2010 Orioles Line-Up

With the addition of Miguel Tejada recently, it seems that the players in the 2010 starting line-up are largely set. That then leaves the issue of how best to arrange them. As of now, here are my offensive projections for the O’s hitters (let me know how you think they will do: 2010 O’s Projections):

Matt Wieters – .287/.352/.452, with a .350 wOBA
Garrett Atkins – .263/.332/.422, with a .329 wOBA
Brian Roberts – .282/.355/.431, with a .345 wOBA
Miguel Tejada – .296/.332/.435, with a .333 wOBA
Cesar Izturis – .259/.304/.339, with a .285 wOBA
Nolan Reimold – .278/.360/.459, with a .357 wOBA
Adam Jones – .283/.339/.466, with a .347 wOBA
Nick Markakis – .297/.366/.484, with a .369 wOBA
Luke Scott – .263/.343/.477, with a .353 wOBA

First up, let’s assemble the line-up using the guidelines put forth in The Book (if you haven’t read it, you should):

“Your three best hitter should bat somewhere in the #1, #2, and #4 slots. You fourth- and fifth-best hitters should occupy the #3 and #5 slots. The #1 and #2 slots will have players with more walks than those in the #4 and #5 slots. From slot #6 through #9, put the players in descending order of quality.”

Step 1 is to figure out who the three best hitters are. Nick Markakis looks like the clear #1, with Nolan Reimold, Luke Scott, Matt Wieters, Adam Jones, and Brian Roberts all being pretty close together thereafter. Roberts is the one whose wOBA is tilted more towards the OBP side, so he goes first (and because he’s fast and stuff). One could concievably make the case for Reimold though, since he walks and is actually a pretty good base-runner. Markakis has the highest OBP on the team, and isn’t as much of a home run hitter, so he would bat second. Because the #2 hitter bats with runners on first so often, putting a guy like Tejada there – since he grounds into a lot of double plays – would actually be problematic. The singles and lack of strike-outs helps leverage Roberts’ base-stealing, but I don’t know if I like that trade-off. For the clean-up spot the choice is really between Reimold and Scott, though Wieters might also have a claim. Strike-outs are worse in the four-spot since there are often runners in scoring position, but Scott’s edge in power gives him the nod. The #3 hitter leverages home runs more and OBP less than the #5 hitter (who should be better overall), so I guess Wieters would bat third and Reimold fifth. After that, it’s descending order by quality. That would make the line-up:

Roberts – S
Markakis – L
Wieters – S
Scott – L
Reimold – R
Jones – R
Tejada – R
Atkins – R
 Izturis – S

So just how optimal is this line-ups? To find out, I headed over to the line-up analysis tool at Baseball Musings (it’s not the most precise thing in the world, but it’s handy for a quick look). Plugging in the projected OBPs and SLGs, we get that the above line-up would score 5.191 runs per game (841 per season). That’s pretty darn good, considering the optimum line-up is at 5.213 runs per game (845 per season):

Reimold
Markakis
Atkins
Scott
Wieters
Jones
Tejada
Izturis
Roberts

That strong OBP is what pushes Nolan into the #1 slot, with Roberts batting last so that he can still get on-base for the top of the line-up. Atkins in the three-hole is interesting though, and likely a bit of a quirk in the system. The first line-up we get with Roberts leading off is #27, at 5.205 runs per game (843 per season): Roberts Markakis Wieters Scott Reimold Jones Tejada Izturis Atkins If you switch Izzy* and Atkins, then that would be one of the more realistic looking ones while also doing a little better on the run scoring than the ones I constructed above. The difference between Izturis-Atkins and Atkins-Izturis comes out to about 0.014 runs per game (a bit over 2 runs for the season), so it’s not exactly a huge concern.

* Every single one of the top 30 line-ups – and that’s as far as it goes – has Cesar batting 8th instead of 9th. It’s worth it to give him a few more plate appearances to allow someone else who can get on-base more often to bat infront of the top of the line-up.

The take-away for me is that there are a lot of guys who can bat in various slots, and so the order makes even a little less difference than it usually would. In any case, the line-up would score in the neighborhood of 800 runs, or a bit over (the line-up tool tends to overestimate, given that there will be injuries and such). That would be a 60+ run improvement from 2009, and would have ranked in the top 4-6 in the league last year. The worst line-up the team could put out would apparently be:

Tejada
Izturis
Jones
Roberts
Atkins
Reimold
Wieters
Markakis
Scott

That comes in at 4.968 runs per game (805 per season). For all the complaining about how to construct line-ups, the difference between the best possible one and the worst possible one is only 40 runs (4 wins). That’s certainly not insignificant, but given that the just about the worst realistic line-up I can construct is:

Roberts
Tejada
Jones
Markakis
Atkins
Scott
Reimold
 Wieters
Izturis

And it’s still at 5.119 runs per game (829 for the season). That’s suboptimal by only about 15 runs (1.5 wins), but that is also an overestimation of the disparity. And given the general precision level, it’s certainly possible that this line-up might actually be better than the previous one (though I doubt it). My main concern with the line-up is whether batting Nick second will encourage him to take more walks again… that’s pretty much it.