In four important division games since Friday, the Angels are 1-3. I thought I'd point out an interesting trend I noticed:
Record when Mike Trout starts: 1-0
Record when Mike Trout doesn't start: 0-3
Therefore, my incredibly oversimplified analysis of the situation is that the Angels are now incapable of winning games without Mike Trout.
Hey, we still use pitcher wins, right? Aren't Trout Wins just as legitimate? Aren't they?
No, of course not. Whether a batter is in a game that his team wins is even MORE prone to luck than the starting pitcher. The batter only gets abour 4 or 5 at bats in a game, plus however many defensive plays he may be involved in. A starting pitcher will usually be somewhere between 25-30 batters faced. A batter could go 4 for 4 with 4 home runs, and still lose 15-4 if everyone else on his team stinks. Pitcher wins are dumb. Batter wins are way, way dumber.
Still, the fact that not starting Mike Trout probably isn't the reason the Angels keep losing when they don't start him doesn't mean they shouldn't be starting him. (Did I put enough negatives in that sentence for you? Did it hurt your head? See, that's how I feel when I read about pitcher wins. Feel free to proceed).
There's a tool called "Win Probablity Added" going around the stat community. It does exactly what it says - measures each thing you do on the field and how much it changed your team's chances of winning.
I'm not a huge fan of the tool as a judge of "value" - it's kind of like RBI for us stat nerds, considering the situation and not the event. A player who hits a home run in the 8th inning of a tie game didn't do anything more valuable than the one who hits a home run to lead off the game, but it will obviously change the probability more. Because of this, it falsely inflates the "value" of closers, and devalues the value of starting pitchers who are on teams with a good offense - C.C. Sabathia giving up 1 run in 8 innings and winning 9-1 pitched just as well as Jered Weaver giving up 1 run in 8 innings and winning 2-1. FanGraphs directly states that it has no predictive value, and that there are better stats to use for raw player evaluations.
What WPA does well, on the other hand, is enhance the narrative of individual games. Joe Posnanski gives a nice post about that here, discussing the top 10 games by WPA.
(Aside. That #10 game, where Dwight Evans hit a home run to tie the game in the 8th inning, and then won it in the 10th, is my favorite non-playoff game in my Red Sox watching lifetime, and will be getting a follow-up post.)
Looking at the WPA for some of the great games in history is pretty fascinating. For example, I remember watching Game 4 of the 2001 World Series - the game where the Yankees beat the Diamondbacks with two home runs of Byung-hyn Kim, first by Tino Martinez in the 9th inning, then by Derek Jeter in the 10th. That's one of the most memorable game endings in recent history, but it's not what I remember it for. Nope, I remember that three times in that game, Tony Womack led off his inning with a hit. He hit singles in the first and third inning, and a double in the fifth. In EACH of those circumstances, Bob Brenly had Craig Counsell bunt. And, in EACH of those innings, the Diamondbacks failed to score any runs. I remember thinking how stupid it seemed that Brenly would waste all of those outs. Well, WPA confims it - those sacrifice bunts increased the Yankees chances of winning by 2%, each. WPA is not done retroactively - it doesn't recalculate based on what Martinez and Jeter did later on. It simply tells us what the consequences of a play are on the game and how they change the game at that time. What they tell us, in this situation, is that Brenly killed his team. Counsell had a .359 OBP in 2001 (from 2000-2005, it was .348, so that's not fluky). So, it stands to reason that, in three plate appearances, Counsell would get on base once. Or, in other words, NOT MAKE AN OUT once. Instead, because of Brenly's strategy, he made an out all three times. On purpose. We can't say for sure if the Diamondbacks lost because of that, but we know that their bad strategy didn't help them and that they lost by one run.
Wait, wait, I thought we were talking about Mike Trout? Where is this going, why are you writing about a playoff game from 10 years ago?!
Sorry, I got a little carried away there. I still can't think of Bob Brenly without thinking about that game.
