March 23, 2011

AL East Lineups & Expected Runs Per Game

A few days ago, I said that as long as the nine regulars are healthy, it won't matter much what order they bat in. Still, it's fun to debate and wonder ...

Terry Francona has dropped a few hints. Thinking about those, along with his spring training lineups, we can assume the top half of the order will look something like: Jacoby Ellsbury, Dustin Pedroia, Carl Crawford, Kevin Youkilis, and Adrian Gonzalez. Tito has said he might sometimes flip the 2/3 hitters and the 4/5 hitters.

Dave Cameron of FanGraphs used the ideas from "The Book: Playing the Percentages in Baseball", by Tom Tango and Mitchel Lichtman, to see what the optimum Boston lineup would be. Cameron's article is behind ESPN's Insider paywall, but this is the lineup (with projected 2011 OPS):
Drew .835 Youkilis .887 Ortiz .872 Gonzalez .976 Pedroia .841 Crawford .850 Scutaro .722 Ellsbury .733 Saltalamacchia .690
I'm betting we do not see this lineup for even one game in 2011.

Baseball Musings' David Pinto has been taking quick looks at each team's offense. He gets a probable batting order, plugs each player's OBP/SLG forecasts from Marcel the Monkey into his Lineup Analysis Tool, and gives us the results.

First, here are the probable lineups for Boston, New York, and Tampa Bay -- the top three run scoring teams in MLB last year (and the only clubs to top 800 runs):
Red Sox         Yankees       Rays
Ellsbury        Jeter         Jaso
Pedroia         Granderson    Damon
Crawford        Teixeira      Longoria
Gonzalez        Rodriguez     Ramirez
Youkilis        Cano          Joyce
Ortiz           Swisher       Zobrist
Drew            Posada        Johnson
Saltalamacchia  Martin        Upton
Scutaro         Gardner       Brignac
Here are the Runs per Game:
                  Red Sox   Yankees    Rays
Probable lineup     5.35      5.25     5.15
Best lineup         5.39      5.30     5.25
Worst lineup        5.12      5.03     4.97
Regressed lineup    4.85      4.79     4.71
I asked Pinto what "regressed lineup" means and he replied: "It's the best fit of the projected lineup onto how teams actually score during the season. Most teams tend to score less than the projected lineup, since that lineup suffers injuries and doesn't play every day."

The best Red Sox lineup -- Gonzalez, Youkilis, Ellsbury, Drew, Pedroia, Ortiz, Crawford, Saltalamacchia, Scutaro -- would score 0.04 runs per game more than Francona's probable lineup -- a difference of 6.5 runs over a 162-game schedule or an improvement of only two-thirds of one win.

Even using the absolute worst possible lineup -- Saltalamacchia, Ellsbury, Drew, Scutaro, Crawford, Pedroia, Gonzalez, Youkilis, Ortiz -- every single day would cost the Red Sox only 3.7 wins over Tito's probable lineup.

I'll still probably complain about some particular day's lineup, though.

9 comments:

Benjamin said...

I think there's a typo -- the FKRs' best lineup can't be lower than their probable lineup, or it wouldn't be their best, no?

allan said...

Yes. I did it with the MFY, too. I moved Probable ahead of Best (from what Pinto posted) and forgot to switch the numbers. Fixed now, I hope.

laura k said...

I'll still probably complain about some particular day's lineup, though.

We could do a contest on the first day you complain about the lineup.

I'd go with May 1, and I'm being generous.

Rick (f/k/a westcoastsox) said...

Good to see Martin Gardner batting 8th/9th for the Yankees

Iron Allison said...

I don't like math, but I appreciate how many different ways you can apply it to baseball statistics! And it usually gives a fairly accurate indication of what could be or what will be, but fortunately the baseball season is full of surprises!

s1c said...

I'll still probably complain about some particular day's lineup, though.

That of course is what is fun about being a fan of baseball. Statistics are always fun and in this game with old school and new school tools the debates can just go on and on!!!

btw - over under on complaint will be game 3!!!

dj1480 said...

Did Pinto's method take base-stealing into account? Does the old-school wisdom of leading off with your best base-stealer survive the statistical analysis?

allan said...

Does the old-school wisdom of leading off with your best base-stealer survive the statistical analysis?

It shouldn't, since if you can't get on base very often, who cares how fast you are?

Pokerwolf said...

CI hitting leadoff for the MFY? Seriously? As a Sox fan, I won't complain, but I think that's a big risk based on how he did last year.