Authentic gameplay: Bunting

SteveP · 01-31-2009, 05:52 PM

First the bad news: I’ve found it impossible to get bunting as a whole to work in OOTP in a way that comes at all close to authentic gameplay in my 1960s based league, no matter what I did with options and settings. OOTP appears to handle bunting even less well than pinch-hitting. The good news is that it’s possible to make some noticeable improvements to sac bunting.

OOTP appears to have a systemic problem with bunt ratings for historical players, at least in this era. Either that or the game engine uses these ratings badly. I plan to try some experiments with rating adjustments, to see what I can learn. For one thing, the AI never calls bunt-for-hit plays in my test league. This can’t be fixed with any of the settings. The AI also never calls squeeze plays, which may or may not be ratings related.

The most important problem with sac bunting is that pitchers SH too often, and position players not nearly enough. In 1960, pitchers accounted for less than 40% of all SH plays. With default settings in OOTP, the proportion is around 90%. This is not only unhistorical, but makes the game too predictable and too boring. IRL, almost every position player executed a sac bunt at some time during the season.

Other SH related problems: the AI sac bunts too often with a runner on second, one out; the AI sac bunts with runners on 1st & 3rd (which happens IRL only as a squeeze play); and the AI sac bunts with two strikes too often.

The settings that I ended up using do help get SH closer to authentic gameplay, however, they need to be tested for other eras.

First, I set the League strategy option for bunting to “Very Often”. This increases the proportion of SH plays by position players significantly -- much closer to historical levels. Doing this requires that overall SH plays be reduced substantially. The only way this can be done effectively is in the League Totals Modifier for sac bunts. For my test league, the best setting seems to be about 25% of normal (i.e., .250 if the normal setting would otherwise be 1.000), but I think that may still be too high.

With these adjustments, you can use normal settings for sac bunting in the Team strategy pages (which I like to vary from team to team). One caution: if you use my settings, you may find some instances of sac bunting by position players disconcerting, as the AI tends to be eccentric in its play calling. One thing I’ve done is to modify the team strategy pages for all teams so that sac bunting is set to “Never” when the run differential is 4+ in either direction. This adds some work to maintaining these pages, however, so I don’t know if I’m going to stick with doing that.

FWIW

Spritze · 02-01-2009, 02:05 AM

Here are the actual MLB numbers for the percentage of SH by pitchers for 1895-2008.

Perhaps if you compare these percentages on a yearly basis to what OOTP produces it will give the game programmer something to go by when fixing this.

Garlon · 02-01-2009, 06:09 AM

The League Totals Modifiers will automatically calculate the appropriate value for Sac Bunts each season to keep the game on pace with the historical SAC/AB ratios.

Pitchers may rack up more Sac's than you may like, but the league as a whole will be on target. I don't think we should worry too much about this issue right now, as there are other more important things that need some work in historicals.

SteveP · 02-01-2009, 02:53 PM

My effort to look at what makes for authentic gameplay is not about season or career stats, except indirectly. It's about how closely individual games played during a season in OOTP compare to how games were really played in that season. Also, I am not posting these to raise issues or complain about game design (unless, of course, Markus decides to take more of an interest in gameplay at this level

). My primary objective is to see what can be done with the available tools to get more authentic gameplay.

One of the main themes of this thread is that having pitchers automatically sac bunt with men on base is both substantially unhistorical and boring -- as is position players rarely sac bunting in these situations. The good news is that this can be much improved with the available tools, for people like myself who care about what happens in individual games. I hope readers will regard this and similar threads in that spirit.

pstrickert · 02-01-2009, 05:01 PM

Quote:

Originally Posted by SteveP

My effort to look at what makes for authentic gameplay is not about season or career stats, except indirectly. It's about how closely individual games played during a season in OOTP compare to how games were really played in that season. Also, I am not posting these to raise issues or complain about game design (unless, of course, Markus decides to take more of an interest in gameplay at this level

). My primary objective is to see what can be done with the available tools to get more authentic gameplay.

One of the main themes of this thread is that having pitchers automatically sac bunt with men on base is both substantially unhistorical and boring -- as is position players rarely sac bunting in these situations. The good news is that this can be much improved with the available tools, for people like myself who care about what happens in individual games. I hope readers will regard this and similar threads in that spirit.

