I got home from work in time to see the ninth through eleventh innings of ALCS Game Three last night. And I want to preface this by saying that there is no doubt in my mind that both Mike Scioscia and Joe Girardi make for much, much better managers of baseball teams than I would, for any number of reasons. That said, it looked to me like both managers wanted very badly to lose this game. Here are just a few of the things I noticed that I think were at least questionable (note that not all of these things impacted the game, and in fact some had a positive impact, but it's the process, not the result, and the process was bat-feces crazy):

• Bottom 7: Izturis pinch hits for Napoli. First comes one I didn't actually see, and that I'm not sure was actually a bad move (but I'm leaning that way): in the bottom of the seventh, Scioscia sent up Maicer Izturis in place of Mike Napoli. On one hand, there was a runner on third and one out, and Napoli strikes out three times as often as Izturis, so your odds of getting one run there via a sac fly or something similar are at least a little better with Izturis. On the other hand, (a) Napoli is clearly a much better hitter, and your odds of getting more than one run that inning were much better with Napoli, and (b) removing Napoli that early meant there was a good chance that backup catcher Jeff Mathis would get at least one at-bat in the game, meaning you've swapped Napoli's very good .364 wOBA for Mathis' very bad .263. In the end, it "worked," as Izturis gave them that sacrifice fly, which at the time put the Angels up 4-3. On the other hand, maybe Napoli homers, or doubles, or does something that led to their scoring two runs that inning and avoiding the whole extra innings thing. Or maybe he comes through in the ninth. Of course, we know what Mathis did. But before the fact, it's much, much better to have three PA by Napoli than one by Izturis and two by Mathis. At any rate, it was better than what Scioscia did with Napoli on Saturday (why does he hate that guy?), but I think on balance it was a bad move.
  (click here to read more)
  
• Top 9: Fuentes intentionally walks A-Rod. Now, you've got a tie game in the ninth inning, your "closer" on the mound, nobody on base, and two outs. Granted, light-hitting Brett Gardner is coming up next, having pinch run for Matsui in the 8th (a move which backfired in the worst possible way -- Gardner getting thrown out stealing right in front of Jorge Posada's game-tying HR -- but again, it's about the process, and I actually think that move made sense). But to put the winning run on base, you've essentially got to determine that Alex Rodriguez's odds of hitting a home run are better than the odds of (a) Gardner's pinch hitter, Jerry Hairston, doubling/tripling/homering A-Rod home, or (b) Hairston walking or singling followed by the next hitter, Posada, coming up with a hit that gets the run(s) home. (Of course, if you pitched to A-Rod, he could double, but then if you walk him, he could steal second, with his 94% success rate in '09 -- we'll just stick with the HR scenario.)
  On the season, both against lefties and overall, A-Rod homered in 6.7% of his AB. In his best homer-hitting season, he homered in 7.9% (just 5.0% vs. lefties). So let's be generous (to Scioscia) and say that A-Rod had, I don't know, a 7.5% chance of homering right there. Hairston got an extra-base hit in 8.9% of his at-bats this year. If that didn't happen, he had a 21.7% chance to single or walk, followed by a 29% chance that Posada gets a hit (which combines for a 6.2% chance that both Hairston reaches and Posada drives the run home). So you walked him against a 7.5% chance that A-Rod homered, forfeiting a 60% chance that you get A-Rod out, and then were faced with a 16.1% chance that some combination of Hairston and Posada drive him in. There's just no way that's a good move.
  
  
• Bottom 10: Mathis is permitted to run for himself. This one was so dumb, even Buck and McCarver noticed. Reggie Willits, rotting away on the Angels' bench, is a great runner; Mathis is a catcher. The Angels decided to carry three catchers on their postseason roster. Mathis was standing on second base with nobody out. If there's ever any reason to carry three catchers, this would pretty much be it. Again, as it worked out, it was a good thing Mathis was in there for the 11th. But there's also a decent chance that Willits scores in Mathis' place in the tenth. And Mathis, two huge doubles yesterday and all, is a terrible hitter. Again, result good, process bad.
  
