Pitcher Pinch Hitting Needs to be Tweaked.

markmcghee · 03-21-2016, 11:56 AM

I'm seeing a lot of pitchers batting in the 7th, 8th, and 9th innings when his team is losing. Pinch hitting needs to be tweaked.

bwburke94 · 03-21-2016, 05:43 PM

Not again...

Anyway, there are two main reasons for this: either there are no position players left on the bench, or the AI wants to keep the pitcher in for bullpen rest reasons. Considering the change to pitcher rest made in 17.2, it's important to ask whether you've patched?

markmcghee · 03-21-2016, 06:12 PM

Yes, I'm running 17.2. Do I need to create new league for the change in pitcher rest to take effect? I see bad pinch hitting in a 1938 league and a 1970 league both created under 17.1.

Matt Arnold · 03-21-2016, 06:17 PM

No, in-game logic always applies to games in progress.

Be aware that older leagues tend to have lower league settings for "pinch hit for pitchers", so they will often stay in much later than they do in the modern game.

RchW · 03-21-2016, 06:40 PM

Quote:

Originally Posted by markmcghee

Yes, I'm running 17.2. Do I need to create new league for the change in pitcher rest to take effect? I see bad pinch hitting in a 1938 league and a 1970 league both created under 17.1.

Define "bad". Too much? Too little? They will be different for sure.

markmcghee · 03-21-2016, 07:14 PM

By bad pinch hitting I mean situations in which the AI's team is down at least a run in the 8th or later, and the pitcher comes to bat and the AI decides to let the pitcher hit.

Sorry for the vagueness. I was talking about too little late-inning pinch hitting for pitchers when the pitcher's team is losing. I will see if I can screen-capture a couple game logs.

I have 3 candidate screen captures illustrating poor pinch hitting decisions by AI.

In the first one Kansas City starter Dick Drago leads off the 8th trailing Oakland 3-1. 100% of the time a pinch hitter should be used in this instance.

In #2 Houston starter Don Wilson leads off the 8th with Houston trailing Pittsburgh 3-0. A pinch hitter should be used 100% of the time in this instance.

#3 is a little bit less obvious. Philadelphia A's starter George Caster bats with two outs nobody on in the 7th trailing Washington 2-0. I would always pinch hit here. There might be a case to be made that this is a 30's era game so therefore Caster might have batted here. I'm pretty sure the most managers would have pinch hit.

It is easy to spot these instances of no-pinchhitting. I simply look for boxscores in which the losing starter pitched the entire game. In almost all of these games the AI should have pinch hit for the pitcher at some point in the 7th, 8th, or 9th.

RchW · 03-21-2016, 07:55 PM

Quote:

Originally Posted by markmcghee

By bad pinch hitting I mean situations in which the AI's team is down at least a run in the 8th or later, and the pitcher comes to bat and the AI decides to let the pitcher hit.

Sorry for the vagueness. I was talking about too little late-inning pinch hitting for pitchers when the pitcher's team is losing. I will see if I can screen-capture a couple game logs.

In 1938 it may be normal for a pitcher to bat in the 8th down one run. In 1970 it may be very dependent on the pitcher. Jenkins, Seaver, Hunter et al might have batted often.

FYI PH plate appearances (calculated from AB) in 1938 were at a rate of 0.56/game. That's just over 1PH PA every 2 games. You should be overjoyed that PH are low. That means the game is working as designed.

in 1970 PH PA were higher approx. 1.44/G.

markmcghee · 03-21-2016, 08:04 PM

I'm all for PH rates matching the era. By and large OOTP17 nails it. It is just that, during some critical situations, the AI is ruining a ton of potential late inning rallies by allowing his pitcher to bat when losing, often times leading off the key late inning. If it can be tweaked, great. If not, no biggy. If you look at my screen shots in my last post I think you will get what I am talking about. Like I said, it's not about PH rates. It's about the AI under-appreciating key pinch hitting situations.

pstrickert · 03-21-2016, 08:08 PM

Look at the team strategy settings. The same strategy applies to being plus or minus one, two, or three runs in the late innings. In reality, these situations call for different strategies, obviously. But one size fits all in OOTP. I don't know if this affects AI decisions. Perhaps not. But it'd be nice if Markus or Matt could verify.

