In the comments section this week, someone brought up a potential batting order for the Mets this year and it got me thinking.
Specifically, it got me thinking: “Hey, I should run the Mets hitters through David Pinto’s Batting Order Optimizer and see what comes.”
Pinto’s tool, based on the work of Cyril Morong, Ken Arneson and Ryan Armbrust, uses players’ on-base and slugging percentages to generate the optimal lineup for each team, by runs per game. It’s fun to play with, though since it actually involves plugging player names and stats into a spreadsheet, it’s one of the nerdiest baseball-related pursuits you’ll ever enjoy.
Anyway, I probably should have considered platoon splits and all, but it’s Friday. I simply plugged in the Mets’ position players’ CHONE projections for 2010 and a .150/.150 line for “Pitcher McGee” and this is what it spit out, a lineup that would, it claims, score 4.78 runs per game:
1. David Wright .391 / .502
2. Jason Bay .376 / .514
3. Angel Pagan .334 / .428
4. Daniel Murphy .328 / .429
5. Jose Reyes .360 / .458
6. Jeff Francoeur .317 / .435
7. Omir Santos .296 / .359
8. Pitcher McGee .150 / .150
9. Luis Castillo .367 / .350
Before you freak out, I’m not saying that’s what should happen. Obviously it won’t, for one thing — there’s no way the Mets will hit Wright leadoff and Bay second — and it’s all based on the CHONE projections, which are only projections. It doesn’t account for handedness or egos or anything else. This is merely what would be optimal for that group of guys based on Morong’s assessment of how to best weight OBP vs. SLG with respect to batting-order positions. Wright bats first because he projects to have the highest OBP, simple as that.
I imagine the Mets’ actual lineup to start the season will look something more along the lines of:
I have no evidence that will be the order, plus there’s always a chance someone gets hurt. This was just my best guess at how it will shake out. Francoeur and Murphy are flip-floppable, but I figured hypothetical Jerry Manuel would want to use Murph to break up the righties in the order.
That lineup, according to the Baseball Musings tool, would average 4.536 runs per game if the players held to their CHONE projections, which they probably won’t.
Here’s something, though: All of the top suggested lineups for the 2010 Mets, according to the optimizer, include the pitcher batting eighth and Luis Castillo batting ninth.
I remember when Tony La Russa first started batting his pitchers eighth a couple of years ago. I ripped him apart for it. Not in print, thankfully, but man, I called him all sorts of nasty things to anyone who would listen. What kind of moron would voluntarily give his pitcher an at-bat earlier in the game than a position player, and more at-bats total?
But Tom Boorstein, in the midst of one of my rants, told me there was a lot of evidence that showed that, indeed, there was an advantage to batting the pitcher eighth and a high-OBP, low-SLG guy ninth. I don’t remember exactly what article he pointed me to online — it might have been this one — but I’ve since come to realize that the idea actually makes a lot of sense.
After all, a team tends to concentrate its very best hitters near the top of the order — as it should — to maximize their at-bats and increase its chances of scoring runs early in the game.
But after one time through the lineup, it’s no safe bet the batting order will reset with the leadoff hitter starting off the inning. Putting a hitter like Castillo in the ninth slot decreases the odds that the top of the Mets’ order — the best hitters — comes up with outs on the board and ups the chances they’ve got someone on base to drive in.
Since Jose Reyes has pretty decent power for a leadoff man, having Castillo hitting ahead of him the second turn through the order would give Reyes more opportunities to drive runs in, rather than forcing him to hit after Omir Santos and the pitcher, near-automatic outs.
Plus, part of my original argument — that moving the pitcher up in the order ultimately gives more at-bats to pitchers — is just silly. Starting pitchers rarely bat four times in a game, and when they do, they’re probably pitching well enough that it doesn’t really matter that they’re hitting so frequently.
More likely, a starting pitcher is being replaced in the lineup by a far superior hitter after his second or third at-bat, and every time his spot in the order comes up after that.
Going back to the Baseball Musings tool, I plugged in this lineup:
The outcome? 4.737 runs a game, .21 higher than the one I’m guessing the Mets will actually go with, and about 34 runs more over the course of a 162-game season.
That’s all just in theory, of course, and I’m just having fun with some nerdery on a Friday afternoon. So calm yourselves down.