  
• Bottom 10: Hairston replaces Damon. After Hughes and Mariano Rivera had combined to load the bases with one out, Hairston, who had pinch hit into the DH position, goes out to left field to replace Johnny Damon, forfeiting the DH position, meaning that Rivera would be due up third if the top of the eleventh ever came. Now, Damon has a notoriously terrible arm, and Hairston's is undoubtedly better. But so much better that it's worth giving up the DH should you get to the next inning? There's not a great chance that he hits it to left field anyway, and if he does, it would have to be hit just so to be the kind of play where either (a) Hairston could throw the runner out but Damon couldn't, or (b) Hairston's arm scares the runner into staying put where he would've tested Damon's (and Hairston is still primarily an infielder; it's not like they subbed in Roberto Clemente or something, fear-wise). I know managers always want to do whatever they can to stay in the game, and I get that, but the benefit there was very, very minimal, and the cost should you somehow get out of the inning alive -- forfeiting your DH, when you're already running out of players -- was huge. They got out of the inning, of course, and somewhat miraculously. But Hairston, predictably, didn't have anything to do with that.
  
  
• Top 11: Cervelli pinch hits for Rivera. I imagine this will be my most controversial call, but I hated this move. After Cabrera and Jeter made two quick outs in the top of the 11th, having already foolishly forfeited the DH, Girardi sent up his third catcher, Francisco Cervelli, in place of Rivera, who had thrown 17 pitches in wriggling out of the bottom of the 10th. In 100 plate appearances in the bigs, Cervelli has hit .298, but with no power or patience (the rare wOBA, .283, that was lower than his batting average). In the minors this year, he hit .233/.290/.346, with 3 HR in 146 PA. He can't hit. The odds of Cervelli, a right-handed hitter, getting on base against Ervin Santana, a good and completely rested right-handed pitcher, were very, very low (25% or so, maybe?). Then, you'd still need Teixeira to drive him home, which, good as Tex is, wouldn't have had more than a 20% or so chance of happening. The Yankees just were not going to get a run in that situation. Why not let Rivera wave at the ball three times (something not without precedent) and send the still-fresh best closer in history out to get them to the 12th? I would expect most managers to do more or less what Girardi did, but I think sticking with Rivera would've been the better, if gutsier, move.
  
  
• Bottom 11: Aceves replaces Robertson. This is the one Yankees fans are no doubt grumbling about this morning, because this is the one that most visibly didn't work out. And to be fair, it was really, really bad. Righty David Robertson (a really nice-looking young pitcher with great stuff, getting 13 strikeouts per 9 in 43 innings and posting a great 3.05 FIP with the big club in '09) came in to start the inning, requiring 11 pitches to pick up two outs. Righty Howie Kendrick was coming up, so of course Girardi (after the Fox camera captured him taking three seconds or so to look something up in a big scouting binder) goes out to replace Robertson with another righty, Alfredo Aceves, a slightly less nice-looking and less young pitcher, with a 3.75 FIP in 84 innings.
  First, I have to say: even if there was a platoon matchup or some other compelling reason to make this change, you probably don't do it here. This game could go 20 innings, for all you know, and the only pitcher you've got left after Aceves that you might feasibly use is Chad Gaudin.
  But that aside, why did he do it? According to the man himself, Girardi "liked the match-up better," and declined to elaborate. Kendrick was 1-for-2 lifetime against Robertson and had never faced Aceves, but he could have 40 or 50 at-bats against one of them, and that still wouldn't be nearly enough to go on. I guess the thinking could have been that Robertson throws a lot more fastballs than Aceves, and according to Fangraphs, Kendrick did a lot more damage against fastballs than any other pitch (though, of course, he also saw a lot more fastballs, so I'm not sure what to take away from that). That would seem to me to be at least offset by the fact that Kendrick is a free swinger, and Aceves threw a lot more strikes; Robertson walked nearly three times as many as Aceves, and was still much more effective than Aceves overall. It would seem to me, then, that Robertson is the perfect guy to get a hitter like Kendrick out.
  And at any rate, even with all the advanced scouting and whatnot that I'll never see that Girardi no doubt has, there's just no possible way to justify that move. If you like Robertson to open the 11th inning, you've got to stick with Robertson to finish the 11th inning, and maybe the 12th too.