teams IRL). I may have one or two teams set to two notches above “Never,” which this only works for teams that don’t fall easily into fielding substitution problems.

I also set pinch-running in Team strategy pages pretty much as I do pinch-hitting, again with some variation from team to team. This not only produces a frequency level that is reasonably authentic, but it greatly reduces the tendency of the AI to bench a fast baserunner for one that is slightly faster.

IRL, managers displayed a variety of behaviors as to when they would use pinch-hitters and pinch-runners. I don’t think there is any way to duplicate this variety in OOTP. So, I am satisfied if I can get the frequency per inning (which mostly means the 9th inning and sometimes the 8th inning) close to authentic.

There is one other setting that may affect the frequency of pinch-hitting. This is L/R matchup preference in the Team strategy pages. It appears to me that when this is set toward “Don’t care,” there is a slightly greater frequency of pinch-hitting per inning, whereas the opposite occurs if it is set toward “Prefer.” Logically so, I think. So, this may provide a tool for fine-tuning AI decision-making about frequency of PHing as well as matchups. Don’t overlook it in your own experimentation.

FWIW

pstrickert · 02-05-2009, 03:14 PM

This is helpful to know, SteveP. I've been re-acquainting myself with BBW 5.5 lately. One of the remarkable things about BBW is the creation and utilization of different computer managers. Users have created dozens and dozens of different AI managers (each with noticeably different AI "personalities") for use in BBW. I found one site that sorts these managers according to decade. So, it's possible to create AI managers with tendencies appropriate to the era (i.e., relief pitcher usage, PH/PR usage, sac bunts, etc.). And remember, BBW has not been updated in about ten years. Even so, it is "ahead of its time" in a number of respects.

SteveP · 02-06-2009, 03:39 PM

Quote:

Originally Posted by pstrickert

This is helpful to know, SteveP. I've been re-acquainting myself with BBW 5.5 lately. One of the remarkable things about BBW is the creation and utilization of different computer managers. Users have created dozens and dozens of different AI managers (each with noticeably different AI "personalities") for use in BBW. I found one site that sorts these managers according to decade. So, it's possible to create AI managers with tendencies appropriate to the era (i.e., relief pitcher usage, PH/PR usage, sac bunts, etc.). And remember, BBW has not been updated in about ten years. Even so, it is "ahead of its time" in a number of respects.

I know there is some interest among OOTP players about a feature like this. I think it might be something that works better in conception than in practice, but it might also be an improvement over the randomizing approach to manager attributes and strategies that we have now. Just don't know.

We would still need the ability to modify those managers, because unless things like PH/PR frequency is fixed, we need the team strategy pages as a tool for correcting unauthentic AI decision-making and gameplay that are systemic. So, right now, I'm glad we have that tool available to us. Since I want to vary settings in these pages, from team to team, I try to match it with what the real teams actually seemed to do in 1960 as much as I can. Part of this is just my interest with authentic gameplay, but part of it is also about getting smarter as to how much these sliders actually affect AI behavior. Eventually, I want to know with some precision what I will get when I adjust a setting or a slider to a lesser or greater degree.

PhillieFever · 02-07-2009, 07:50 PM

Great work Steve,its funny,I had every team's pinch hitting/def. sub settings at 1 click from never and didn't really get the results that I was looking for,once I moved it to never,presto

pstrickert · 02-07-2009, 09:01 PM

Quote:

Originally Posted by SteveP

I know there is some interest among OOTP players about a feature like this. I think it might be something that works better in conception than in practice, but it might also be an improvement over the randomizing approach to manager attributes and strategies that we have now. Just don't know.

We would still need the ability to modify those managers, because unless things like PH/PR frequency is fixed, we need the team strategy pages as a tool for correcting unauthentic AI decision-making and gameplay that are systemic. So, right now, I'm glad we have that tool available to us. Since I want to vary settings in these pages, from team to team, I try to match it with what the real teams actually seemed to do in 1960 as much as I can. Part of this is just my interest with authentic gameplay, but part of it is also about getting smarter as to how much these sliders actually affect AI behavior. Eventually, I want to know with some precision what I will get when I adjust a setting or a slider to a lesser or greater degree.

It would help if we had the option to save our strategy settings and share them. Then we could create an assortment of AI managers to choose from. Why Markus has not given us that option, I do not know.

SteveP · 02-08-2009, 12:39 AM

Quote:

Originally Posted by PhillieFever

Great work Steve,its funny,I had every team's pinch hitting/def. sub settings at 1 click from never and didn't really get the results that I was looking for,once I moved it to never,presto

You may find, as I did, that the best setting for most teams on pinch hitting is probably halfway between "never" and one notch (click) above "never." Oh, well. I'm happy to have PH behavior that on a league-wide basis looks pretty authentic. I wish some of the other behavior issues were that easy to fix.