I think this is where Garlon and I see things differently. Garlon is more interested in macro-level results. I, like you, want things to look "right" at the micro-level. Based on the number of single-season replay type games out there, I'll bet many people share our interest. Indeed, many people have fond memories of specific seasons and enjoy "reliving" them. I hope OOTP10 would offer more options for this type of user.

PhillieFever · 02-01-2009, 09:51 PM

Quote:

Originally Posted by pstrickert

I think this is where Garlon and I see things differently. Garlon is more interested in macro-level results. I, like you, want things to look "right" at the micro-level. Based on the number of single-season replay type games out there, I'll bet many people share our interest. Indeed, many people have fond memories of specific seasons and enjoy "reliving" them. I hope OOTP10 would offer more options for this type of user.

I'm with you guys,I play out every game as well and it really takes away from the sim when you know the game isn't being played the right way.Thats going to be my main concern during beta,stats on the micro-level.

Spritze · 02-01-2009, 11:53 PM

I'm a "with you guys" guy too.

I play out most games and notice the same things. Numbers are fine and dandy but they ain't "real".

Really real is why guys like Joe Ginsburg used to lead the league in intentional walks. It wasn't because of his dangerous bat it was because he always hit eighth in front of pitchers who'd strike out a lot. This sort of thing doesn't track in OOTP.

SteveP · 02-05-2009, 01:34 PM

An update on my experimentation in this area. I did a wholesale adjustment of player bunt ratings. Roughly speaking, the original rosters created from the Lahman DB had the peak of the distribution curve for bunt-for-hit ratings somewhere in the 15-20 range (on the 0-250 scale). So nobody bunts. I raised these so that the peak is somewhere in the 90-100 range. I also raised the sac bunt ratings for position players, but only about half as much. There was no science behind this, just trial-and-error. Since I haven't a clue how the original ratings were derived or how they are supposed to work (and, of course, they don't work anyway

), I was just taking my best shot. The point is that the result of this much change is quite striking. Bunting is now truly part of the game, and I can now use the available tools to adjust this to get pretty authentic results. Some people might think it unhistorical to see a Willie Mays or Hank Aaron occasionally lead off an inning with a bunt, but in fact this did happen. I now can get this sort of behavior in the game (along with the AI's somewhat eccentric decision-making of course, but that's unavoidable

).

OTOH, I still have yet to see the AI call a squeeze bunt. Has anyone? I suspect it is either not in the AI's repertoire, or it's been given a much higher rarity value than should be the case.

Garlon · 02-05-2009, 07:03 PM

SteveP -

...and now you have a ton of guys in the league racking up 30-40 bunt hits per year.

Give us a list of the .300 hitters in the league now.

There was a reason why we made the Bunt-For-Hit ratings the way they are in historicals.

SteveP · 02-05-2009, 08:54 PM

Quote:

...and now you have a ton of guys in the league racking up 30-40 bunt hits per year.

No, not so. I've gone from having no bunting whatsoever, to the occasional bunt for hit. If you have read my OP in this and related threads, you will note that I am comparing on a game-by-game basis what happens in OOTP versus what happened in real 1960 games. If there was too much bunting going on in my league, it would be obvious right away. Unfortunately, there are no stats kept on this (except for SH), so that's the only way to test this.

Part of the confusion may have to do with the way the game engine decides about offensive plays. It seems to make a new decision on each pitch. So, I've noticed there are quite a few bunt attempts with my new ratings (batter squares, but takes the pitch or misses the bunt) on pitches prior to 2 strikes. I have no way of comparing the frequency of these attempts to what may have happened in real life -- it may be a reasonable simulation of historical gameplay, or it may just be an unavoidable consequence of how the game engine works. Don't know. So, I only look at how often the ball is put into play on a bunt.

Quote:

There was a reason why we made the Bunt-For-Hit ratings the way they are in historicals.

It would be good to know what that reason is. Perhaps it was to suppress bunt attempts of the sort I describe above. However, the result is no bunting at all. We've also established that the game does not come close to historical experience in the relative distribution of sac bunts between pitchers and position players, if you use anything like default settings. So, I think there is something systemically wrong with bunting in OOTP (no complaints or issue-raising intended, just stating an observation

). I think that ratings would be a big factor in that, or else it's the way in which the game engine uses those ratings.

