Your LeBron James Elbow Update
April 29, 20102010 NBA Playoffs Eastern Conference Semifinals Preview: Cavaliers vs Celtics
April 29, 2010Just because we were told this would be a rough year for the Indians, that shouldn’t stop us from whining every now and then. Today we’re going to open a whole new bottle of whine (see what I did there?), and along the way, we’ll hopefully develop a better idea of the players who are most at fault for the sub-standard performance so far this season. A quick note: the stats in this piece were compiled on Tuesday, when the team’s record was 8-11. That may come in handy to those of you following along at home.
A few weeks back, we discussed the concept of Wins above Replacement (WAR). Basically WAR tells us how many wins a player generated for his team over and above what a cheap AAA player would have contributed. But as you might remember, WAR is a context-independent measurement: it doesn’t value a two-run HR late in a one-run game any more than a grand slam when your team is already leading by 10. The skill (hitting the HR) is what’s being measured, and you don’t take into account when the event took place in the course of a game.
This is mostly a good thing, because it’s been demonstrated over and over again that there really isn’t such a skill as “clutch hitting.” In other words, no one has been able to demonstrate that certain players have an ability for hitting better when the game is on the line than when it’s not. So you cut the context out and just measure the “skill.”
But sometimes we want the context, especially when we’re trying to evaluate how and why our team won or lost a game. Furthermore, when you talk about bullpens, context can be very important. Late in a close game, the game is certainly “on the line”, and in these games the bullpen’s performance becomes of utmost importance.
So let’s talk about a different stat, called Win Probability Added, or WPA for short. WPA measures the impact any given player had on affecting his team’s chance to win a game. At the beginning of any game, each team has a .500 chance of winning the game, and by the end of a game, the winner has added .500 points, moving to 1.000, while the losing team moved to zero. WPA measures each player’s contribution to this movement, and, when added over the course of season, WPA tells you how many wins—above average—a player contributed to his team.
WPA works by measuring how much more or less likely a team is to win after each plate appearance. In this way, any points a pitcher gains, a hitter loses—and vice versa. A nifty zero sum game, as it were. And the measurements are done using a win expectancy matrix that measures how likely a team is to win a game given the score, the number of outs, and the number of baserunners. We looked at these matrices when we discussed the strategy behind bunts and stolen bases, if you remember.
Let’s use a recent Indians game to illustrate what I’m talking about. Here’s a WPA chart for the Indians’ 5-3 extra-inning victory against the White Sox on April 8th:
In the first couple of innings, the White Sox did more to improve their chances of winning than the Indians did by getting runners on base. In the bottom of the second inning, we see a slight spike in the White Sox likelihood of winning the game. That’s because Andruw Jones was hit by a pitch from Masterson, and then Pierzynski singled to move Jones to third with one out. Since their chances of scoring (and winning) went up with each of these events, those players were credited with positive WPAs while Masterson was debited the same amount. Jones getting hit awarded him .026 WPA and Pierzynski’s single gave him .066 WPA. For those plate appearances, Masterson lost .026 and .066. Even though the Sox threatened in the second, they didn’t score, so you see each team’s chances go back at 50-50 as the third inning begins.
As most baseball fans know, all plate appearances are not created equally. Look to the 11th inning. The game was tied at three when Asdrubal Cabrera singled to drive in Luis Valbuena from second. Before Cabrera’s single, the White Sox had a 53.7% chance of winning the game. After Cabrera’s RBI, the White Sox had only an 18.8% chance of winning. So here’s what his single was worth in WPA:
.537 – .188 = .349.
What’s my point? Well, Cabrera’s single was much more valuable than Pierzynski’s was in the second inning (.349 for Cabrera versus .066 for Pierzynski), because it did more to affect his team’s chance of winning, both because it was later in the game and it added a run. In this way, WPA accounts for context and “clutch” performances. While it’s not necessarily a great tool for predicting future performance (since clutch “skill” doesn’t really exist), WPA does a great job of telling us how much a particular player has contributed to his team’s likelihood of winning or losing. And for this reason, it helps us assign praise or (in the Indians’ case) blame to the right players.
With that in mind, let’s take a look at the Indians’ bullpen performance to date since most people seem to think they’ve pitched terribly. Here are the ‘pen’s WPAs:
Name | WPA |
Jensen Lewis | 0.43 |
Tony Sipp | 0.35 |
Aaron Laffey | 0.18 |
Jamey Wright | -0.04 |
Joe Smith | -0.05 |
Rafael Perez | -0.08 |
Chris Perez | -0.79 |
TOTAL | 0.00 |
Individually, none of these WPAs is terribly surprising. Keep in mind that an “average” player would have a WPA of 0—he would neither help nor hurt his team win games. A team full of the “average” players would win exactly 81 games over the course of a season. For a guy who I’ve never completely been sold on, Jensen Lewis has performed admirably so far this year, contributing 0.43 wins above what an average player would have contributed in the same situations. Tony Sipp has also helped to create more wins than losses for the year.
On the other hand, Chris Perez has been a bit of a disaster in the closer’s role so far. Through the first three weeks of the season, Perez has cost his team considerably: if he had performed at merely an average level (zero WPA), the bullpen would have a combined WPA of 0.79; instead, the bullpen has a WPA of exactly zero. In other words, Chris Perez’s performance has single-handedly taken the bullpen from above average to just…average.