On that note, and since you mentioned it, I haven't done an "authentic gameplay" thread on the defensive sub option yet. That's mostly because I can't quite figure out what exactly this option does, and how it best matches up with authentic gameplay. I'm not exactly sure what you meant by suggesting you set this as "1 click from never," as well. Was that a typo?

PhillieFever · 02-08-2009, 02:06 PM

Quote:

Originally Posted by SteveP

You may find, as I did, that the best setting for most teams on pinch hitting is probably halfway between "never" and one notch (click) above "never." Oh, well. I'm happy to have PH behavior that on a league-wide basis looks pretty authentic. I wish some of the other behavior issues were that easy to fix.

On that note, and since you mentioned it, I haven't done an "authentic gameplay" thread on the defensive sub option yet. That's mostly because I can't quite figure out what exactly this option does, and how it best matches up with authentic gameplay. I'm not exactly sure what you meant by suggesting you set this as "1 click from never," as well. Was that a typo?

It absolutely was a typo,I meant pinch running,good catch.I had just been looking at my setting for defensive substitution right before I posted,FWIW I have defensive subs on very rarely.

SteveP · 02-08-2009, 03:29 PM

Quote:

Originally Posted by PhillieFever

I have defensive subs on very rarely.

Have you figured out exactly what happens with each setting? I just haven't focused enough on it yet. What I would be looking for is some analysis that might look like this:

If set to very rarely, defensive subs only come in when the game is tied or the team is one up. If set to rarely, the subs come in only when the team is three up or less. If set to normal, the subs come in when the team is five up or less. Etc. Something like that.

So far, my nonscientific look at this hasn't suggested a clear pattern like that. So I have it set to "Very rarely" as well, but that is only because it looks to me like defensive subs IRL generally happened less often than they do in OOTP with default settings.

endgame · 02-08-2009, 04:10 PM

Good comments, SteveP. Earlier in my season, I did dial down - drastically - the pinch hit for position players setting to Never or within three notches of Never across the league for every individual team, and saw the improvement I had been hoping to achieve. In addition - and everyone will have their individual styles with this - dialed up or down (usually down) the use of pinch running to never exceed the frequency with which the team's utilize the hit and run. I thought that somewhat indicative of the management's treatment of their team's speed. That's my rationale, at least.

The final result is a much more realistic - IMHO - approach to lineup and substitutions than I'd been realizing. Of course, as a natural consequence, this is expected to drive up the plate appearances over the last seasons for positional players, but I'll be interested to see by just how much and whether it's of any real impact, where, and in what form. It could easily be a simple effect like lowering league ERA because better hitters stay in longer. I may find a slight reduction in stolen bases, as the pinch runners - typically faster - are not on the bases as often late in games creating those numbers. Again, I can't imagine it being of any magnitude, but it will be interesting to do some comparisons at year's end. I'm a solo fictional player, but I assume the feedback will be helpful nonetheless.

Thanks for the affirmations in your tips. It's moving the game's visual logic - at least to me - in the right direction as a game plays out. I play out most of my games, so it'll be a while. Regardless, I'll update you at the end of the season.

SteveP · 02-08-2009, 06:50 PM

Quote:

Originally Posted by endgame

Good comments, SteveP. Earlier in my season, I did dial down - drastically - the pinch hit for position players setting to Never or within three notches of Never across the league for every individual team, and saw the improvement I had been hoping to achieve. In addition - and everyone will have their individual styles with this - dialed up or down (usually down) the use of pinch running to never exceed the frequency with which the team's utilize the hit and run. I thought that somewhat indicative of the management's treatment of their team's speed. That's my rationale, at least.

My objectives are pretty simple (I tend to fall into the KISS camp, anyway

): 1) to stop the AI clearing the benches almost every game with PHs and PRs, which just plain looked stupid; and 2) to drastically cut down on the fielding substitution problems when it did this. It drives me crazy to routinely see OFs or INFs playing catcher in the 9th or 10th inning. I assume the game design made it so that anyone could play catcher for just this reason (and, OK, it did happen IRL when all of the astrological signs were lined up in just the right way), but I wish they had thought first to cut down on the PH and PR so stuff like this wouldn't happen so often.

Anyway, I'm glad there are tools in the game that allow us to get a reasonably good fix ourselves. One of things that I really like a lot in OOTP, is that it will PH for top players when the game is a blowout, so that does add some variety to what happens. OTOH, as we all know, the game does not make adjustments at the end of a season, based on who is in a pennant race or not, or for post-season play, or the like. That's just the way it is. So, do it yourself. One possibility is at the time of roster expansion, up the PH/PR a notch or two for teams who out of contention. Also up the PH/PR a notch for teams in post-season. That's one idea, anyway. Otherwise, my inclination is to leave it alone, because it does not take much in the way of extra PH/PR activity to trigger those dreaded fielding substitution problems.