RchW · 02-05-2009, 08:54 PM

Quote:

Originally Posted by SteveP

An update on my experimentation in this area. I did a wholesale adjustment of player bunt ratings. Roughly speaking, the original rosters created from the Lahman DB had the peak of the distribution curve for bunt-for-hit ratings somewhere in the 15-20 range (on the 0-250 scale). So nobody bunts. I raised these so that the peak is somewhere in the 90-100 range. I also raised the sac bunt ratings for position players, but only about half as much. There was no science behind this, just trial-and-error. Since I haven't a clue how the original ratings were derived or how they are supposed to work (and, of course, they don't work anyway

), I was just taking my best shot. The point is that the result of this much change is quite striking. Bunting is now truly part of the game, and I can now use the available tools to adjust this to get pretty authentic results. Some people might think it unhistorical to see a Willie Mays or Hank Aaron occasionally lead off an inning with a bunt, but in fact this did happen. I now can get this sort of behavior in the game (along with the AI's somewhat eccentric decision-making of course, but that's unavoidable

).

OTOH, I still have yet to see the AI call a squeeze bunt. Has anyone? I suspect it is either not in the AI's repertoire, or it's been given a much higher rarity value than should be the case.

I think I understand what you are trying to do but using Mays and Aaron as examples is not the best. Mays had 13 sac in 22 seasons 6 in the decade of the 60's. Aaron quite amazingly had 21 in 22 seasons but 18 of those were in his first 3 seasons 1954-56. He did not have a sac in the 1960's. So I would be on the "unhistorical" side if Mays and Aaron are bunting in a 60's league.

BTW that applies to many power hitters in the 1960's.

Yazstremski 13/23 seasons
Frank Howard 7/16 seasons
Frank Robinson 17/21 seasons
Harmon Killebrew 0/16 seasons
Mantle 14/18 seasons
McCovey 5/22 seasons

SteveP · 02-05-2009, 09:02 PM

Quote:

Originally Posted by RchW

I think I understand what you are trying to do but using Mays and Aaron as examples is not the best. Mays had 13 sac in 22 seasons 6 in the decade of the 60's. Aaron quite amazingly had 21 in 22 seasons but 18 of those were in his first 3 seasons 1954-56. He did not have a sac in the 1960's. So I would be on the "unhistorical" side if Mays and Aaron are bunting in a 60's league.

I can point you to specific games played in 1960 in which Mays or Aaron hit bunts to get on base as leadoff batters in an inning. If I see something like that happen once or twice in a season, I want to know it's possible for it to happen once or twice when I play my 1960 league. What I am trying to do is no more complicated than that.

You are correct that they did not SH. And I don't suspect either will in my league, despite what I did with the ratings -- though I would certainly notice if it happened.

RchW · 02-05-2009, 10:24 PM

Quote:

Originally Posted by SteveP

I can point you to specific games played in 1960 in which Mays or Aaron hit bunts to get on base as leadoff batters in an inning. If I see something like that happen once or twice in a season, I want to know it's possible for it to happen once or twice when I play my 1960 league. What I am trying to do is no more complicated than that.

You are correct that they did not SH. And I don't suspect either will in my league, despite what I did with the ratings -- though I would certainly notice if it happened.

Fair enough. I don't believe OOTP9 keeps stats on bunt hits, which is too bad. I guess I was misled by your OP which seemed to focus on SH.

SH are a strategic AI decision. Bunting for a hit has some AI strategy to it but is really a player skill that can be substantially affected by the player strategy settings (set to override). Have you looked at that?

SteveP · 02-06-2009, 03:06 PM

Quote:

Fair enough. I don't believe OOTP9 keeps stats on bunt hits, which is too bad. I guess I was misled by your OP which seemed to focus on SH.

At the time I posted my OP, there was nothing much I could say about bunt-for-hits and squeeze bunts, other than that they never happened. So it made sense to focus more on the SH aspect of bunting, about which I had more useful info to offer. The problems with SH (excepting squeeze plays) can be corrected to a great extent (though not completely) with the available tools. But I now think there is a systemic problem with bunting overall that is probably best solved by modifying ratings. However, only people with a strong interest in getting authentic gameplay will want to do that.

