Who should be the Twins starter in ALDS Game 1?

With another White Sox loss (7 in a row, geesh) and another Twins victory last night, Minnesota's magic number is 2, meaning they will almost certainly be the first team to clinch, perhaps as soon as tonight. The question will then be, how eagerly will Gardenhire pursue home-field advantage?  The Twins are neck-and-neck with the Yankees and Rays at this point, so they'll have to play dominant ball over the next two weeks in order to come out with the top record, especially since following the New York series, Tampa will be getting rich on Seattle, Baltimore, and Kansas City.  

Gardy's already announced that he's going with a six-man rotation for the next couple turns, but that doesn't necessarily mean the Twins can't win a lot of those games, facing teams like Cleveland, Detroit, and Kansas City.  It would, however, probably require him to continue to lean heavily on relievers like Jesse Crain and Matt Capps, while playing Joe Mauer and Jime Thome almost everyday, and my inclination is that all those guys could use some rest.

I know Gardy isn't exactly a stats guys, but when it comes to setting a playoff rotation, it's hard to ignore these:

Scott Baker - Home: 8-3, 3.98 ERA, 1.33 WHIP, Away: 4-6, 5.24 ERA, 1.34 WHIP

Nick Blackburn - Home: 6-3, 4.04 ERA, 1.31 WHIP, Away: 3-7, 6.99 ERA, 1.70 WHIP

Brian Duensing - Home: 6-1, 1.59 ERA, 1.06 WHIP, Away: 4-1, 2.86 ERA, 1.22 WHIP

Francisco Liriano - Home: 7-4, 2.83 ERA, 1.15 WHIP, Away: 7-4, 4.19 ERA, 1.38 WHIP

Carl Pavano - Home: 8-4, 3.90 ERA, 1.24 WHIP, Away: 9-7, 3.38 ERA, 1.09 WHIP

Kevin Slowey - Home: 8-4, 3.67 ERA, 1.20 WHIP, Away: 5-2, 4.91 ERA, 1.36 WHIP

What this suggests to me is that home-field advantage makes a big difference to the Twins, not necessarily in whether or not they can win a five-game series, but in terms of which pitchers give them the best opportunity for doing so.  Frankly, there's no way I give Blackburn a start away from Target Field, regardless of his recent hot streak.  I also make sure that Pavano is starting at least one of my road games, two if that's possible.  If all other factors are moot, here's how I think you'd set it up:

w/ Home Field:

Liriano (H), Duensing (H), Pavano (R), Slowey (R), Liriano (H)

w/o Home Field:

Pavano (R), Liriano (R), Duensing (H), Blackburn (H), Pavano (R)

The other thing Gardy should take into account is the fact that Yankee Stadium is tailored to left-handed hitters, so left-handed pitchers usually fair better there, thus perhaps tempting him to go against Liriano and Duensing's home/road splits.  Undoubtedly, he's waiting to see who his opponent is before setting the rotation.  The Yankees have a .625 W% against RHP this year, compared to .574 against LHP. Tampa is the exact opposite: .561 v. RHP, .667 v. LHP.  So, basically, the argument for Pavano over Liriano to get two starts makes much more sense if the Twins are starting the series in Tampa Bay, rather than in New York or in Minnesota.  

At the end of the day, however, while I think these stats should be taken into account when deciding whether to go with Blackburn over Slowey or Baker, I want my best pitcher taking the hill for two starts in a short series, no matter what the splits say.  And the best pitcher on the Twins is Liriano, and it's not really close:

Liriano - 14-8, 3.44 ERA, 1.25 WHIP, 9.38 K/9, 2.75 BB/9, 0.25 HR/9, 2.44 FIP (Fielding Independent Pitching), 6.3 WAR (Wins Above Replacement)

Pavano - 17-11, 3.60 ERA, 1.16 WHIP, 4.80 K/9, 1.54 BB/9, 0.90 HR/9, 3.92 FIP, 3.2 WAR

Let me break this down a little.  The Yankees and Rays rank 3rd and 5th in the AL in homers this season.  Both teams, but especially the Yankees, score a high percentage of their runs off the long ball.  Pavano is three times more likely to give up homers.  Pavano has also been a little lucky this season in terms of batting average on balls in play (BABIP).  Generally, when hitters manage to put the ball in play, it drops for a hit around 30% of the time.  Pavano has been are 28%, while Liriano has been around 34%.  Neither is an incredible outlier, but this explains to some extent why the FIP statistic theorizes that over a prolonged period of time, if Pavano and Liriano pitched under the same exact circumstance, Pavano's ERA would be a little higher than it is (3.92), while Liriano's would be a full run lower than it is.

But what really strikes me here is the Wins Above Replacement gap, with Liriano being three wins more valuable than Pavano.  WAR takes into account things like the quality of competition, the places the pitcher made his specific starts, etc.  Basically, what WAR is telling us is that although Liriano has been a bit unlucky in terms of wins and ERA, he's about twice as valuable a pitcher as Pavano.  This, I think you'll agree, is backed up by watching them pitch.  While Liriano has suffered from injury and inconsistency over the course of his career, he has frequently shown the ability to be a dominant "Ace," as he has been doing throughout this year.  Pavano, despite an excellent performance in 2010, has never really been that kind of pitcher.  Maybe he was in 2004 with the Marlins (18-8, 3.00 ERA), but the 34-year-old version is considerably changed with long intervening track record of mediocrity.

Obviously, it would be great to see Pavano dominate the Yankees in New York, wreaking vengeance for years of ridicule and abuse.  However, this story is all the more powerful if it unfolds in game two or game three, especially if the Twins already have a series lead.