RchW · 03-21-2016, 09:01 PM

Quote:

Originally Posted by markmcghee

I'm all for PH rates matching the era. By and large OOTP17 nails it. It is just that, during some critical situations, the AI is ruining a ton of potential late inning rallies by allowing his pitcher to bat when losing, often times leading off the key late inning. If it can be tweaked, great. If not, no biggy. If you look at my screen shots in my last post I think you will get what I am talking about. Like I said, it's not about PH rates. It's about the AI under-appreciating key pinch hitting situations.

Are you saying that the strategic imperative to pinch hit in certain base/out states and scores is the same through all baseball era's. I don't think that has been the case.

markmcghee · 03-22-2016, 06:58 AM

I'm saying that 100% of big league managers past, present, and future would pinch hit for Don Wilson leading off the 8th down by three runs as Houston was in my screenshot. That's a no-brainer, regardless of era. I'm seeing quite a few instances in which the losing pitcher finishes the game out never having been pinch hit for in the late innings. It would be nice to see the AI get the no-brainers right. And like I said, if it can be tweaked fine. If it's not possible, no biggy, it's still a great game. The best yet.

wolverinebball11 · 03-22-2016, 08:20 AM

In modern day baseball, you're absolutely right. During the 2015 season, pitchers only threw a complete game in games their team lost 6 times. (I'm only counting non-DH games here). Of those six, the pitcher never hit after the 6th inning twice and allowed the game winning run in his final inning three times. The sixth game was rain shortened so a "complete game" was actually only 5 1/3 innings. So yeah, in 2015 a manager is going to pinch hit in those scenarios 100% of the time.

However, historically a scenario like this isn't as unusual as you might think. In 1970, a pitcher threw a complete game in a game his team lost 110 times. Only seven of those 110 were games decided by more than three runs. In 1938, it happened a whopping 178 times! Of those 178, only 27 were decided by more than three runs.

Moral of the story here is that things like this were a lot more common back then.

Markus Heinsohn · 03-22-2016, 08:41 AM

Quote:

Originally Posted by wolverinebball11

In modern day baseball, you're absolutely right. During the 2015 season, pitchers only threw a complete game in games their team lost 6 times. (I'm only counting non-DH games here). Of those six, the pitcher never hit after the 6th inning twice and allowed the game winning run in his final inning three times. The sixth game was rain shortened so a "complete game" was actually only 5 1/3 innings. So yeah, in 2015 a manager is going to pinch hit in those scenarios 100% of the time.

However, historically a scenario like this isn't as unusual as you might think. In 1970, a pitcher threw a complete game in a game his team lost 110 times. Only seven of those 110 were games decided by more than three runs. In 1938, it happened a whopping 178 times! Of those 178, only 27 were decided by more than three runs.

Moral of the story here is that things like this were a lot more common back then.

This.

It all depends on the league strategy settings. If someone is not satisfied with the defaults, he can always tweak the PH settings.

markmcghee · 03-22-2016, 08:43 AM

Wow! Ok, then. I stand corrected. I guess what's a no-brainer today was not always a no-brainer no matter how no-brainy it seems. I was concerned that I had a boat load of those types of situations in the first few weeks of the 38 and the 70 season. Probably another case of small sample size.

03-21-2016, 11:56 AM	#1
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	Pitcher Pinch Hitting Needs to be Tweaked. I'm seeing a lot of pitchers batting in the 7th, 8th, and 9th innings when his team is losing. Pinch hitting needs to be tweaked.

03-21-2016, 07:14 PM	#6
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	By bad pinch hitting I mean situations in which the AI's team is down at least a run in the 8th or later, and the pitcher comes to bat and the AI decides to let the pitcher hit. Sorry for the vagueness. I was talking about too little late-inning pinch hitting for pitchers when the pitcher's team is losing. I will see if I can screen-capture a couple game logs. I have 3 candidate screen captures illustrating poor pinch hitting decisions by AI. In the first one Kansas City starter Dick Drago leads off the 8th trailing Oakland 3-1. 100% of the time a pinch hitter should be used in this instance. In #2 Houston starter Don Wilson leads off the 8th with Houston trailing Pittsburgh 3-0. A pinch hitter should be used 100% of the time in this instance. #3 is a little bit less obvious. Philadelphia A's starter George Caster bats with two outs nobody on in the 7th trailing Washington 2-0. I would always pinch hit here. There might be a case to be made that this is a 30's era game so therefore Caster might have batted here. I'm pretty sure the most managers would have pinch hit. It is easy to spot these instances of no-pinchhitting. I simply look for boxscores in which the losing starter pitched the entire game. In almost all of these games the AI should have pinch hit for the pitcher at some point in the 7th, 8th, or 9th. Attached Images Last edited by markmcghee; 03-21-2016 at 07:56 PM.