Quote:

SH are a strategic AI decision. Bunting for a hit has some AI strategy to it but is really a player skill that can be substantially affected by the player strategy settings (set to override). Have you looked at that?

By setting bunt-for-hit in all of the team strategy pages all the way to frequently, as well as having the league strategy option for bunting set all the way to "Very Often," I think I've exhausted what the available tools will do (and I still get no bunting). I don't think modifying the player strategies will add anything to frequency at that point. Moreover, as much trouble as it is to tweak and maintain 16 team strategy pages in order to get authentic gameplay for some dimension of the game, I'm not going to start tweaking 500 individual player strategy pages.

SteveP · 02-06-2009, 03:26 PM

I did a little check on the frequency of real life bunt-for-hit plays in 1960 compared to what I was getting in my league. A bit time-consuming, so I could only check 20 games. I may try to keep track of this manually going forward, but having looked at this season in detail a couple of times already, I think what I am looking at is tolerably representative.

IRL, there was one bunt-for-hit attempt (from one team or the other) roughly every 2-3 games (this is exactly the sort of metric I like to help me figure out if I'm getting authentic gameplay

). It's difficult to be perfectly precise because you can't always be certain what the intent was related to a bunt with men on base. If you see a player who is known to bunt-for-hit with no one on base, and he bunts with a man on, how do you know he was sac bunting? I assume that OOTP always looks to the sac bunt rating in any sac bunt opportunity, but if so this is probably too constrained to be realistic.

I am getting a frequency in my league that is slightly higher than this, but I have been tweaking the relevant settings, so I will see what it looks like going forward.

Side note: in the sample from real games I looked at, a position player bunted with a man on 2nd, 2 outs. I counted that as a bunt-for-hit. However, I suspect I will never see a play like that in OOTP, no matter what I might modify, because of the combination of a man on base and 2 outs. Oh, well.

Garlon · 02-06-2009, 06:59 PM

SteveP -

Make a list of your top Bunt For Hit rated guys in the league, and then give us a sim BA vs real BA for each of them. You will see that they consistently overperform by a considerable margin. All those bunt hits become bonus hits on top of the hits that they will normally get. Their bunt hits won't replace line drive singles to CF, they will just be in addition to those hits.

We can discuss the reasons why Bunt For Hit is curently the way it is in historicals when we get a beta board for OOTP10.

Essentially, it was turning .250 hitters into .350 hitters, over 500 AB that's like 40-50 bunt singles. And there ended up being thousands of these bunt hits in the league every year. That was causing great hitters like Ty Cobb struggle to hit .315, while Joe Schmoes were consistently hitting .340+ each season. There are only so many hits to spread across the league in historicals, and the high Bunt For Hit guys just rack up way too many that they don't deserve.

SteveP · 02-07-2009, 05:36 PM

Quote:

Originally Posted by Garlon

Make a list of your top Bunt For Hit rated guys in the league, and then give us a sim BA vs real BA for each of them. You will see that they consistently overperform by a considerable margin. All those bunt hits become bonus hits on top of the hits that they will normally get. Their bunt hits won't replace line drive singles to CF, they will just be in addition to those hits.

OK, I get the picture. This is helpful information. I can't go around raising bunt ratings while leaving BABIP ratings (I assume that's what we are talking about here) alone, without messing up the BAs. Probably should have guessed that, but that's one of the reasons I started these threads -- so people like me could better understand what was going on with some of this stuff.

Is that also the reason why position players don't get to sac bunt much in OOTP? Because some of those sac bunts will turn into singles? (And who cares about pitcher BAs, anyway

)

Are you also saying that this in an inherent limitation of the game engine -- that we will never see bunting behave in a historically realistic way? There are always some limitations in a simulation game, so maybe that's one of them for OOTP. Would be good to know.

Quote:

Essentially, it was turning .250 hitters into .350 hitters, over 500 AB that's like 40-50 bunt singles. And there ended up being thousands of these bunt hits in the league every year. That was causing great hitters like Ty Cobb struggle to hit .315, while Joe Schmoes were consistently hitting .340+ each season. There are only so many hits to spread across the league in historicals, and the high Bunt For Hit guys just rack up way too many that they don't deserve.