But what interests me isn’t that Chris Perez hasn’t performed well—everyone already knows that. Rather, I’m surprised to see that the bullpen as a whole hasn’t been where most of our losses have come from. They’re performing at an average level—or basically the level you’d expect from a .500 team—but I get the feeling that most people feel like the bullpen has cost us most of our losses, without remembering that they’ve been integral to plenty of our wins as well. Those middle innings that the bullpen ate up in the first and second weeks of the season were crucial to keeping us in close games.
So let’s look to see which component of the team—bullpen, starters, or position players—has helped or hurt our chances of winning thus far. Here are the starters’ WPAs:
Name | WPA |
Fausto Carmona | 0.59 |
Mitch Talbot | 0.25 |
Jake Westbrook | -0.20 |
David Huff | -0.23 |
Justin Masterson | -0.54 |
TOTAL | -0.13 |
As a whole, the starters have negatively affected the team’s likelihood of winning games, David Huff and Justin Masterson have been the two biggest offenders so far this year, but other than those two, the starters haven’t been terrible. The rotation’s -0.13 WPA for the season puts them at slightly below the level a .500 team would have, but I’d say most of us expected that this season. It’s hard to get too worked up about a mediocre pitching staff pitching, well, mediocrely.
But more importantly, the starters haven’t been the team’s biggest problem. As of this writing, the Indians sit at three games below .500. That means that their combined team WPA should be -1.500 (-.500 for each loss and .500 for each win). And only .13 of that -1.500 has been generated by the pitching staff, and none of it by the bullpen. So guess who’s next?
Here are your “sluggers” measured by WPA:
Name | WPA |
Shin-Soo Choo | 1.13 |
Austin Kearns | 0.19 |
Grady Sizemore | 0.09 |
Russell Branyan | 0.04 |
Mark Grudzielanek | 0.03 |
Andy Marte | 0.01 |
Luis Valbuena | -0.01 |
Michael Brantley | -0.17 |
Asdrubal Cabrera | -0.20 |
Lou Marson | -0.26 |
Mike Redmond | -0.29 |
Jhonny Peralta | -0.45 |
Travis Hafner | -0.55 |
Matt LaPorta | -0.92 |
TOTAL | -1.36 |
And this is the real point. Our pitching staff is keeping us in games; their combined performance so far this season would basically translate to a .500 record if they were being complemented with anything resembling an “average” offense. But our hitting has been flat-out awful.
Sure, Choo by himself has been responsible for about one extra win over what an average hitter would have produced, but the rest of that list looks pretty anemic: when Austin Kearns, a part-time player, is your second most productive hitter, you’re gonna have some problems.
One thing to keep in mind before we make too many conclusions: WPA incorporates a lot of context. So just because Matt LaPorta has the worst WPA on the team, that doesn’t mean he has the worst batting average (that distinction belongs to Sweet Lou Marson) or the worst slugging percentage (Marson again; funny how that works) or the most strikeouts (surprisingly, Choo). What it does mean is that LaPorta has come up short in big moments: when he had a chance to win a game or drive in a big run, he grounded into a double-play or struck out more often than not. And because of those performances, he’s cost his team about a win compared to what an average player would have. In other words, these charts don’t tell you how good a player is—that’s the job of WAR—but rather how much that player has contributed to his team’s won-loss record. One tells us what to expect going forward (WAR); the other tells us how to interpret what has already happened (WPA).
So if you’re looking for a group to blame for the slow start, you might want to consider the position players. And if you’re wondering how we know what we know (epistemology!), keep an eye on those WPA charts.
Until the offense can get it going, it’s silly to fault our pitching staff for the fact that we’re losing more games than we’re winning. Next time, we’ll take a look at the offense as a whole and a few important players to figure out some of the whys behind this shaky start.
See you then!
-Jon Steiner
* Sorry for the nerdy title; the Indians’ poor play has required me to entertain myself lately, and titular references to epistemology are all I’ve got.
7 Comments
That was a mind-numbingly great article!
Ugh. Not about the article, which was great, but the fact that every position player but one is below .2, putting them all below Talbot, Carmona, Sipp, Lewis, and on par with Laffey. Bleah.
Don’t know much about history/
Don’t know much about epistemology . . .
(Seriously, that’s an incredible article)
“In the first couple of innings, the White Sox did more to improve their chances of winning than the Indians did by getting runners on base.”
Wait, so THAT is the trick?
Just wanted to give an update on the numbers, since I put those together a few days ago. Here are the current WPAs:
Bullpen WPA = -0.05 (Would be significantly positive, but offset by C. Perez’s -1.14)
Rotation WPA = 0.14 (Led by Talbot and Fausto, both above 0.50)
Hitter’s WPA = -1.59 (With Choo at 1.25 and Kearns at 0.40 and everyone else less than 0.08 and A LOT of negative numbers)
So our total team WPA is still at -1.50 (since we’re still 3 games below .500). The offense (-1.59 WPA) is still MORE THAN completely responsible for the losing record.
That’s just gross. I’m glad to see that my anger towards our hitters hasn’t been in vain. If the Indians locker room was a high school, leaving runners on base would be even cooler than smoking or pretending like you had sex.
I coulda told you that… All in all, if we were scoring close to a modest amount of runs per game (5 or 6, or even 4) we would probably be over 500 right now, as our pitchers have NOT really been that bad! Even C. Perez would have had more chances to pitch, giving him more confidence and experience in the role, resulting in a better WPA for him. Our bats looked like they may have started to wake up in the last 2 games in Anaheim, so here’s hoping for a heavy-hitting homestand.