Jacoby Ellsbury wasn't the Red Sox's best hitter in 2011, but he was real close. Only Adrian Gonzalez (155) and David Ortiz (154) had higher OPS+ in a lineup that scored nearly five-and-a-half runs per game. He got to that point with some new-found power, more than even the optimistic scouting reports from his college days could have envisioned: Ellsbury smacked a team-leading 32 homers, and a 83 extra-base hits overall.
Even in a lineup full of heavy hitters such as these, one wonders if the Red Sox would be better served moving this more powerful iteration of Ellsbury down in the lineup. Alex Speier approached this question from a specific angle the other day, asking if it was worth giving up more plate appearances over the course of the season in exchange for more plate appearances with runners on. Essentially, should Ellsbury hit first to set the table, or should he be the one clearing the table off?
For obvious reasons, no spot in the Red Sox lineup stepped to the plate more times with the bases empty than the team's leadoff hitter. The No. 1 overall hitter in the lineup batted 274 times with runners on base, three fewer than the 277 times that the team's No. 9 hitter batted with base runners and more than 100 fewer times than the batters in the No. 4 spot (390 times) and No. 3 spot (378) stepped to the plate with runners on base.
To break things down even further, we can look at the total number of runners on base for the Red Sox hitters who saw at least 300 plate appearances in 2011:
"OBI" is "Others Batted In" -- essentially runs batted in minus home runs. OBI percentage is the rate at which hitters drove other runners in. This number isn't predictive -- it's basically the result of players lines with runners on base from a single year -- so don't expect a replication of these exact figures in 2012.
However, we can see that Ellsbury was as qualified as anyone else in Boston's lineup to hit in the middle as far as 2011 goes. He just didn't get the same number of opportunities, as he had 96 fewer plate appearances with runners on than Gonzalez (despite 14 more PA overall) and 114 fewer runners on base total. In fact, despite having 212 more plate appearances than Kevin Youkilis, he had six fewer plate appearances with runners on base, and 13 fewer runners on total.
It's not surprising that a good hitter would be capable of driving runners in, but you can see the stark contrast between leading off the game and then following the Jarrod Saltalamacchias of the world, and being Adrian Gonzalez or David Ortiz, who get the benefit of having two or three of the game's very best hitters getting on base prior to their turn at the plate.
Rearranging the lineup so Ellsbury bats in the middle would surely increase his own run production, but would it benefit the Red Sox? There is an opportunity cost to moving Ellsbury, as it also means other hitters need to move as well, and their opportunities might be lessened in favor of Ellsbury's. Using the lineup simulator from the aptly-named LineupSimulator.com, we can get a sketch of what shuffling the Red Sox lineup around might do.
PECOTA projections for 2012 were plugged in, with an Ellsbury's forecast somewhere in between his pre-2011 and 2011 figures (203 hits, 53 extra-base hits, 53 steals). Various lineups were created. The standard, with Ellsbury leading off. Carl Crawford leading off with Ellsbury third, and the rest of the lineup situated to break up the lefties. Ryan Sweeney leading off, with Adrian Gonzalez or Dustin Pedroia second, and Jacoby Ellsbury third, with Youkilis and Ortiz coming up after. Even though Pedroia won't lead off, a lineup with that situation was produced just to see. Ellsbury was plugged in to bat fourth in a few iterations, to see if that meant more than hitting third. The bases were covered to see how the Red Sox could most effectively clear them.
Regardless of what was plugged in, the scenario with Ellsbury leading off produced the most runs (4.99 per game. 808 for the year). The difference from that lineup and ones with Ellsbury batting elsewhere were small, though, as the lowest output from realistic lineups was 4.96 runs per game, or 804 runs. One is "better" in a sense of the word, but we're talking hundreds of a run difference per contest.
Next, Ellsbury's 2012 projection was replaced with his 2011 numbers, to see if a more powerful Ellsbury deserved to be any lower in the lineup than the compromise version from his projections. The Red Sox now were set to score 5.15 runs per game, or 834 on the season. Shifting the lineups around in the same way described above, the result was the same: Ellsbury leading off produced the most runs. The differences between the best and worst lineups were still about 3-5 runs over the course of the entire season.*
*This is just a simulation, so don't take it as gospel for what the Red Sox will or won't do in 2012. As a rough sketch for the idea goes, though, it fits the bill.
Moving Ellsbury down in the lineup creates more opportunities for him to drive runners in. However, due to the heavily left-handed lineup in Boston, the move means that Adrian Gonzalez, David Ortiz, and Dustin Pedroia could end up with fewer plate appearances on the year, and therefore fewer opportunities to drive in runners. It also causes someone like Crawford or Sweeney -- good hitters in their own right, but lightweights compared to these offensive behemoths -- to end up with more time at the plate.
More Jacoby Ellsbury overall might be more important than more Jacoby Ellsbury with runners on. The difference is so slight from a game-to-game basis, though, that it almost doesn't matter what the Red Sox decide to do. As they are lacking in terms of other legitimate options to lead off, though, keeping Ellsbury there -- even in his 30-homer form -- makes the most sense. It worked out fine for Boston last year, and the lineup has the same ridiculous depth now as it did then.