Wednesday, April 13, 2005

Line 'em up, sit 'em down

Sports Illustrated always focuses on a minor aspect of a sport during their season preview issues. In their 2005 baseball preview, they chose to look at the art of selecting a lineup. Most of the article reflects the new emphasis on on-base percentage, and chides teams like Washington who bat low OBP guys like Cristian ".303 Lifetime OBP" Guzman in the 2-spot just because he is speedy. Ultimately, the article isn't much of a revelation, because in the analysis of each spot, the author concludes "you need a guy with high OBP." In a perfect world, every player would have a .400 OBP and slug .500 (numbers that would indicate an elite batter). But even the huge clubs like the Yankees and the Red Sox can only afford to populate half of their lineups with such hitters. Small market teams can barely afford one such hitter. The result is that most teams' lineups have a pocket of 3 or 4 good hitters in the middle surrounded by mediocre to poor hitters. Here's how a traditional lineup would look (assuming the team has players that fit each category):

1. Fastest hitter
2. Best contact hitter/bunter
3. Best overall hitter
4. Best power hitter
5. Power hitter
6. Balanced hitter with speed
7. Balanced hitter
8. Contact hitter
9. Pitcher (NL)/ worst hitter

Baseball Prospectus recently published an article that questioned the merits of this traditional style, based on the fact that as you descend one place in the order, you can expect a decrease of 20-30 plate appearances per year. BP reasoned, "Why deny your best overall hitter 40-60 potential plate appearences by batting him third? Move him up to leadoff and give him the chance to get as many extra base hits as possible." BP suggested moving the 4 and 5 hitters up two spots as well, the logic being that a team's power hitters should come to bat more often than a team's singles hitters. Though traditionalists would clamor that the best hitters on the team would now have no one to drive in, that's technically only true in the 1st inning. The reward of 40-60 more plate appearances (which for most great hitters means at least 3 more homeruns) would outweigh the risk of starting with empty bases. The key to making this kind of lineup work is to put the high OBP low SLG guys (in other words, singles hitters) at the bottom of the order. They would still hit ahead of the best hitters, just not until the middle and late innings--a small price to pay for the extra 50 plate appearances for the big boppers.

This type of lineup would be a snap to pull off in the AL because they play fake (DH) baseball--most smart AL teams already have a "second leadoff hitter" in the 9 spot. It would be trickier in the NL because of the pitcher's stranglehold on the 9 spot. On one hand, batting the pitcher 9th makes sense to me--why give a terrible hitter any extra at bats? Realistically, however, unless they pitch a dominant game, starting pitchers only make 2-3 plate appearances. From around the 6th inning on, pinch hitters take the pitcher's spot. With this in mind, BP suggests batting the pitcher 8th, surrounding him with two high OBP low SLG guys. What a concept, eh? Actually, it's already been done before--during Mark McGwire's 70 HR 1998, Cards manager Tony LaRussa often hit the pitcher 8th, so that he could put three high OBP hitters in front of McGwire.

Obviously, combining the "3-4-5 to 1-2-3" and the "1-2 to 7-9" and the "9 to 8" shifts are very drastic. BP didn't really do much speculation, but I will. I honestly think that such a lineup would be wildly successful. Think of the Cardinals lineup--it'd go something like Pujols, Edmonds, Rolen, Walker--that's a lineup that would get you at least a run in the 1st inning more often than not. What starting pitcher doesn't like to start out with a lead? And think of the implications for the end of games as well--how many times has a game ended with the best hitter waiting on deck while some speedy guy who can't hit strikes out? Why not increase the chance that it will be your best hitters at the plate in the bottom of the 9th?

The newfangled lineup also makes sense in this era of baseball, in which the stolen base has become vastly devalued. A guy like Scott Podsednik leads off because he can steal bases. But these days, with home runs flying out at record rates, a stolen base doesn't really gain much. The time when a stolen base is really worth the risk (unless the catcher can't throw), is when there are two outs and a weak hitter is at the plate. Hmm...sounds like the perfect place for Podsednik to hit is in the 7 spot in front of the pitcher in the neo-lineup!

