THE BOOK--Playing The Percentages In Baseball

Somewhat related, we had an umpire blog last year:
http://www.insidethebook.com/ee/index.php/site/comments/do_umpires_have_their_own_strike_zone/

Certainly umpire strike zones are a “skill.” You only have to watch a few games to realize that (in fact, you would probably be better off watching games than looking at ball/strike data - kind of like speed - you only need to watch a guy a couple of times to figure out how fast he is).  I question whether the other things - in fact most things - in the articles I am referring to, are.  And in any case, as I stated, you can’t just present the data as if it “means something” without doing the analysis that Tango did in his above-referenced thread.

For some reason, I get much more upset when an article on The Hardball Times would make a faux pas than at BP.  For the most but not all part, Hardball Times is sabermetrics, while BP is intelligent mainstream.  Maybe I measure my annoyance level by my expectations.

I agree, but still there are plenty of hard core sabermetricians at BP.  I guess you can’t have those few individuals review every single article that goes up there.  By and large I am a fan of BP as well as THT.

I agree.  I’m a fan of BP and THT, as well as Mike and the Mad Dog at WFAN.  It’s all tempered by my expectation levels. 

For example, with Dan Fox, I’d hold him to very high sabermetric standards.  For Joe Sheehan, I’d hold him to high intelligent standards.  They each deliver.

I would have liked that Sheehan would have used Leverage Index in today’s article and tackled things from a sabermetric viewpoint:
http://www.baseballprospectus.com/article.php?articleid=6421&mode=print&nocache=1183669601

For example, if you go here:
http://www.fangraphs.com/winss.aspx?team=Indians&season=2007

You will see that Betancourt is by far their best reliever, and his LI is a healthy 1.43, which is a shade above a standard setup reliever.  Based on his identified role, he’s being used as expected.  Borowski is having an ok year, and his LI is a high 2.11, meaning he’s really been used in high leverage situations, as that level is around the typical league leader.  The guy that Joe had a problem with, Mastny, has a seasonal LI of 1.15, which is just fine.  He’s been pretty good with RISP (for whatever reason).  Rafael Perez and Fernando Cabrera looks like they should be used in higher leverage situations, but between them, they have 10 HR allowed on 44 hits.

So, for what Joe did, he delivered the goods.  It wasn’t a sabermetric evaluation, though.

Betancourt is one of the better relievers in baseball and should be used as a closer.  I have heard the Indians announcers on TV say that the Indians know he is very good, but they just don’t think he can handle the closer role.  Who knows?  The Indians are one of the smarter organizations.  I would certainly give him a chance, and a lengthy one at that.  Then again, 1.43 is nothing to sneeze at for a great reliever.  The only problem is that Borowski is one of the worst closers in baseball, so that to swap them would help the Indians a lot.  In fact, if I were the Indians, I would let Borowski rack up a ton of saves, which any closer would do for that team, and then get something of value for him and let Betancourt close.

I concur with MGL on the intelligence of the Indians organization.  They’ve been sabermetrically-inclined for years, and have top guys working there.

1.43 is a fairly high LI, so they’re definitely letting Betancourt rack up pressure innings.

Todd Jones is leading the league at 2.36, followed by Hoffman at 2.33.  Borowski is 6th, and Betancourt is 49th (among relievers with at least 10 relief appearances). 

Among non-closers with high LI:
Brandon Lyon, 2.1 (2 saves)
Zumaya, 1.9 (1)
Duscherer, 1.9 (0)
Turnbow, 1.7 (1)
Linebrink, 1.7 (1)
Morrow, 1.6 (0)

As you can see, it’s possible to get your setup guys lots of leverage situations, without closing the game.

Closers being coddled (LI less than Betancourt):
Mariano Rivera, 1.24 (10 saves)
Eric Gagne, 1.35 (11 saves)
Billy Wagner, 1.37 (16 saves)
Al Reyes, 1.42 (17 saves)

Yanks: the relievers LI is 0.92.  But even so, Joe’s either doing a terrible job with Mo, or Mo has been such a disaster (by his standards) that Joe doesn’t trust him as much.

Mets LI is only 0.85, so there’s not much room to use Wagner.  Still Heilman is second at 1.23, so Willie is giving Aaron lots of pressure situations that could go to Billy.

Tampa: what’s going on here?  Shawn Camp has an LI of 1.35, and the team LI is 0.97.  How does Reyes already have 6 HR allowed?