03-21-2016, 08:04 PM	#8
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	I'm all for PH rates matching the era. By and large OOTP17 nails it. It is just that, during some critical situations, the AI is ruining a ton of potential late inning rallies by allowing his pitcher to bat when losing, often times leading off the key late inning. If it can be tweaked, great. If not, no biggy. If you look at my screen shots in my last post I think you will get what I am talking about. Like I said, it's not about PH rates. It's about the AI under-appreciating key pinch hitting situations. Last edited by markmcghee; 03-21-2016 at 08:08 PM.

03-21-2016, 05:43 PM	#2
bwburke94 Hall Of Famer Join Date: Jun 2008 Location: Belchertown, MA, USA Posts: 4,522	Not again... Anyway, there are two main reasons for this: either there are no position players left on the bench, or the AI wants to keep the pitcher in for bullpen rest reasons. Considering the change to pitcher rest made in 17.2, it's important to ask whether you've patched?

03-21-2016, 06:12 PM	#3
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	Yes, I'm running 17.2. Do I need to create new league for the change in pitcher rest to take effect? I see bad pinch hitting in a 1938 league and a 1970 league both created under 17.1.

03-21-2016, 06:17 PM	#4
Matt Arnold OOTP Developer Join Date: Jun 2009 Location: Here and there Posts: 16,228	No, in-game logic always applies to games in progress. Be aware that older leagues tend to have lower league settings for "pinch hit for pitchers", so they will often stay in much later than they do in the modern game.

03-21-2016, 08:08 PM	#9
pstrickert Hall Of Famer Join Date: Dec 2005 Posts: 16,997	Look at the team strategy settings. The same strategy applies to being plus or minus one, two, or three runs in the late innings. In reality, these situations call for different strategies, obviously. But one size fits all in OOTP. I don't know if this affects AI decisions. Perhaps not. But it'd be nice if Markus or Matt could verify.

03-22-2016, 06:58 AM	#11
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	I'm saying that 100% of big league managers past, present, and future would pinch hit for Don Wilson leading off the 8th down by three runs as Houston was in my screenshot. That's a no-brainer, regardless of era. I'm seeing quite a few instances in which the losing pitcher finishes the game out never having been pinch hit for in the late innings. It would be nice to see the AI get the no-brainers right. And like I said, if it can be tweaked fine. If it's not possible, no biggy, it's still a great game. The best yet.

03-22-2016, 08:20 AM	#12
wolverinebball11 Minors (Rookie Ball) Join Date: Mar 2016 Posts: 39	In modern day baseball, you're absolutely right. During the 2015 season, pitchers only threw a complete game in games their team lost 6 times. (I'm only counting non-DH games here). Of those six, the pitcher never hit after the 6th inning twice and allowed the game winning run in his final inning three times. The sixth game was rain shortened so a "complete game" was actually only 5 1/3 innings. So yeah, in 2015 a manager is going to pinch hit in those scenarios 100% of the time. However, historically a scenario like this isn't as unusual as you might think. In 1970, a pitcher threw a complete game in a game his team lost 110 times. Only seven of those 110 were games decided by more than three runs. In 1938, it happened a whopping 178 times! Of those 178, only 27 were decided by more than three runs. Moral of the story here is that things like this were a lot more common back then.

03-22-2016, 08:43 AM	#14
markmcghee Minors (Double A) Join Date: Dec 2003 Location: Ft.Lauderdale Posts: 151	Wow! Ok, then. I stand corrected. I guess what's a no-brainer today was not always a no-brainer no matter how no-brainy it seems. I was concerned that I had a boat load of those types of situations in the first few weeks of the 38 and the 70 season. Probably another case of small sample size.