The Least Important Batting Order Ever?
by Ken Arneson
MLB.com’s new A’s beat writer Jane Lee tweeted her suggested A’s lineup today. I found it hard to argue for or against her suggested order. It seems like every player in the lineup is roughly a .280/.335/.410 player, so it didn’t seem to matter much to me what order you put them in.
To test my hypothesis, I ran the 2010 Marcel Projections through David Pinto’s Lineup Analysis Tool. I don’t think the tool produces particularly realistic or accurate results, even though I had a little hand in developing it. But if it’s useful for something, it’s getting an estimate on the theoretical size difference between the best and worst possible lineups.
When I’ve run this before on potential A’s lineups, the difference between the best and worst lineups has been about 45 runs per year. For the projected 2010 lineup, the difference is 29 runs. And since no one is going to bat Coco Crisp cleanup with Jack Cust and Kevin Kouzmanoff eighth and ninth, you can probably say that any reasonable batting order Bob Geren decides to run out there this year will be about as good as any other.
I’m too lazy/busy to run the numbers, but it makes you wonder, how many teams in baseball history have had a lineup where the batting order mattered less than the 2010 A’s?
Hello, Probably the ’62 Mets lineup order wouldn’t have mattered much either. Lineups, in general, are overrated. How often does a lead-off hitter actually lead off an inning, other than the first inning? And why is it more important to have a big guy to “clean up” the bases in the four-hole than in the sixth-hole? He is just as likely to find runners on base in any given inning batting sixth as he is batting fourth. Ah, well. It’s the nature of the business. Good post. Bill Miller (ondeckcircle.wordpress.com)
Regarding how much the batting order matters for a particular team, it’s not really a question of how good or bad the team is. It’s really more a question of how big the difference is between your best players and your worst ones. If they’re all good, or all bad, the order won’t matter. If some are good and some are bad, then the order matters more.
It’s true that a lot more energy is spent discussing lineups than it’s basically worth. 45 runs over a whole season between the best and worst of the 362,800 possible batting orders isn’t really very much, especially when you consider that most of the arguments aren’t about possible orders at opposite extremes.
The choices won’t make very much difference at the level of an individual game, maybe a fifth of a run per game, but they can add up over the course of 162 games to a win or two in the standings if you consistently do something dumb, like put your lowest OBP guy in the leadoff spot.
The people who have done the math on this (see http://insidethebook.com/) have figured out the basic heuristic for optimizing lineups to maximize a team’s runs scored. You take your three best hitters and hit them 1st, 2nd, and 4th. Your leadoff guy should be the OBP guy of those three, the 4th should be the SLG guy, and the 2nd guy should be the most balanced between OBP and SLG. Then you basically fill the rest of the lineup in order of how good they are.
I imagine there must be some difference between NL batting orders vs. AL batting orders, too, given the existence of the DH in the AL. Once you start using a pinch hitter for the pitcher in the NL as early as the 6th or 7th inning, lineups get all screwed up anyway. The same is true for late-inning “defensive” replacements who get unexpected at-bats in extra innings. Having your best OBP guy as your lead-off hitter is certainly important, but suppose your best OBP guy is also your best slugger? Then, does he hit 1st, or 4th? There are so many variables, I just don’t think it matters all that much, although as you said, one or two victories per season can (but usually doesn’t) make a big difference in the standings. Didn’t Billy Martin once pull his lineup out of a hat just to see what would happen? Or was it Tony La Russa? Anyway, good conversation. Thanks, Bill
Yeah, there are some differences. The one interesting thing that I’d never heard before that came out of the mathematical studies was that in the NL, there is some added value in having someone who can steal bases batting 6th. That way, they can get into scoring position for the 7th and 8th hitters before the pitcher comes up and ends the inning with an out.