<$BlogRSDURL$>

Thursday, March 03, 2005

4-3 APOLOGETICS
Today the LA Times prints its annual article on the subject of Darin Erstad being valuable even though he doesn't put up good statistics.

Some choice quotes:

Erstad: "I'm the first to admit, I feel I've underachieved the last few years. My power numbers haven't been where they should be. I should hit 20 home runs and 35 doubles every year. But I'm not going to jeopardize this team for the benefit of personal statistics."

And: "I know my power numbers are not on par [with other first basemen], but making productive outs is more important to me. The 'Moneyball' approach is a different philosophy, a strong philosophy. I don't walk a ton, and my on-base percentage isn't as high as it should be. But I also roll about 30 ground balls a year to second base, getting runners to third."

(All emphases are mine.)

This is a curious line of thinking, to say the least. There is somehow this notion that racking up gaudy OBP and SLG numbers is a selfish act that can hurt the team; nothing could be further from the truth.

In other sports, particularly sports like basketball or metric football, a selfish player can damage the team by hogging the ball and scoring opportunities for himself. But in baseball, the most individual of team sports, the team does not suffer for one man driving in runs. Everyone benefits.

Darin Erstad literally thinks it's his job, if he comes up with a man on second with no one out, to ground out to the second baseman and advance the runner. That is not what a major league hitter should consider his job. He should be looking to drive the ball and get the run in -- and advancing the runner via an out is a consolation, it's making the best out of the worst alternative.

Let's say you have a runner on second and no one out. In descending order, and neglecting fielder errors, what are the most preferable results from the plate appearance?

1. Home run
2. Triple
3. Double
4. RBI single
5. Single with runner holding at third
6. Walk (or infield single on which runner cannot advance)
7. Single with runner being thrown out at home
8. Making an out that advances the runner to third
9. Making an out that does not advance the runner
10. Lining into a double play

Erstad's sense of "unselfish play" is to skip straight to the eighth-best result. Does that make any sense to you?

Now, sure, I'm sure you can say, "Yes, but if he tries for a home run, it increases his chances of striking out" or something. Yes; but I'm not saying he should go up there looking for a home run. I'm saying he should go up there, looking for a ball to hit hard and preferably pull -- or just to the left of the shortstop. If you do well, you have a hit and probably an RBI -- and that's obviously good for the team. And if you fail and make an out, you have probably done so in a way that can advance the runner, making the most of the out. But the out should not be a concession.

As you might guess I have no major league playing experience, you might suspect I am full of hot air on this. I have discussed this in passing before, so I'll just point you there. (It's the second bullet point.)

In case you're wondering, Erstad really did make 30 "productive outs" last year, which is better than making nonproductive outs. Erstad says, "I could lay down a bunt and sacrifice, and my average would be better." So let's give him the benefit of the doubt and say that each of those 30 productive outs could have been sacrifice hits. What then?

Instead of hitting 295/346/400 for an OPS+ of 95, he would have hit 314/367/426 for an OPS+ of 108, which would have ranked ahead of six regular AL first basemen, whereas his actual 95 ranked ahead of only one. Of course, that's assuming that no other AL 1B would get to benefit from this game. I mean, Scott Hatteberg had 18 "productive outs," so if you give him those 18 plate appearances back he moves from 284/367/420 with an OPS+ of 104 to a 293/378/434 with an OPS+ of 110.

I'm all for thinking of the team first, and being an unselfish teammate. But we shouldn't lose sight of the fact that, in baseball, team success is built on individual achievement. Darin Erstad knows that his OBP and SLG should be higher; I would love to see him do something about it.

UPDATE: I had some interesting comments, but no real room in the comment section to respond, so I'm doing it here.

To respond to The Alchemist:

I think that in almost all situations, 1st & 2nd with no out is better than man on 3rd with one out.

I was looking for a recent Run Expectation Matrix; here's one from 2003, which says that the average team will score 1.0303 runs for every time they have a runner on 3rd with one out, but will score 1.5384 for every time they have 1st and 2nd with no outs.

