Now that I've totally settled the roster questions facing the A's, what about the lineup? David Pinto's lineup analysis tool turns out to be rather accurate, so let's use that along with PECOTA to figure out how people should be hitting in the two major lineups the A's will likely be using.

The everyday players: Mark Ellis will be the second baseman, Orlando Cabrera will be at shortstop, Ryan Sweeney will be in center, and Matt Holliday will be in left. Most catchers don't play every day, but Kurt Suzuki saw action in a whopping 148 games last year, so he's on this list as well. Finally, while we all probably think Eric Chavez isn't long for the field, since he never is, we'll include him as an everyday guy, because as long as he's healthy, he will be out there every day.

Now, add to those guys Travis Buck in right, Jack Cust at DH, and Jason Giambi at first. Unfortunately, the very best lineups for this crew of players all have Cust leading off, which we know will never happen. Those best lineups score over 4.9 runs a game, a total that boggles the mind of A's fans given recent offensive performances.

A plausible lineup that Geren might build from this crew is: Sweeney, Ellis, Cust, Holliday, Giambi, Chavez, Cabrera, Buck, Suzuki. That lineup, with lots of nice LR splits (the only back-to-back hitters are Giambi and Chavez), projects to 4.881 runs per game, just 0.05 runs fewer than the best possible lineups.

How about slotting in Daric Barton instead of Travis Buck, playing Barton at first, shifting Giambi to DH and Cust to right field? I'd figure the same lineup as above, with Barton hitting eighth instead of Buck. They're both lefties, and neither has earned a higher spot in the lineup, so a straight swap seems plausible. Barton's higher projected OBP and practically equivalent power means that the Barton lineup projects to 4.897 runs per game, out 0.07 worst than the best lineups, which again have Cust leading off.

Is 0.016 runs per game on offense worth the defensive hit that results from replacing Buck in right with Jack Cust? Cust's career right field UZR/150 comes out to -0.142 runs per game, while Buck rates out at +0.062, a difference of 0.204 runs per game. That's a pretty massive difference, and it dwarfs the lineup difference.

Of course, that lineup would also include Barton at first base instead of Giambi. Giambi's career UZR/150 gives him a -0.037 runs per game, while Barton's is +0.046, a difference of 0.083. That cuts into but does not utterly dissipate the difference in defense between Buck and Cust. The difference between the two lineups is basically a run every ten games, which is not an insubstantial amount -- that's sixteen runs, or a win and a half, over the course of the season. In a division where the winner might go to October with 85 wins, these margins matter.