Sunday, October 7, 2012

2012 ALDS Preview - Orioles vs. Yankees

After battling down to the wire, the Yankees were able to stave off the Orioles to win the AL East by two games.  Their reward for that is to play the Orioles on the road for two games to start the ALDS.

All season long, we’ve heard/read/said that the Orioles were not for real, that they were lucky and that they were due to crash at any point.  That was mainly a function of a run differential that was negative for almost the entire season as well as seemingly unsustainable records in both one run games (29-9) and extra inning games (16-2).  In general, teams approach about .500 in those games although having a strong bullpen can help a team win a higher percentage of close games and that’s something the Orioles have had all season.

None of what the 2012 Orioles got lucky with matters right now.  The only thing that matters is how good the 25 man roster they will have for the ALDS is, and how it may match up against the Yankees.  Because of that, I’m going to look at projections for the Orioles and Yankees which will account for their performances this year as well as in recent prior seasons.  This will account for the fact that some players may have over/under performed expectations and are possibly better or worse than their 2012 numbers, but more importantly it will account for the fact that the rosters and playing time distributions in a five game series are a lot different than they are in a 162 game season and simply comparing two teams’ over 2012 Pythagenpat records is the height of laziness when it comes to statistical analysis.  Well that or using FIP for pitcher WAR.  But I digress.

For the projections, I’m using Oliver from the Hardball Times.  They’ve been updated as of last week and include 2012 MLEs for players that spent time in the minors so I feel they are the most complete version of forecasts available right now.

So, as I mentioned, it’s all about rosters and playing time.  Based on what is out there, here is my rough guess at those two things.  First up, the Orioles’ lineup and bench.

wOBA: Weighted on base average
BR: Linear weights batting runs

The idea here is to try and allocate PA over a 5 games series.  So basically it was a case of adding PA until the team got to 135 outs (27 outs times 5 games).  That includes double plays and obviously there may be games where teams don’t need to bat for 27 outs (home wins) but as long as the scale is the same for both teams it doesn’t matter.

The Yankees will be throwing lefties CC Sabathia and Andy Pettitte in the first two games and if there was a game 5 it’d probably be CC again so that might impact some of the PAs for the lefties.  So I gave some DH PA to Lew Ford instead of Jim Thome and some 2B PA to Robert Andino instead of Ryan Flaherty.  For the most part I don’t expect the other bench guys to play much.

Basically this estimates that the Orioles would score about 24.5 runs in an average five game series using this allocation of playing time.

And here’s what the Yankees lineup and bench should like.

Yes, I know Andruw Jones and Chris Dickerson are not on the roster.  They are just place holders.  Anyway, the only spot I see some finagling of playing time is DH with some combination of Raul Ibanez, Eduardo Nunez, Alex Rodriguez and Eric Chavez.  I threw Jayson Nix three PA for the hell of it as well. 

So the Yankees have a very slight offensive edge here, 26.4 runs to 24.5 runs.

How about the pitching?  Here’s my stab at the Orioles first.

Apparently the Orioles will be using Jason hammel, Wei-Yin Chen and Miguel Gonzalez in the first three games.  It’s expected that Chris Tillman would pitch the fourth game if necessary although it could also be Zach Britton or Joe Saunders.  So a lot of this could be impacted by how that plays out.

Although Hammel is on tab to pitch the opener which should make him available for Game 5 if needed, he hasn’t been healthy in the second half so I restricted him to one start and gave a start to Joe Saunders.  Because Oliver doesn’t like most of the O’s starting pitchers after Hammel and Chen and I think it may be underselling them a bit I limited them to five innings and gave more innings to the bullpen where they have better projected performers.

The big question for Baltimore is whether Gonzalez and Tillman are closer to the 3.25 and 2.93 ERAs they posted in the regular season or to their 4.78 and 5.22 projected ERAs.  It’s the difference of two runs in this scenario.  That may not seem like much, but it’s the equivalent of a swing of about six wins in a five game series but I’ll get into that in more detail shortly.

Same crap for the Yankees.

The Yankees should be in very good shape if the CC Sabathia we saw in his last three starts is the CC Sabathia we see this postseason.  His velocity still hasn’t been overly impressive to me and the fact that he pitched well against an awful Blue Jays team and an even worse Red Sox team isn’t overly impressive to me either.  But let’s hope for the best.

So we’re looking at a Yankee pitching staff that projects to allow around 20 runs vs. a Baltimore staff that projects to allow 23.  Running the offensive and pitching numbers through Pythagenpat looks like this.

If you instead think that Gonzalez and Tillman are the guys they were in 2012 it looks like this.

That seems a bit closer to reality for me.  The Yankees should be favorites to win this series, but not overwhelmingly so.