I have to idea what you did to get such results. I won't see anything like that happen in my league, assuming I calibrate it closely enough to historical experience. I should get less than 30 bunt singles per team in the season from raising the bunt-for-hit ratings plus the other adjustments I am using. I'm thinking that a more active bunter on a team (leadoff batter, for example) gets 4-6 of these in a season -- maybe 10+ points more on his BA. The league as a whole will get about 2-2.5% more hits in the season. So the effect after season-end re-calibration is that non-bunters get a slight reduction in BA, occasional bunters get BAs about right, active bunters get a boost of 7-8 points. Not apocalyptic, but not desirable either. I could reduce bunting to half of what it was IRL. Bunts every 5-6 games would at least make individual games in OOTP more varied and less boringly predictable.

Anyway, it's food for thought, and I hope useful information for any readers of this thread.

RchW · 02-07-2009, 05:47 PM

Quote:

Originally Posted by SteveP

By setting bunt-for-hit in all of the team strategy pages all the way to frequently, as well as having the league strategy option for bunting set all the way to "Very Often," I think I've exhausted what the available tools will do (and I still get no bunting). I don't think modifying the player strategies will add anything to frequency at that point. Moreover, as much trouble as it is to tweak and maintain 16 team strategy pages in order to get authentic gameplay for some dimension of the game, I'm not going to start tweaking 500 individual player strategy pages.

Would you really have to do more than 2-3 players per team?

SteveP · 02-08-2009, 12:30 AM

Quote:

Originally Posted by RchW

Would you really have to do more than 2-3 players per team?

I probably shouldn't have added that, as it didn't clarify anything. My main point is that I had everything dialed up to maximum on a league wide basis, and for every team, and no one was bunting. Dialing up some individual players as well wouldn't cause them to bunt (it isn't additive to the other settings).

If you are suggesting that I leave everyone else at default levels, but dial up some individual players, I could have done that, but they wouldn't have bunted either (as I just said). The only way to get any of the 1960 players to bunt was to raise their ratings (for which I have been heartily admonished and counseled by leading experts

). I understand now that keeping bunt ratings low to curtail bunting was a deliberate game design decision. I know how to get around it; I'm just trying to figure out if I want to, and by how much.

SteveP · 02-13-2009, 01:47 PM

I'm going to post on this one more time, because I've done a further experiment, and I think it's the last one I'm going to do. I think I've done as much as it is reasonably possible to do, given the available tools, to make the overall bunting game more authentic.

Basically, what I did was leave bunt-for-hit ratings where I last set them (with the peak of the distribution in the 90-100 range). I boosted the sac bunt ratings for position players so that the more active sac bunters got ratings similar to pitchers. I have all bunt related settings essentially at default, except that I blocked any type of bunting for run differentials of 4+ (on the team strategy pages). I also had to adjust the LTM for sac bunting for the first season. Here are the results:

1. Position players get about half as many sac bunts as pitchers. It should be the other way around, but I think this is the best one can get, because the game engine is set to make pitchers sac bunt much more often than ratings would dictate.

2. Pitchers sometimes hit away with two strikes, rather than continuing to sac bunt. Pitchers sometimes hit away without attempting to sac bunt at all.

3. I still have never seen the AI call a squeeze play, though hope never dies.

4. There is no cure for the sac bunt with men on 1st & 3rd. The game has it wrong, and we have to live with it.

5. There still seems to be too much sac bunting with one out, but it's less, and I don't really have any statistical basis to say that it's way out of line.

6. Batters in OOTP are incompetent bunters compared to their real life counterparts. That could be something deliberately designed into the game, or just that the game designers never took an interest in this dimension. Because of that, from what I see, any adverse impact on BAs from these rating changes is so small as to be undetectable.

I think that covers it. Before I started working this problem, bunting in OOTP bore no relationship to real-life 1960s baseball. It's now about 50%. But that's enough to add some interest and vitality to individual games.

