THE BOOK--Playing The Percentages In Baseball

I pinch hit.  The starter will begin seeing the lineup the 3rd time through and there are relievers who would be more effective at that point in a close game.

Pinch hit.  If the guy wasn’t pitching a great game, this would be an easy call ... your pitcher isn’t that much better (if at all) than the bullpen, so why throw away so much WE on the plate appearance just to keep him in the game?

So now the question is ... how much do you change your assessment of your pitchers ability today based on the good performance (6/1 K:BB)?  Probably not much, so stick with the pinch hitter.

My answer: You pinch-hit. I believe MGL has written that the pitcher should practically never hit, much less in a one-run game. Plus, his “true” RA is apparently equal to (or worse than given that he will be facing the lineup a third time) that of the reliever you would be bringing in for the sixth inning. I’d guess the marginal value of pinch-hitting is something like .3 runs in that situation, greater than any difference that would exist between the pithcer’s RA from that point forward and any decent reliever.

Xei’s answer: You let the pitcher suicide squeeze. It’s the most exciting play in baseball, and you have to assume it will take the defense off guard. You need to put pressure on the defense.

I send in a pettition to get the DH in the NL, so we can be done with this tomfoolery once and for all.

Pinch hitting seems so obvious that I am half-expecting mgl to blow my mind explaining why he should stay in.

I would be interested to find out at what point the SP becomes so good that you should keep him in

I would be interested to find out at what point the SP becomes so good that you should keep him in

When he comes up in the top of the 9th with a 5-run lead and is pitching a no-hitter/perfect game.

Unless of course you’re the Chunichi Dragons manager Ochiai during the Japan Series. In which case it’s time to pull your pitcher for your veteran relief ace.

I still can’t get my teeth around that boneheaded decision.

I will also ask, what percentage of major league managers would pinch hit and what percentage would leave the starter in?  Then someone can look it up.

Also, as you would assume, your starter has only thrown around 60-70 pitches.

… unless of course you’re the Chunichi Dragons manager Ochiai during the Japan Series ...

Hey Sal Paradise, can you tell me the exact date when it happened?

MGL - Only one situation exactly like that in the last four years.  PHI at PIT 8/19/2007 Ian Snell pitching for PIT due up with the bases loaded and 2 out in the bottom of the fifth and....is allowed to bat, whereupon he promptly strikes out.  He pitches the next inning where he allows a three run homer to fall behind 4-0. He stays in to pitch a scoreless 7th inning.  In the bottom of the seventh Jim Tracy’s master plan becomes apparent.  The Pirates get 2 on with 2 out with the number 8 hitter coming to the plate.  Having perserved his pinch hitters in the 5th, he now uses Josh Phelps to pinch hit for Corlos Maldonado and Phelps is able to draw a walk loading the bases for ..... Ian Snell.  This time Tracy pinch hits Matt Kata for Snell. Kata doubles in 3 runs. The pirates go on to score 4 more runs with 2 outs, winning the game 8-4 for winning pitcher ...Ian Snell.  Just the way they drew it up in the locker room before the game. The moral is don’t pinch hit for your pitcher in the fifth inning when pinch hitting for him in the 7th is going to net you 7 runs.

To Sal Paradise:
If you’re talking about game five of the 2007 Japan Series, I believe the score was 1-0 when Ochiai pulled Masa and sent in Iwase to finish the perfect game and the Series. If they had been leading by five, I’m sure Ochiai would have let Masa finish the game on his own.

I guess I would wanna know how tired my bullpen is. If you’ve only got 2 to 3 guys available that might be good enough reason to leave the pitcher in… On second thought I can’t really see when that situation would come up unless you played like a 15 inning game the night before.

So definitely pinch hit here. I think maybe 15% of managers would pinch hit in this situation, though I can’t come up with a specific person. I know Pinella wouldn’t. I don’t think he’d pinch hit in the 6th either.

I am a little surprised that that situation came up only once in the last 4 years.  But you can easily lump together a bunch of similar situations.  Same thing but with runners on 2nd and 3rd.  Or 1st and 2nd.  Or even 2nd or 3rd only. Or the 4th inning or top of the 5th. Or top of the 6th.  Or even bottom of the 6th if you want to stretch the envelope a little.  If you use the bottom of the 6th, though, you really have to make sure that the pitcher is pitching a very good game so far and that he has only thrown 75, maybe 80, or fewer pitches, otherwise the alternatives start to become too obvious.

I want to say that NO managers would pinch hit in that situation, but whenever you think that something happens zero percent of the time, and that “something” is not ridiculous, it is usually on the order of 10-15%, as Scott above guesses.

