ALCS Primer: Game 1

After three grueling days without American League baseball, the ALCS finally begins tonight. I've already written about the rotations, lineups, bullpens, and defenses. And as if that wasn't enough, I discovered that we really shouldn't just look at the Pythagorean records of these two teams and assume the Red Sox are a lot better; it's actually pretty close.

Game 1 Lineups
Against the righty Beckett, I would assume won't do anything weird- Victor catching, Garko at first, Asdrubal at second, Peralta at short, Blake at third, Kenny in left, Grady in center, Gutierrez in right, and Pronk DHing.

Boston is going to start Kielty in right, since he has the platoon advantage against Sabathia, and has hit CC well in his career (.310/.375/.655 in 32 PAs). Beyond that, I would assume it'll be Youk, Pedroia, Lugo, and Lowell from 1st to 3rd, Manny and Coco joining Kielty in the OF, and Ortiz DHing.

Game 1 Starters
Sabathia vs. Beckett is a pretty ridiculous matchup. I had somehow missed this, but Knuckle Curve points out that against lefties this year, Sabathia has walked four while striking out 75. Yes, his K:BB ratio against lefties this year is better than 18:1. Add in 3 HBPs, and 3 HRs allowed, and his Fielding Independent ERA against LHB is 1.49 this season.

That's pretty amazing, although it isn't going to be particularly helpful against the Sox, since with Drew out of the lineup they'll have only one lefty (Ortiz). Sabathia's FIP against righties is 3.48.

I saw on ESPNNews this morning that Beckett has thrown a shutout in three of his last four playoff starts. This is over the span of three series and five years, with a relief appearance in between, but that's still pretty amazing. In his last five October appearances, he has struck out 41 while walking six in 38.1 innings. He has allowed one homer, and his ERA is 0.70. Wow.

Game 2 Bullpens

Well, everybody is rested. There's a good chance that Papelbon will be the only reliever the Red Sox need, if they need any at all. For Cleveland, I would expect Betancourt to see more action than Perez because of all the righties Boston is starting, although they do have a few lefty bats on the bench (Drew, Hinske, Ellsbury, Cora).

One gambling note: here are the odds from BetUS.com for the exact result of the series.
I think that "Red Sox, 4-3" line is pretty good. G7 would either be Beckett (if they pitched him on short rest in G4) or Matsuzaka, against Jake Westbrook. In Boston. These lines only give Boston a 56.5% chance of winning that game. If it comes to that, even if it's Matsuzaka against Westbrook, Boston is going to be a much bigger favorite than -130, which is the corresponding line for 56.5%.

Completely unrelated, but can everybody (read: ESPN) please shut up about this interference call last night. It was a good call, he went out of his way (intentionally or not) to prevent Matsui from completing the DP. Where is the debate here?

G1 starts in about 3.5 hours, so it's finally time for a prediction. I'm gonna go with Indians in 6. I think Boston wins tonight, but the Indians take G2. Then the Indians take 2/3 in Cleveland (I'm not gonna bother trying to predict the exact games, since we don't really know who Boston's G4 starter is), then Fausto ends it in G6. I'm sticking with my "Fausto becomes famous" prediction.

Update: I would just like to add that my "gorilla math*" has CLE +151 as giving the bettor about a 10% advantage tonight.

*The Urban Dictionary definition for that term is awesome.

Completely unrelated, but a new college football blog written by some friends of mine who are reasonably intelligent: Week 7 Picks [Little Men on Campus]

Are the Indians "Built for October"?

When trying to gauge the relative strengths of teams in October, people tend to use run differential. The Red Sox were +210, the Indians +107, thus the Red Sox must be significantly better.

The runs scored portion of this makes sense to me. Boston averaged 5.35 R/G, Cleveland 5.01. Those numbers are probably pretty representative of the strength of their offenses.

In terms of pitching, however, run differential doesn't strike as a very effective way to judge teams in October. First of all, a lot of these guys aren't even on the playoff rosters. Does it really matter that Jeremy Sowers allowed 49 runs in 67.1 innings? Should we really care that Roberto Hernandez had a 6.23 ERA in 26 innings? As incredible as Clay Buchholz was in limited work (22.7IP, 4ER, 22K), that really isn't going to help the Red Sox against the Indians.

I think we also need to take into account the extra off days. During the regular season, C.C. Sabathia made 34 starts, which is 21% of the Indians' games. Regardless of the length of this series, his share of October starts will be higher. If the series is a sweep, he will have pitched in 25% of the games. If it's not, he will pitch game five, which means he'll pitch in at least 28.5% (2/7) of Cleveland's games.

I have always thought that the Indians are "built for October", with the strength at the top of their rotation, and their three excellent relievers (plus Borowski). So I decided to try to figure otu if there was any truth to that.

Here's what I did. For the bullpens, I took the FIPs of each pitcher. I then tried to figure out about what percentage of their team's relief innings each pitcher will throw. Here are the results of that little exercise:

These %s are obviously quite unscientific, but I think they're pretty decent approximations. Using these, we can reach a weighted ERA for each bullpen. Cleveland's comes to 3.11, Boston's to 3.27. This doesn't take fielding into account at all (thus the "Fielding Independent" portion of FIP), so I added each teams unearned runs per nine innings; 0.24 for Boston, 0.28 for CLE. So, the expected bullpen RA is 3.35 for Cleveland, and 3.55 for the Red Sox.

For the starters, I just used their individual RAs. I also took into account each guy's average innings per start. For example, Sabathia has an RA fo 3.51, and averages 7.1 IP/GS. That leaves 1.9 innings to the bullpen. So I combined Sabathia's 3.51 RA, and the Indians' bullpen RA of 3.35 (obviously weighting Sabathia's much more heavily), to come to 3.46 as the Indians' expected RA in a game Sabathia starts. Here are the numbers for each of the eight starters:

Okay, now we just have to account for the number of starts each pitcher will get. For this, I turned to the Vegas lines (of course I did). BetUS.com thinks there's a 15% chance of a sweep, 25% chance it ends in 5, 30% in 6, and 30% it goes 7.

Accordingly, the G1 starters will make an average of 1.85 starts, G2 starters 1.60, G3 starters 1.30, and G4 starters 1.oo (obviously). With this information, we can weight each SP's "TM RA" appropriately, and take into account that Sabathia and Beckett will inevitably be bigger factors in this series than Byrd and Wakefield.

Finally, we have how many runs we should expect each team to allow per game. Combining this with their R/G, here are the results:

These results are very different from what we get when we simply use RS & RA from the entire season. Doing that, we get an EXP W% of .623 for BOS, and .562 for CLE.

The Red Sox are still the better team, and they have home-field, but it's probably closer than most people think. Here are the W%s for each team in each game, taking the location of the game into account.

The Red Sox are favored in G1 (-155 at Bodog), and that makes sense. It will be interesting to see what the line is for G2; I would guess Boston will be slightly favored, and that may be legit if you want to talke Schilling's postseason dominance into account.

I think Boston will probably be favored in G3, as Matsuzaka inspires more confidence than Westbrook. I'm not so sure about G4 though- this analysis doesn't know that Wakefield is struggling through injuries.

Finally, if the Indians are going to advance, it looks like they're going to have to do it in G6, as things aren't looking too good for them if this thing goes the distance.

Photo: Cleveland.com.