01-31-2009, 05:52 PM	#1
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	Authentic gameplay: Bunting First the bad news: I’ve found it impossible to get bunting as a whole to work in OOTP in a way that comes at all close to authentic gameplay in my 1960s based league, no matter what I did with options and settings. OOTP appears to handle bunting even less well than pinch-hitting. The good news is that it’s possible to make some noticeable improvements to sac bunting. OOTP appears to have a systemic problem with bunt ratings for historical players, at least in this era. Either that or the game engine uses these ratings badly. I plan to try some experiments with rating adjustments, to see what I can learn. For one thing, the AI never calls bunt-for-hit plays in my test league. This can’t be fixed with any of the settings. The AI also never calls squeeze plays, which may or may not be ratings related. The most important problem with sac bunting is that pitchers SH too often, and position players not nearly enough. In 1960, pitchers accounted for less than 40% of all SH plays. With default settings in OOTP, the proportion is around 90%. This is not only unhistorical, but makes the game too predictable and too boring. IRL, almost every position player executed a sac bunt at some time during the season. Other SH related problems: the AI sac bunts too often with a runner on second, one out; the AI sac bunts with runners on 1st & 3rd (which happens IRL only as a squeeze play); and the AI sac bunts with two strikes too often. The settings that I ended up using do help get SH closer to authentic gameplay, however, they need to be tested for other eras. First, I set the League strategy option for bunting to “Very Often”. This increases the proportion of SH plays by position players significantly -- much closer to historical levels. Doing this requires that overall SH plays be reduced substantially. The only way this can be done effectively is in the League Totals Modifier for sac bunts. For my test league, the best setting seems to be about 25% of normal (i.e., .250 if the normal setting would otherwise be 1.000), but I think that may still be too high. With these adjustments, you can use normal settings for sac bunting in the Team strategy pages (which I like to vary from team to team). One caution: if you use my settings, you may find some instances of sac bunting by position players disconcerting, as the AI tends to be eccentric in its play calling. One thing I’ve done is to modify the team strategy pages for all teams so that sac bunting is set to “Never” when the run differential is 4+ in either direction. This adds some work to maintaining these pages, however, so I don’t know if I’m going to stick with doing that. FWIW

02-01-2009, 06:09 AM	#3
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,301	The League Totals Modifiers will automatically calculate the appropriate value for Sac Bunts each season to keep the game on pace with the historical SAC/AB ratios. Pitchers may rack up more Sac's than you may like, but the league as a whole will be on target. I don't think we should worry too much about this issue right now, as there are other more important things that need some work in historicals.

02-01-2009, 02:53 PM	#4
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	My effort to look at what makes for authentic gameplay is not about season or career stats, except indirectly. It's about how closely individual games played during a season in OOTP compare to how games were really played in that season. Also, I am not posting these to raise issues or complain about game design (unless, of course, Markus decides to take more of an interest in gameplay at this level ). My primary objective is to see what can be done with the available tools to get more authentic gameplay. One of the main themes of this thread is that having pitchers automatically sac bunt with men on base is both substantially unhistorical and boring -- as is position players rarely sac bunting in these situations. The good news is that this can be much improved with the available tools, for people like myself who care about what happens in individual games. I hope readers will regard this and similar threads in that spirit.

02-01-2009, 11:53 PM	#7
Spritze OOTP Historical Czar Join Date: Dec 2001 Location: Bothell Wa Posts: 7,253	I'm a "with you guys" guy too. I play out most games and notice the same things. Numbers are fine and dandy but they ain't "real". Really real is why guys like Joe Ginsburg used to lead the league in intentional walks. It wasn't because of his dangerous bat it was because he always hit eighth in front of pitchers who'd strike out a lot. This sort of thing doesn't track in OOTP.

02-05-2009, 01:34 PM	#8
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	An update on my experimentation in this area. I did a wholesale adjustment of player bunt ratings. Roughly speaking, the original rosters created from the Lahman DB had the peak of the distribution curve for bunt-for-hit ratings somewhere in the 15-20 range (on the 0-250 scale). So nobody bunts. I raised these so that the peak is somewhere in the 90-100 range. I also raised the sac bunt ratings for position players, but only about half as much. There was no science behind this, just trial-and-error. Since I haven't a clue how the original ratings were derived or how they are supposed to work (and, of course, they don't work anyway ), I was just taking my best shot. The point is that the result of this much change is quite striking. Bunting is now truly part of the game, and I can now use the available tools to adjust this to get pretty authentic results. Some people might think it unhistorical to see a Willie Mays or Hank Aaron occasionally lead off an inning with a bunt, but in fact this did happen. I now can get this sort of behavior in the game (along with the AI's somewhat eccentric decision-making of course, but that's unavoidable ). OTOH, I still have yet to see the AI call a squeeze bunt. Has anyone? I suspect it is either not in the AI's repertoire, or it's been given a much higher rarity value than should be the case.