I am a little surprised that no more than one person has brought up the issue of preserving the pen, or saving the pinch hitters, or whether or not the pen was already tired or overused, since all the other questions were riddled with complications, some of them pertinent and some of them not.

That is an interesting situation, Peter.  Tracy is not one of my favorite managers. Considering that I hate most of them, that doesn’t say too much about him, in my eyes at least.

It seems to me that managers have decided to give up the obvious benefit of a pinch hitter in situations like this in order to manage the workloads. They figure, if I can’t leave my starter in for 125 pitches anymore, I’m certainly gonna get my 95-100 pitches if he is pitching well. They figure the overall seasonal impact of such an approach trumps the win impact for this single game.

And you know, I don’t have a problem with that.

WHY don’t you have a problem with that?

I’d leave him in and have the guy on 3rd steal home.  It might be slightly more optimal to pinch hit, but you run into maverick risk if you do that.

D’oh!  My play would likely open me up to even more second-guessing, but if you’re gonna be unconventional, why not go whole hog?

MGL - The reason that this exact situation only occurred once in 4 years (and frankly I am surprised that it occurred at all) is the requirement of having the pitcher bat with the bases loaded and two outs in the bottom of the fifth and 2 out when his team has scored 0 runs can only happen in two ways.  Either it is the end of the second time through the order and the opposing pitcher has pitched the first 4 innings without leaving a man on base, or it is the end of the third time through the order and the opposing pitcher has left 9 men on base in the first 4 innings while having no runs score.  Either of these situations is pretty rare in itself without the further requirement that the visiting team has only scored 1 run.  The PIT game was the first of these two, four perfect innings before the fifth for PHI.

I am not sure where Charlie Manuel ranks on your list of Hall of Shame managers but he should probably get some negative points for leaving Lohse in to pitch to the pinch hitter Kata in the 7th when Lohse had already thrown 98 pitches.

If it’s Micah Owings you’re going to let him hit.

Sure, David.  But you are assuming that he isn’t.

BTW, I like Peter’s retrograde analysis.  Every now and then I think about trying to solve play by plays from boxscores, but never get around to it.

Stealing home, I like it!  Thing is, it’s either your #4, 5, or 6 hitter, and I’d guess that the traditional lineup does not have a decent home-stealer in that lineup position.  But then again, if the pitcher is going from the windup,who knows, maybe dude’s got a 1-in-3 shot at it.

My gut reaction was pinch hit, for sure, but with just 2 hits and 1 BB allowed, maybe letting him stay in the game and utilizing a steal-home attempt or a squeeze attempt is worth a shot. 

Average pitches per PA is 3.8, so my guy has pitched 15 outs, 1 BB, and 2 hits, so 68.4 pitches.  He’s probably another good inning in him, and my top of the lineup will bat at least twice more.

It depends on who’s on my bench and who I can replace him with from the bullpen.  If I have an adequate Long Reliever I’d probably pinch hit, so to if I have a good pinch hitter or a starter on the bench that day i’d also probably pinch hit.  How good is the pitcher at hitting? Is it Micah Owings is it Carlos Zambrano?  There are other variables that need to be accounted for.

I would assume “average” context if MGL is not explicitly stating the context. 

Clearly, he is not asking about Micah Owings, who is the most extreme case of a pitcher at bat, and would only apply to one team in 30.  He’s also not talking about how Albert Pujols should hit in various other questions either.

I feel that it is an obvious pinch hit here.
The pitcher is up in a very high leverage situation (bases loaded, 2 out, we are down 1). 

I would think that many managers would do the same hting, but I dont watch many NL games.

My first thought was “what is the state of my bullpen?” As in: 1) how much talent do I have; and 2) are they reasonably rested, such that I can get 4 innings out of them?  So that’s a two-part first question (heh).  The second question is who do I have on my bench to pinch hit?

Assuming I have a decent ‘pen that hasn’t just been depleted by a 15-inning marathon the night before, and I have a pinch hitter of any worth, the answer is to pinch hit for my 3rd starter who hits like an average pitcher.

If I have a collection of arsonists and my “pinch hitter” is a utility infielder who hits like an average pitcher… maybe not.

Well what is the “average” context? How rested is the bullpen and are any players of the players on the bench injured? Who is the opposing pitcher? What is your pitcher’s pitch count and how much endurance does he actually have?

I’m not trying to nit-picky but I don’t really think there is a fully assumed, “average”, context.

How much is the typical bullpen rested on average?  One game you might need the starter to go 7, another game you can live with the starter going 4.  If the typical bullpen requires a starter go 6 innings in an average situation, and since mgl didn’t say how much the bullpen was rested, then presume that removing the pitcher in the 5th means taxing the bullpen a bit.

