Lineup Analysis: 2009 and 2010

I was beaten to the punch on lineup simulations in the comments yesterday, I know. But since when have I let that stop me?

-----

As it stands, the Seattle Mariners are currently the proud owners of an improved offence over last year. Granted, that wasn't very hard to do, as the team scored an AL-low 640 runs en route to a lucky (by pythagorean standards) 85-77 finish. Current calculations have the team at around 40 runs better than last year's edition, which would imply that we're in for a whopping total of 680 runs scored next year, which would make us... still worse than the Royals. This, say many people, is a problem. Our offence is still going to suck, and we were lucky to be above .500 last year. The Mariners are going to struggle to compete in the AL West.

I feel that these people are being overly hasty in dismissing the team. Observe:

2009 Seattle Mariners: 640 runs; 3.95 R/G.

2009 Seattle Mariners run through a lineup simulator (optimal lineup): 710 runs; 4.382 runs per game.

Hmmmmmm. I used the actual numbers from each position, too, so it's not like we're totally neglecting the fact that Ronny Cedeno exists or that Mike Sweeney held down first base for a while. For those who are curious, I've included the optimal lineup below, including on-base-percentage and slugging.

Table 1: Optimal Lineup for the 2009 Mariners

It looks pretty familiar, actually. Ichiro up top is a no-brainer, and Wakamatsu was experimenting with Branyan at #2 a lot. Essentially, we didn't really deviate that far from an optimal lineup last year, and yet for some reason massively underperformed what we should have done even accounting for our frankly abysmal slash numbers. Not to go tumbling off on a tangent here, but it deserves saying again: our hitting was a total, total shambles last year. Four positions contrived to produce an OPS lower than .650, and two of those positions didn't even manage to reach the lofty peaks of .600. For comparison, Joey Gathright's career numbers would have been a significant improvement over our shortstops and left fielders last year. Yikes.

Anyway, for some reason, the Mariners scored less runs than they should have, by between 40-70 runs. The simplest explanation is one of non-clutchness (or luck, in other words). The 09 Mariners were completely incapable of stringing together hits in such a way as to actually score runs. Amazingly, this bad luck almost totally cancels out the good luck from winning all of those close games. Our record last year was fluky in such a way that it actually was a decent reflection of the team's talent. So, even without going into WAR or other advanced metrics, we have a much higher hitting base to build on than we thought we might.

Onto 2010. Bradley, Figgins, and Kotchman are in. Branyan and Beltre are out. I've given the 'expected lineup' my best shot, using CHONE for all the non-Ichiro batters and Ichiro's career average line (because projection systems suck at Ichiro). Unfortunately I don't have the patience to account for injuries, player days off, etc, so these numbers are going to be a touch optimistic (possibly countered by CHONE hating Figgins and Bradley).

2010 Seattle Mariners projected through a lineup simulator (optimal lineup): 758 runs; 4.676 runs per game.

And the lineup:

Table 2: Projected Optimal Lineup for the 2010 Mariners

First off, that's a weird looking lineup, and I doubt anyone's going to put Chone Figgins anywhere near the #9 spot.  751 runs would have been good for 9th in the league last year, and it's only 30 runs below league average, a much happier number than the -140 that we posted last season.

I'm going to knock off maybe 15 runs for our bench players filling in (remember, we're probably going to be adding another bat, so this is probably overly conservative) and run a quick pythag check on our wins, using our run prevention from last year and knocking it down a little bit because of CLIFF LEE WOOO. So, a back of the envelope guesstimate:

RS: 736
RA: 685
Pythag*: 0.528
Wins: 85

Not too bad for a conservative estimate, is it?

*Using Patriot's method.

PS: This would have looked much better with Matt Holliday.