Tango Tiger has some interesting data for 1999-2002. Take a look:

              % of time team scores X runs
Bases Outs 0 1 2 3 4 5+
1,2 0 57 16 11 9 4 3
3 1 34 48 11 5 2 1
So, getting the runner to third will increase your chance of scoring one run, but cut down on big inning potential. There are some situations where you might want to do that, of course. So you could go ahead and revise my list for those specific situations where one run is particularly significant.

An anonymous commenter wonders how, theoretically, putting Finley at 1B and Erstad at CF would affect things. We can only guess, but it's spring training, so let's have fun and try.

I say Darin Erstad projects to be roughly -10 runs offensively against average this year, and Steve Finley projects to be +2.

Now, we should also adjust for position in some way. I am partial to using Tango Tiger's positional adjustments, derived for UZR. This is by no means universal, but I like them.

Using the adjustments, we find that a full-time CF is +5 runs better defensively than a generic defensive player, and that a 1B is -9 runs. So, with Finley in CF and Erstad at 1B, we have Finley up to +7 with Erstad at -19.

But we have yet to take into account the quality of their defense. Erstad, of course, is a very good defensive 1B. How good? I don't know, but from David Pinto's PMR we can divine that he was about +24 above an average 1B defensively last year, which would project up to about +30 for a full season. Let's keep it conservative and estimate him at +25. That gets him up to +6 overall.

Finley's defense is something of a question, but few statistical methods are saying he's good; the question seems to be "Is he bad or just mediocre?" Using PMR we find that he was roughly -6 a year ago. I know UZR takes a harder line on him, but he was -6 by Baseball Prospectus' figures, as well, so I think that's a reasonable estimate.

So Finley was +7, and we take off 6 for the quality of his defense, so he's at +1.

So Finley and Erstad together are roughly +7 runs above average. What a bargain!

But what if we put Erstad back in CF? His offense is still -10, but now he gets a 5-run boost for being a CF to get him to -5. He's a very, very good CF, and should be worth about 20-25 runs against average defending that position. Let's stay conservative and give him +20; that will get his overall total up to +15.

Finley, we said, was +2 offensively. We take off 9 for playing 1B and get him to -7. The question becomes, "How is his defense?"

If Steve Finley would be an average 1B, moving him doesn't really make any sense, because he'd be at -7, and with Erstad a +15 in center, that gets us to +8, which is virtually identical to what we got with Finley in CF and Erstad at 1B.

If Finley were to be a bad 1B, the move would be counterproductive. Finley would, theoretically, have to be an excellent defensive 1B for the move to balance out.

I wouldn't have advocated contemplating such a move even if this back-of-the-envelope said it should, so the fact that even looking at it in a cursory and nonsophisticated manner indicates that it's not likely to be a great idea ... of course, there many, many big assumptions even in this, especially in the difference between Finley's offense and Erstad's offense, so I would hardly go by this quick analysis being The Truth.

Comments:
I think your order is wrong. Man on second with no outs is definitely a case where advancing the runner to third is more preferable than getting on base without advancing the runner. You are moving from a HIGHLY unlikely double play situation to a fairly likely one (depending on who is up next) plus the possibility of a force at third. In a close game, moving that runner to third with one out is better than setting up the force out at third. Of course, it does depend on closeness of the game, who is on base, who is batting, and who is up next (Vlad, for instance, grounds into double plays at a reasonably high clip). And actually swinging instead of sacrifice bunting makes it so that the defense may not be expecting the sacrifice, which can make it easier to pull off. Erstad probably sacrifices more than he should, but he doesn't care about stats...just winning. And certain situations dictate different strategies for winning, including using or not using productive outs.
 
Can a Finley CF/Erstad 1B combo be compared to a potential Finley 1B/Erstad CF combo? I mean Erstad is going to play one way or the other. Can a comparison be done to net the offensive + defensive combinations of both to see how much worse Erstad at first is? Is it substantial? It'd take a lot of assumptions on the Finley part, I know.
 
Oh, and it ignores Kotchman of course.
 
Thanks for the time on this. Obviously, a better scenario would involve Erstad in center and Kotchman at first with Finley on another team. I guess I was curious about the only potential reality of moving Erstad to center given what the real situation is. Looks like a wash.
 
Post a Comment

This page is powered by Blogger. Isn't yours?