Do the same thing with every one of your points. 

Or, specify what you would need for the PH to be in play and what you would need for the pitcher to bat.

Otherwise, what is the point of having a discussion and arguing about managerial moves, if every answer will always be “it depends”.  Yeah, ok, it depends.  Tell me what it would depend on, for you to decide one way, and what it would depend on for you to decide to go the other way.

Pinch hit.  He most likely only has one inning left to pitch anyways.  This is a high leverage situation.  Can’t let the pitcher hit here.  Other factors needed are how good is your bullpen and how rested are they.  But those would have to be extreme answers on one side to change my mind.
vr, Xei

Peter good point about why this situation hardly ever occurs.  I’d like to see how often the pitcher is pinch hit for (and not) in the 4th, 5th, and 6th innings when he has thrown less than 85 pitches, he has allowed 3 runs or less, and the leverage index of the AB is 2 or higher. I am guessing it is less than 25%.

Pinch hit, mostly because the pitcher isn’t as good as his current performance* indicates and because of the higher RE with a pinch hitter.

*To paraphrase MGL from a thread from a while ago (hopefully I remember this right): a manager should feel fortunate that his below-average starter gave him a good performance for 5-6 innings and should pull him before the wheels fall off.

MGL/14: “WHY don’t you have a problem with that?”

I am well aware that my answer won’t meet your standards of evidence-based study, as it’s mostly my observation of the trends in modern baseball. One of the strongest trends is increasing regularity of roles and workload for pitchers. One of the ways this is being achieved is leaving your starter in for the 6 IP/95 or so pitches--unless he is getting shelled. One of the assumptions for this question is an “average rested” bullpen. But one of the ways managers achieve this is not pulling your effective starter early just for a pinch hitting move. So this assumption is not independent of the tendency to make/not make the type of move in question here.

I’m aware of the leverage of using a PH in the given situation, as well as the idea that this “below-average” starter is likely to regress towards his usual performance if left in the game. But I happen to place a significant weight, by observation, on these other considerations which, IMO, have not come close to being quantified.

David you are not the only one who thinks that the bullpen load should be a consideration.  I disagree strongly, as you can replace your #3 starter who is .4 runs worse than an average starter and going into his third time through the lineup (which makes him about .6 runs worse) with a replacement level reliever and lose NOTHING in the pitching department.  And of course you gain a lot in the hitting department.

So where is the strain on the bullpen if you can replace those 1-2 more innings of your starter with an equivalent replacement reliever?  By definition, you have an almost unlimited supply of replacement relievers.  I am not saying that there is NO strain on the bullpen with this strategy as you are limited to 25 roster spots. But, since you can use ANYONE to relieve our starting pitcher AND you gain as much as you can possibly gain from any strategy move by pinch hitting in that spot, this is a no-brainer to me.  An absolute no-brainer.

I have said this before and I will say it again until the cows come home (or the Cubs win the World Series), by far and away the biggest single gain you can make on a team WE for the season is to take your weaker starting pitchers out as soon as possible when a high leveraqe situation comes up and he is due to bat.  I would take my pitcher out in the third inning.  I thought I made this question easy by having it occur in the 5th inning.  Actually, it was apparently easy, as almost everyone agrees that it is an easy pinch hit.  Of course you have ego considerations.  Most of your starting pitchers who were pitching well would expect to hit in that situation. If I were the manager, I would just say, “Sorry guys, orders from the front office!”

I’d like to see how often the pitcher is pinch hit for (and not) in the 4th, 5th, and 6th innings when he has thrown less than 85 pitches, he has allowed 3 runs or less, and the leverage index of the AB is 2 or higher. I am guessing it is less than 25%.

Answer. 2 of 165? in top of the 4th, 2 of 173 in the bottom of the 4th, 6 of 145? top of the 5th, 10 of 72 in the bottom of the 5th, 20 of 64? in the top of the 6th, 19 of 44 in the bottom of the 6th.  The bottom of the inning numbers I am sure of.  I didn’t have my queries fine tuned when I did the top of the innings.  For the top of innings 4 and 5 the number removed is a maximum, some of them may be relief pitchers or pitchers with more than 85 pitches, but there are so few of the removed that it proves your point anyway.  The top of the 6th pitchers removed is correct but the total number of pitchers may include a relief pitcher or pitcher with more than 85 pitches.  So might the totals for the tops of innings 4 and 5.