Rangers: Otsuka has a higher LI than Gagne (1.65).  And Benoit is at 1.30.  Gagne is getting the glory, but the other two are used in just as tough situations, and performing just as well.  Otsuka is the guy I’d go after.  Dude’s in the shadows of Gagne and Hoffman?  I know he’s 35, but check out his career (2004-2007) stats: 232 IP, 217 K, 65 NIBB, 12 HR on 190 hits. 

Scouts should be all over this guy.

I thought it was a good article by Joe.  Nothing earth shattering and certainly nothing that hasn’t been said a zillion times already.

I don’t think that managers misuse closers because of the save rule or because closers would get pissed off if they weren’t given the maximum number of save opportunities (presumabaly to increase their monetary value).  I think that is a myth.  I think it is just that managers do not understand leverage (at least the nuances), and they overvalue their perception of “pressure” in the 9th inning.  They truly think that protecting a 2-run lead in the 9th is more important than a tie game in the 8th, for example.  And there is risk aversion as usual that comes into play.  The manager would be crucified if he did not have his closer pitch the 9th with a 2 or 3 run lead and his team loses (which will happen 10% or whatever of the time).  If he does not bring in his closer in the 8th or in a tie game in extra innings, or some other non-save situation, no one but a sabermetric-type cares.  Managers get tricked into believing that what makes them look good is the right thing to do.  Regular people get tricked into that too.

So, to recap, we have 3 things that work to keep a manager from using his closer optimally.  One, his ignorance of leverage (including overvaluing the importance of the 9th inning).  Two, his perception that only his closer can handle the pressure of preserving a 9th inning lead.  And three, his desrire, mostly subconscious, to avoid criticism.

Sure, the save rule sort of helps him to use a certain set strategy which is clearly not optimal.  But NOT because closers want to rack up saves.  Again, at least not consciously by managers.  I don’t think many managers say to themsleves, “I would really like to use my closer in the 8th or not use him in the 9th with a 3 run lead, but he is going to be really mad at me, so I won’t.”

#7, RE: Mariano…

There have been several games this season where the Yankees lost in the ninth or extra innings with Rivera never entering the game. Most notably, they lost in 13(!) innings at San Francisco without ever calling on Rivera (who was used once in the previous week to boot).

Torre simply will not use him in a tie game on the road. It arguably cost them the 2003 World Series (turning to Jeff freaking Weaver in lieu of Rivera in a tie game) and he hasn’t learned since. I’ll never understand that reluctance.

Mo’s LI, aside from this year, has been pretty good (1.82 to 2.23 from 2002-06).  His setup guys (Gordon, Nelson, Karsay) have been around 1.4.  Joe did a good job.  But not this year. 

I think his lack, or peceived lack, of a quality setup guy has totally flummoxed Joe, especially since he said he’d be using Mo as 1-inning reliever because of his health and age.  (Dude is almost 38 and has 32 K in 32 IP, with only 5 walks.  But his BABIP, a career under .280, which is fabulous, is a career high .332 this year.)

***

MGL: I know someone who talked to a big-league manager, and that manager told him that his guys told the manager that they like their set routine by innings.  They prepare themselves by time, not situation.  And that some guys don’t want the extra pressure of the ninth inning.  So, you really have human attitudes among the pitchers to consider.

I agree that a manager does not like to pull a better reliever for a worse one.  So, if you have the heart of the order in the 7th inning of a close game, with men on base, if he brings in his top reliever, he’ll expect that reliever to finish the game (ala old-Gossage). 

It really isn’t worth the aggravation to the manager to be second-guessed by armchair coaches who expect perfection.

The only change I would make is for the manager to stop using the ace with a 3-run lead in the 9th (especially if the heart of the order is not due up), in favor of more 5 out games.

I agree that there are all kinds of considerations at work, including what the players want and like.  That does NOT mean of course that managers cannot use their relievers and especially their closers more effectively than they presently do.

Yes, if I had to dicate to a manager what to do and not to do, very generally, and without too much opportunity to be criticised, it would be to not use the closer with a 3 or more run lead unless they need the work, and to use him more often in non-save situations, like in the 8th innings and tied games in the 9th or later.  When you bring in your closer in the 8th, you can play it by ear whether he finishes the game or not, depending on the score in the 9th, how many pitches he has thrown, who is coming up, home or away, how much work he has had lately, and who else you have available in the pen.

If I could dictate to a manager what to do.........dang, just think of the fun you could have, playing chess with all those pieces.

But since we’re talking about relievers, I wouldn’t even have a closer. You could use your relief pitchers in the manner that gives you the best chances to win games, not by what role they’re supposed to have. And it sure would save the team a good chunk of change, too.

Silly the way things have evolved in the bullpen, no?