Granted, this is all a pipe dream, because it would take one hell of a gutsy manager to bat Manny Ramirez leadoff. It'd be even more difficult to get a guy to accept batting below the pitcher, especially if it were a leadoff man who's used to getting around 700 plate appearances per year. As much as statsheads treat ballplayers like number lines, they are human beings with multi-million dollar egos. As well as this would seem to work on paper, I have a feeling enough agents would tie up front office phone lines to make any experiment a short one. Oh well. I can always experiment in MVP 2005.


Anonymous Sean O said...

The only real comment I can make is something interesting I read on SoSH about a month ago when everyone was debating whether Renteria should be batting second versus lefties, since then his speed would largely be wasted, as he'd never steal with Manny and Ortizzle coming up right after. But some guy had done like a 1000 season sim for two different lineups that were completely different, and the end stats for each were almost identical. It's very counter-intuitive that the order wouldn't make any sense, so I'm not really sure what to make of it.

But it was something interesting to read, and I guess it's not completely illogical that it could be true.

8:51 AM  
Blogger tylernu said...

I did neglect to mention that batting order has virtually no effect on rate statistics like OBP and SLG. The only thing batting order changes are the counting stats that come with more plate appearances.

Also remember that speed isn't necessarily wasted when a runner doesn't steal. Going from 1st to 3rd on a single with less than 2 outs is more valuable than stealing second. Stealing is also of no use unless you don't get thrown out a lot. Renteria'scareer SB% is 72%, which isn't terrible, but less than ideal. Guys with 77% or greater should be the only who attempt to steal often, and even then, they should only try it with 2 outs. Renteria's never been a good on base guy--take out his peak year in 2003, and his career OBP would be much less than .346--so I would say he's best suited batting around 7th or 8th against righties (in a traditional lineup). His career OBP against lefties is .380, so it makes sense to bat him 2nd against them.

10:21 AM  
Blogger Dan said...

I've always thought long and hard about why a baseball lineup is like it is. But to me it really makes sense. Managers play to win on a game-to-game basis. Getting a guy 40-60 more at-bats per season, as opposed to each game having a better chance of having guys on base is worth the risk. Wouldn't you rather have Pudge Rodriguez or Adam Dunn batting with one guy on in the first than nobody on in the first? And this holds out game in and game out. This is especially true for guys in the 3 spot. They typically get on base, but don't hit homers. They still come up in the first no matter what, so why not have them up with guys on in front of them, and hopefully lead to a bigger inning??

I dunno. I mean, i understand what they are saying. But if it truly would help out to change the way a batting order works, i feel like managers would have changed by now! Tradition isn't THAT strong...

Oh, i now have a blog on this site. Not sure if i'm gonna use it yet. It'll only be for sports if i do. But yeah, i'll let you know if i use it.

5:45 PM  
Blogger tylernu said...

I definitely understand your point of view, Dan. I think the real criticism that SI and BP had of managers is that they bat bad hitters 1st and 2nd just because they have speed. It seems to me like speed matters more at the bottom of the lineup, because fewer home runs will be hit down there, meaning it's more important to be able to take the extra base.

As for the 3-4-5 shift, I would argue that it IS a game-to-game thing. The statistic I cited was "40-60 more PAs per year," but on a game-to-game basis, that would mean 5 AB instead of 4, or 4 instead of 3 in a close game. The other key to the 3-4-5 shift is that the heart of the lineup is guaranteed to start an inning with 0 outs. In a traditional lineup, more often than not the 3 hitter comes up with at least 1 out. If the 5 hitter comes up in the 1st, there's almost certain to be be at least 1 out and probably 2. You know as well as I do, the fewer outs the better. So I think that if 3-4-5 started the game, starting with 0 outs would make up for the fact that no one is on base.

Anyway, if you start writing in your blog, make sure to post the link on here, and I'll put it on my little links sidebar.

10:38 AM  

Post a Comment

<< Home