Managers appear to be much more reluctant to remove a starting pitcher who is pitching well prior to his completing the 5 innings necessary to get credit for the win.  Understandably so, since they might have a pitcher revolt if he did it too often since pitchers’ salaries are still influenced by wins.  The pinch hitter rate in the 6th and the bottom of the 5th was higher than I would have guessed.

Thanks Peter!  Interestingly, the bottom of the 5th was 15%, about what we expected.  You thought that was high, but as you said, the big consideration for the manager when the pitcher is pitching a good game is leaving him in for at least 5 in order to possibly get the win. I am also surprised how high it is in the 6th.  What condition did you use for the score and how well the pitcher was pitching?

If this question had been about the 3rd inning, it wouldve been much harder for me.  After 5, I’m not worried much about bullpen fatigue, and you dont have to worry about the pitcher being upset for not getting a chance at his win.  (In fact, his best chance for a win is to pinch for him here, and get runs, to give him the lead).  While you might say that that shouldnt be a factor, its certainly easier to not have to make a decision that might hurt team morale.

But according to MGL, removing the pitcher is even correct in the 3rd, so thats very interesting!

MGL - I used exactly what you gave in Post #28. Pitcher coming to bat has allowed 3 runs or less and thrown 85 or fewer pitches.  Actually, rechecking Post #28 you had said fewer than 85 pitches, so I added an extra pitch.

The other thing that should be noted is that those numbers are over 4 seasons 2005-2008.  So using bottom of 5th, top of 6th, bottom of 6th numbers the situation only occurs about 3 times a year per team.

My intention was not to point out that “it depends” but rather that the assumed “average situation” isn’t obvious.

Call my crazy but I doubt that most people are going to agree that the following is clearly assumed: the starting pitcher has a pitch count of 85 with an average tolerance of 110 pitches, there are 4 possible bullpen options (only 2 of which have been effective), there 5 able players on the bench (however the backup catcher should be not used as a pitch hitter), there are 4 games until the team’s next game off, they are playing against a 0.500-caliber team, the opposing pitcher is a league average starter, of the 4 available pinch hitters 1 is above average, 1 is average, 2 are below average, 3 are right-handed and 1 is left-handed and this can keep on going on and on and on.

I realize that you probably think this is just an argument for argument’s sake, but it isn’t. Making a generalization about the situation of a game is pretty risky because there are so many factors to consider and most of which are not painfully obvious.

Since you asked, I will have to call you… crazy! 

Yes, that has to be assumed, otherwise we’re going to be here reading a 1000 word post by MGL that shows all of the parameters that we can’t reasonably assume by your standards.

Tonight in the COL/STL game, COL was winning in the top of the 7th and they had the bases loaded and no outs and the pitcher, De La Rosa, up to bat.  This was the perfect storm for a manager (Tracy, whom I have already called an idiot) to make a gigantic error (I think).

De La Rosa is an otherwise mediocre pitcher.  He was pitching well.  His pitch count was low.  His team was leading, albeit by a slim margin.  There were no outs, so that even if De La Rosa made an out, his manager would be able to “save face” because there would be at least one or two more batters in the inning (keep in mind that lots of times a manager’s decision is driven by how HE looks in the eyes of the fans and the media and NOT on whether it helps or hurts the team).

BTW, De La Rosa is a poor-hitting pitcher and was facing a lefty reliever in Reyes (he bats lefty).  I assume that he is around a .120 hitter (maybe worse) or so here with lots of K’s and almost no power.

Of course he is only going to pitch another inning or two at best.

Tracy lets him hit, which I think is a gigantic error.  I think letting him hit costs on the order of .5 runs in expectancy.  You can manage a major league team for 100 years and not make a decision which costs .5 runs.  If you brought in me throwing 55 mph for an inning or two in place of De La Rosa, you would not give up .5 runs (OK, maybe with me on the mound, but you get the idea...).

Can anyone figure out around how much in WE Tracy gives up by letting De La Rosa bat rather than a pinch hitter?

I’m not sure if this has been discussed yet...but what about having a reliever warmed up?  How long does it take to warm up a middle reliever?  Is the commercial break between half-innings enough (assume worst case the pinch hitter makes an out on the 1st pitch).  If it takes longer, then when do you get a guy up in the pen, If you need to get a guy up in the pen a couple of batters before it gets to the pitcher’s spot...then what negatives are there (if any) if you don’t wind up pinchhitting?

I am not 100% sure, but I think that you ALWAYS have enough time to warm up a reliever in between half innings.  Plus, in this case, there were no outs, so you presumably had tons of time.

I figure a minimum of 2 batters, or 3-4 minutes.

I thought the original post mentioned 2 outs