Josh Hamilton had just about as bad a game as a hitter can have last night, going 0-5 with two strikeouts and three GIDPs (with Mike Trout on base, no less). I was curious about just how bad it was, so I took a look at his WPA for the game. For those that don't know the stat, WPA stands for win probability added, and gives you a rough idea of the sum of a players contributions to the victory or loss. A team's expected winning percentage starts at 50%, and WPA looks at how each player's results change that. For example, if Jacoby Ellsbury hits a lead-off home run to start a game, and the Red Sox expected winning percentage goes from 50% to 65% (these numbers are totally made up), that plate appearance would be worth 0.15 WPA. It's a counting stat like runs or RBI, in that someone with more at bats or innings pitched will be able to accumulate more. It's also context dependent- a grand slam down three runs is worth a hell of a lot, while a grand slam up five runs is worth very little (as the team is already very likely to win the game).
Hamilton's performance was worth an amazing -0.477 WPA. That performance was so putrid it really got me thinking. First off, it's just amazing that a player can make eight outs in five at bats. Second, it got me wondering just how bad this performance was, in the grand scheme of things. Using Fangraphs custom leaderboard, I took a look at the players that had accumulated the most -WPA (that is, only the negative portion of WPA). I figured this would be a good proxy for players who have really bad games, but it's not ideal since WPA is a counting stat, so I'll end up with guys who have bad games and still have playing time. Unfortunately, I can't do a leader board for individual game WPA.
Here's what the "top" 10 looked like:
1 Melky Cabrera Blue Jays -6.23
2 Adam Dunn White Sox -6.20
3 Martin Prado Diamondbacks -6.13
4 Elvis Andrus Rangers -5.97
5 Josh Hamilton Angels -5.96
6 Alexei Ramirez White Sox -5.95
7 Mike Trout Angels -5.93
8 Alcides Escobar Royals -5.91
9 Jacoby Ellsbury Red Sox -5.82
10 Starlin Castro Cubs -5.80
Our man Josh Hamilton slides in at five, joined by fellow Angel and surprise pick Mike Trout at seven. Mike Trout's name is interesting here, but he does have an overall positive WPA. It's also important to note that WPA is not really predictive- it's a good way to look at what has happened, but that's about it.
Anyway, I took a look at the game logs for each player, and other than Hamilton, none of the top 10 managed to even break -0.4. Travis Hafner did have a terrible game that amounted to -0.53 WPA, but it was a game that ran 18 innings. I also took a look at Aaron Hicks, since he had such an awful start, but even he only managed a worst game of -0.3 WPA. Hamilton's game last night was really, really, really bad, but I suppose these things happen when you use up THIRTY PERCENT of your team's outs while managing to accomplish nothing at all.
5 comments:
I forgot to note that this is hitters only. Pitchers can easily have much worse games, particularly relievers. Tom Wilhelmsen, for example, managed a cool -0.9 WPA when he blew a three run lead against the Mariners and only managed to record one out.
Wow. Glad I traded that guy!
WPA is funny. Earlier this year Mike Trout hit a super-cycle (normal cycle plus a stolen base), which is an incredibly rare event. Trout's was only the 17th in MLB history, which makes the super-cycle more rare than the perfect game. Amazingly, Trout also had a negative WPA for this game. That was the first time anybody ever hit for the cycle and had a negative WPA. Basically, he struck out in the 1st inning when the score was still 0-0, and then his 4 hits all came after the Angels had a significant lead and did not contribute much WPA.
Yeah, WPA is kind of meaningless in a lot of ways, since so much value is context dependent. In this case, Hamilton was really bad in a game that was lost by one run. And he came up with lots of opportunities to drive in runs.
Here's the situations for each at bat:
1 man on base, double play
1 man on base, double play
2 men on base, double play ends inning
1 man in base in scoring position, strikeout to end inning
2 men on base, strikeout to end inning
Even his strikeouts ended the inning! He couldn't even hit into a double play those innings since there were already two outs! It's a perfect storm of suck. Seriously. Even the third double play couldn't have been a triple play! Given the situations, he could not have performed worse (unless he injured Mike Trout with a flying bat or the like).
It's interesting to think about how it's actually possible to have a single-game WPA of less than -1, or greater than 1. Though it would have to be in a pretty wild game with a lot of lead changes.
It would have to be. The highest or lowest a team's WPA could be is +0.5 or -0.5. That would mean some guys did really well, and other guys did really, really, really bad.
Post a Comment