Anyway. Mike Trout. WPA. Right.
On Friday, the Angels lost to the Rangers 11-7 and fell three games behind. They had been losing 11-0, but the back end of the Ranger bullpen apparently decided to give them a sporting chance. What did the non-Trout Angels' outfielders do?
LF Vernon Wells: 2-3, 2B, R; WPA +.002
CF Peter Borjous: 1-5, CS; WPA -.084
RF: Torii Hunter: 1-3; WPA -.061
TOTAL: .364/364/.455; WPA -.143
Recap: It's fair to use this as an example of why WPA isn't so hot sometimes. The Angels lost this game because Dan Haren stunk.
Trout Effect: Minimal. Though he got into the game late as a pinch hitter, going 1-2 with a single, the outfield wasn't the reason they lost. The end of the first inning, where Hunter struck out while Borjous was caught stealing while it was still a 0-0 game is why the WPA for the group is so bad.
On Saturday, Mike Trout started, and the Angels won 8-4. As a team, they hit 5 homers and moved back to two games behind.
LF Vernon Wells: 2-4, HR, 3B, BB, R, RBI; WPA +.127
CF Peter Borjous: 1-5, HR, SB, RBI, 2 R; WPA +.035
RF Mike Trout: 2-4, HR, R, RBI; WPA +.037
TOTAL: .385/.429/1.231; WPA +.199
Recap: Howie Kendrick and Bobby Wilson homered too. It's fair to say C.J. Wilson wasn't sharp in this one. Still, the Angel outfielders played a major role in the one.
Trout Effect: MAKES HIS TEAMMATES BETTER. KNOWS HOW TO WIN. Yeah, back to earth. He had a very good game, and was one of the reasons the Angels won.
On Sunday, Trout did not start, the Angels lost 9-5 and fell back to three games behind.
LF Vernon Wells: 1-4, RBI; WPA +.006
CF Peter Borjous: 0-4; WPA -.065
RF Torii Hunter: 0-3, BB; WPA -.060
TOTAL: .091/.167/.091; WPA -.119
Recap: The Angels were winning this one until Jered Weaver blew up in the 7th. The +.008 of Wells' two out RBI in the 3rd was largely negated by a pop out in the first with two outs and runners on first and second. Torii Hunter's strikeout in the third with a man on second and one out while leading 2-1 wasn't a highlight either.
Trout Effect: Notable. Despite the final, this was a close game for much of the way. A couple outs made by Hunter and Wells with runners in scoring position allowed Colby Lewis to stay in the game through six, which got the ball to the good part of the Ranger bullpen.
Last night, the Angels lost to the lowly Mariners, 5-3. With the Rangers off, the Angels fall to 3.5 games back. Jokes about the patheticness of the Mariner offense may now be passe. Allowing Dustin Ackley and Mike Carp to settle into the 3/4 spots in the lineup is making a big difference. Their .416 SLG as a team in August is 60 points higher than it was any other month, and their .329 OBP ranks 5th in the American League. Sleeper pick for 2012? Not quite ready to jump on that bandwagon, but with their pitching, getting an average offense can make them a serious contender.
LF Vernon Wells: 1-4, R; WPA -.074
CF Peter Borjous: 1-3; WPA -.146
RF Torii Hunter: 1-4, R; WPA -.052
TOTAL: .283/.283/.283; WPA -.272
Recap: Borjous's line drive double play when it was a 3-3 game in the 7th was a killer in the WPA department. The singles all came early in the game, limiting the WPA.
Trout Effect: Moderate. The guy DOES have a .426 SLG, doesn't he? This looks for the world like the kind of game where they needed an extra base hit to swing the balance.
So, over the last three days, Mike Trout has a WPA of +.037, while the rest of the Angel outfielders are at -.372. As we mentioned about, that has no predictive value, only narrative value. So, maybe we can't conclude that playing Mike Trout more would have helped them. What we do know is that, on the whole, the guys they did play didn't help them.