THE BOOK--Playing The Percentages In Baseball

Well, not quite a DP.

I know, it’s a riot.

Suppose someone goes 6-6 (he’s hot!), and then he goes 2-14 (he’s cold!).... and then he’s benched for a game, and then you look and he was 8-20 (6-6 plus 2-14), and say… uh, why did we bench him… he was hitting .400.

The entire process is lunacy.

Now apply that to batter/pitcher matchups… “I had to play him… he was hitting .400 in 20 at bats!” .... “I had to sit him.. he was hitting 2 for his last 14 against him”.

Ideally, managers are never given any split numbers.  I blame Earl Weaver.

Poor Earl.

Well, this post-season has shattered my naive belief that the Yankees had joined the ranks of the smarter, sabremetrically-inclined teams.  Yikes. 

Molina over Posada was the first warning sign.  I understand one can defend it, but it’s still a highly questionable decision.

But man, benching Swisher for Hairston is an EPIC FAIL.

And it’s become clear that Girardi thinks he gets stats but really just does the same dumb matchup stuff Torre did (oh, Enrique Wilson got some hits against Pedro!  Let’s start him!).

MGL, you do realize that if it is a “ridiculable offense” (made-up word!) to tout a player’s hot streak by citing a single game’s great performance (e.g. he’s been hot, and had three more hits tonight!!!), then it must be just as “ridiculable” to mock the hot streak touter with a single game’s lousy performance - right?

Somewhere on a site like insidethebook.net or something, the bizarro-MGL is tearing you a new one as we speak!

MGL:  Just in case you didn’t realize, I’m 99% with you on this, I’m just pointing out that your other, long-form argument is much more persuasive and valid than this sort of zinger (fun as it is!)

Rob in CT - I don’t know, of course, but could Girardi be pulling a Bill Belichick with Swisher, concealing an injury with some mumbo-jumbo about matchups and the like?  There would be a very strong incentive to having the Phillies think Swisher is healthy and available, rather than either not available, or available but hobbled.  Some late inning strategies might be influenced quite a bit by whether the Phils think he’s ready to pinch-hit…

Greg, you are right!  I am the pot calling the kettle black!

understanding that short term statistically history should not be given much if any weight, would you have a problem with a manager (like Girardi) benching a player (like Swisher) due to a recent slump and justifying it by saying something like: “His swing is all messed up, he can’t get on top of the ball, he’s getting fooled by pitches he usually hammers, and we can’t afford to give him so many ABs until our hitting coach can work with him to correct his mechanics and approach”? 
In this situation, the manager is not supporting his decision with bogus statistical evidence, but rather with his own personal observation of a visible problem with the player’s performance.
Would that be a more defensible excuse for a manager such as Girardi to bench someone like Swisher, in your eyes?

#8 bowie: I’m pretty sure the answer here is “yes, if that’s the case.” But the likelihood of it being the case is not so much, and its far more likely there’s some sort of confirmation bias of this “mechanical issue” after seeing the “slump” numbers.

“His swing is way off. Look, the guy’s 0 for his last 12, can’t you see it?” That’s the sort of thinking that may be going into a reasoning like that. Either Tango or MGL said something to this effect in an earlier thread.

Michael: I agree about confirmation bias being the more likely culprit for this rationale if/when it happens.

However, I think we should recognize that not all perceived slumps are just illusions created by normal statistical fluctuation. Sometimes a guy has truly, if temporarily, suffered a decline in his hitting ability due to a correctable problem (i.e., not due to age or infirmity) such as a mechanical flaw or picking up a bad habit or not reviewing video of pitchers he’s going to face, etc.
In these cases, I can see the basis for benching someone who needs an intervention of sorts to get his swing and his approach fixed.

Ditto what Michael said.  I don’t trust manager judgments like that for obvious reasons. Plus, as I also have said many times, when a player who has not changed his true talent at all is hitting well because of the many random (pseudo-random actually) things that can cause a player’s results to be positive, of course he looks like his mechanics are perfect, he is confident, he is seeing the ball well, etc., and when a guy is 0 for his last 15, he LOOKS like his mechanics are all messed up even if his true talent is exactly the same, etc.

Plus even if the observation is true, how do we or the manager know that sitting him for a game is going to change anything?

Again, you or a manager can have an opinion about anything you want, and it may or may not be true.  But without any evidence, the default view has to be that it is not true.  Even things that we consider obvious are based on evidence and data.

mlg--Most managerial discourse is, uh, baloney:  I get that.  Small sample sizes:  I get that. 

But suppose you look at a guy and he hasn’t hit a ball hard in 5 games, has made contact very few times.  He’s faced generally unimpressive pitchers; he’s repeatedly failed to hit predictable pitches in hitter’s counts.  This may not be predicative of anything, over a large population:  because guys turn it around, from good to bad or vice versa, very unpredictably.  But why do we need to think that nothing is wrong with that guy? 

Do you simply believe that we discover if guys can hit major league pitching over large samples and there are no external factors currently susceptible to identification that govern smaller sample size oddities?

nick

“Do you simply believe that we discover if guys can hit major league pitching over large samples and there are no external factors currently susceptible to identification that govern smaller sample size oddities?”

It’s not that their are no factors that are responsible for small sample size oddities, it’s that those factors offer no predictive value.  Obviously, when a player goes 0-4, he probably swung at a few bad pitchers and missed a couple of hittable ones.  However, a player having a bad game or a bad stretch of games doesn’t have any bearing on how he will play the next game (unless it is something constant like an injury).

What MGL appears to be saying is that there is no use for a hitting coach. If not, please tell us what the use is.

#14, Yeah, that’s exactly what I am saying.

Of course there’s a use for a hitting coach.  Dumbo had a use for the white feather.

What teams need to have is someone who will do scouting without knowing the statistics. Keith Law said in a recent chat that he ignores any stats in Arizona Fall League play and focuses just on observing the players. Of course, if you’re watching an at-bat, it’s hard not to view the results as well, whether it’s a roped double into the gap or a strikeout. But if these scouts do indeed just watch the game for mechanics, motions, and other observations, without regards to the recent stats, then that’s valuable information.

But if you look at a streak of 4 for 40 and THEN point out that his mechanics are off, it’s hard to believe that you figured that out without knowing the 4 for 40.

I put this in the Granderson thread, but this looks like a better spot for it:

http://www.insidethebook.com/ee/index.php/site/comments/curtis_grandersons_take_on_hitters_being_hot_or_cold/#29

The meat of my comment was this:

If hot and cold streaks do not exist, or are not predictive, why would a manager ever take a pitcher like Roy Halladay, Johan Santana, Tim Lincecum or C.C. Sabathia out of the game before they hit 100 pitches?

Before you answer, remember, if you believe hot and cold streaks don’t exist, you believe past performance, ESPECIALLY recent, small-sample-size past performance, is meaningless.  And performance in the very game you are making a decision in is as recent, and as small-sample-size as it gets, so this should be the *textbook* example of what you believe should be severely discounted or ignored…