02-05-2009, 07:03 PM	#9
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,301	SteveP - ...and now you have a ton of guys in the league racking up 30-40 bunt hits per year. Give us a list of the .300 hitters in the league now. There was a reason why we made the Bunt-For-Hit ratings the way they are in historicals.

02-06-2009, 03:26 PM	#15
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	I did a little check on the frequency of real life bunt-for-hit plays in 1960 compared to what I was getting in my league. A bit time-consuming, so I could only check 20 games. I may try to keep track of this manually going forward, but having looked at this season in detail a couple of times already, I think what I am looking at is tolerably representative. IRL, there was one bunt-for-hit attempt (from one team or the other) roughly every 2-3 games (this is exactly the sort of metric I like to help me figure out if I'm getting authentic gameplay ). It's difficult to be perfectly precise because you can't always be certain what the intent was related to a bunt with men on base. If you see a player who is known to bunt-for-hit with no one on base, and he bunts with a man on, how do you know he was sac bunting? I assume that OOTP always looks to the sac bunt rating in any sac bunt opportunity, but if so this is probably too constrained to be realistic. I am getting a frequency in my league that is slightly higher than this, but I have been tweaking the relevant settings, so I will see what it looks like going forward. Side note: in the sample from real games I looked at, a position player bunted with a man on 2nd, 2 outs. I counted that as a bunt-for-hit. However, I suspect I will never see a play like that in OOTP, no matter what I might modify, because of the combination of a man on base and 2 outs. Oh, well.

02-06-2009, 06:59 PM	#16
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,301	SteveP - Make a list of your top Bunt For Hit rated guys in the league, and then give us a sim BA vs real BA for each of them. You will see that they consistently overperform by a considerable margin. All those bunt hits become bonus hits on top of the hits that they will normally get. Their bunt hits won't replace line drive singles to CF, they will just be in addition to those hits. We can discuss the reasons why Bunt For Hit is curently the way it is in historicals when we get a beta board for OOTP10. Essentially, it was turning .250 hitters into .350 hitters, over 500 AB that's like 40-50 bunt singles. And there ended up being thousands of these bunt hits in the league every year. That was causing great hitters like Ty Cobb struggle to hit .315, while Joe Schmoes were consistently hitting .340+ each season. There are only so many hits to spread across the league in historicals, and the high Bunt For Hit guys just rack up way too many that they don't deserve.

02-13-2009, 01:47 PM	#20
SteveP Moderator Join Date: Aug 2008 Posts: 3,109	I'm going to post on this one more time, because I've done a further experiment, and I think it's the last one I'm going to do. I think I've done as much as it is reasonably possible to do, given the available tools, to make the overall bunting game more authentic. Basically, what I did was leave bunt-for-hit ratings where I last set them (with the peak of the distribution in the 90-100 range). I boosted the sac bunt ratings for position players so that the more active sac bunters got ratings similar to pitchers. I have all bunt related settings essentially at default, except that I blocked any type of bunting for run differentials of 4+ (on the team strategy pages). I also had to adjust the LTM for sac bunting for the first season. Here are the results: 1. Position players get about half as many sac bunts as pitchers. It should be the other way around, but I think this is the best one can get, because the game engine is set to make pitchers sac bunt much more often than ratings would dictate. 2. Pitchers sometimes hit away with two strikes, rather than continuing to sac bunt. Pitchers sometimes hit away without attempting to sac bunt at all. 3. I still have never seen the AI call a squeeze play, though hope never dies. 4. There is no cure for the sac bunt with men on 1st & 3rd. The game has it wrong, and we have to live with it. 5. There still seems to be too much sac bunting with one out, but it's less, and I don't really have any statistical basis to say that it's way out of line. 6. Batters in OOTP are incompetent bunters compared to their real life counterparts. That could be something deliberately designed into the game, or just that the game designers never took an interest in this dimension. Because of that, from what I see, any adverse impact on BAs from these rating changes is so small as to be undetectable. I think that covers it. Before I started working this problem, bunting in OOTP bore no relationship to real-life 1960s baseball. It's now about 50%. But that's enough to add some interest and vitality to individual games.