02-05-2009, 01:20 PM	#1
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	Authentic gameplay: Frequency of PH & PR There is no need to rehash the problems with excessive PH/PR – and related fielding substitution problems -- in OOTP. However, my study of real 1960 games vs. my 1960 league did produce some useful insights related to this subject. Some time ago, I adopted a practice of adjusting the League strategy setting for PH for position players to “Very rarely,” and the related setting in all team strategy pages near to the “Never” end of the slider. It’s clear now that these settings do produce reasonably authentic gameplay in OOTP. There is the side benefit that fielding substitution problems are greatly reduced. I hit on a good metric to use in calibrating PH frequency. IRL, pinch-hitting for two position players in the same inning occurred seldom or very seldom (post-season play being a possible exception, as well as the rare manager who liked to pinch hit a lot for L/R matchup reasons). Pinch-hitting for three position players was about as rare as triple plays. I don’t believe this is era specific, but so far I’ve only studied 1960. My conclusion is that if you are seeing pinch-hitting for two or more position players in an inning happen fairly often for several teams, you should consider reducing the team strategy settings. With one exception, mentioned below, there’s not much you can do in OOTP to fine-tune this. Because I like some variety in how teams in the league make gameplay decisions, I have many team strategy sliders set to one notch above “Never,” with some set to “Never” (which actually gives pretty authentic behavior compared to many teams IRL). I may have one or two teams set to two notches above “Never,” which this only works for teams that don’t fall easily into fielding substitution problems. I also set pinch-running in Team strategy pages pretty much as I do pinch-hitting, again with some variation from team to team. This not only produces a frequency level that is reasonably authentic, but it greatly reduces the tendency of the AI to bench a fast baserunner for one that is slightly faster. IRL, managers displayed a variety of behaviors as to when they would use pinch-hitters and pinch-runners. I don’t think there is any way to duplicate this variety in OOTP. So, I am satisfied if I can get the frequency per inning (which mostly means the 9th inning and sometimes the 8th inning) close to authentic. There is one other setting that may affect the frequency of pinch-hitting. This is L/R matchup preference in the Team strategy pages. It appears to me that when this is set toward “Don’t care,” there is a slightly greater frequency of pinch-hitting per inning, whereas the opposite occurs if it is set toward “Prefer.” Logically so, I think. So, this may provide a tool for fine-tuning AI decision-making about frequency of PHing as well as matchups. Don’t overlook it in your own experimentation. FWIW

02-07-2009, 07:50 PM	#4
PhillieFever Hall Of Famer Join Date: Apr 2008 Location: Elk Twp. NJ Posts: 6,763	Great work Steve,its funny,I had every team's pinch hitting/def. sub settings at 1 click from never and didn't really get the results that I was looking for,once I moved it to never,presto __________________ We're All Wednesday Aren't We? WAWAW

02-08-2009, 04:10 PM	#9
endgame Hall Of Famer Join Date: Aug 2002 Posts: 16,842	Good comments, SteveP. Earlier in my season, I did dial down - drastically - the pinch hit for position players setting to Never or within three notches of Never across the league for every individual team, and saw the improvement I had been hoping to achieve. In addition - and everyone will have their individual styles with this - dialed up or down (usually down) the use of pinch running to never exceed the frequency with which the team's utilize the hit and run. I thought that somewhat indicative of the management's treatment of their team's speed. That's my rationale, at least. The final result is a much more realistic - IMHO - approach to lineup and substitutions than I'd been realizing. Of course, as a natural consequence, this is expected to drive up the plate appearances over the last seasons for positional players, but I'll be interested to see by just how much and whether it's of any real impact, where, and in what form. It could easily be a simple effect like lowering league ERA because better hitters stay in longer. I may find a slight reduction in stolen bases, as the pinch runners - typically faster - are not on the bases as often late in games creating those numbers. Again, I can't imagine it being of any magnitude, but it will be interesting to do some comparisons at year's end. I'm a solo fictional player, but I assume the feedback will be helpful nonetheless. Thanks for the affirmations in your tips. It's moving the game's visual logic - at least to me - in the right direction as a game plays out. I play out most of my games, so it'll be a while. Regardless, I'll update you at the end of the season. __________________ "Try again. Fail again. Fail better." -- Samuel Beckett _____________________________________________ Last edited by endgame; 02-08-2009 at 04:14 PM.

02-05-2009, 03:14 PM	#2
pstrickert Hall Of Famer Join Date: Dec 2005 Posts: 16,527	This is helpful to know, SteveP. I've been re-acquainting myself with BBW 5.5 lately. One of the remarkable things about BBW is the creation and utilization of different computer managers. Users have created dozens and dozens of different AI managers (each with noticeably different AI "personalities") for use in BBW. I found one site that sorts these managers according to decade. So, it's possible to create AI managers with tendencies appropriate to the era (i.e., relief pitcher usage, PH/PR usage, sac bunts, etc.). And remember, BBW has not been updated in about ten years. Even so, it is "ahead of its time" in a number of respects.