Tuesday, May 18, 2010

Is It Luck?

There is perhaps no more polarizing argument in the baseball statistics world than the argument over "clutch." On the one hand, there's disagreement about what clutch even means, whether a batter up with the bases loaded in the 5th inning of a tie game is in as clutch a situation as a batter up in the ninth inning with two outs, his team down by three runs. The latter is a save situation, of course, and making an out ends the game. But there's a much better chance of actually changing the outcome of the game with the bases loaded in the 5th inning of a tie game. A double, for example, will likely give you a three run lead, where a double in the 9th will put a runner on second, still with two outs and down by three.

On the other hand, there's great disagreement about whether certain players can raise their game to perform better in clutch situations. Is it not the case, for example, that Derek Jeter simply plays better with his team down by a run in the 9th inning? Doesn't he become a better hitter? Likewise - and this is the other side of clutch - doesn't Alex Rodriguez suddenly forget how to hit home runs in the same situation, and turn into a puddle of self-doubt and failure?

The tricky thing, here, is trying to tell the difference between players who perform better in the clutch and players who perform better, period. Of course you want Albert Pujols or Derek Jeter up in the 9th inning with the game on the line, not necessarily because they are clutch, but simply because they are good hitters. The real question of clutch is not how well players perform in high leverage situations, but how much better or worse they perform in those situations relative to how they perform in normal situations.

Fortunately, Fangraphs keeps track of clutch. Their resolution of the first question - what counts as clutch - is based upon the very statistics that they pioneered: Win Probability Added (WPA) and Leverage Index (LI). Basically, for each and every major league game, Fangraphs keeps track of game situation and calculates the odds each team has of winning based on score, inning, outs, and runners on base. From there, they extrapolate the importance of a given situation. Intuitively, a high leverage situation is like the one I described at the outset: bases loaded in the 5th inning of a tie game. That's high leverage because the outcome of the next at bat will have a tremendous influence on the odds of each team to win the game. A strikeout will be huge for the pitching team, while a walk or a hit will give the batting team a run and the lead (and, again intuitively, the team currently with the lead is almost always favored to win the game).

Anyway, the point here is not to dive into WPA or LI, but rather to talk a little bit about clutch and the Colorado Rockies. As in years past, the Rockies are off to an awesome start pitching this season, but can't seem to hit to save their lives. That their defense has been suspect as well has helped to lead the team to a disappointing early-season record of 19-19. Not the end of the world, but .500 isn't about to catapult anyone into the playoffs, and playing .500 ball so far means the Rockies have to play way better than .500 the rest of the way to earn their first ever division title (or at least their first ever back-to-back playoff berths).

What ails the Rockies lineup, however, is not totally clear. While they have lacked star-power this year, what with Todd Helton and Troy Tulowitzki forgetting how to hit home runs, Miguel Olivo has been a pleasant surprise (at least so far), Carlos Gonzalez has been excellent, and Seth Smith might finally be working his way into the everyday lineup. The Rockies don't have the talent to have a top-flight offense, but they should be competent.

Except they can't hit in the clutch.

Whether clutch is a skill or not, the numbers are striking. The Rockies currently rank 28th out of 30 MLB teams in clutch hitting this season. That is, the Rockies perform significantly worse in high leverage situations than in normal ones, so much so that only two other teams (the Cubs and Brewers, incidentally) have been worse so far in 2010.

It is early in 2010, but what disturbs me is where the Rockies stand looking backwards to 2003, Clint Hurdle's first full season:

2009 - 27th
2008 - 23rd
2007 - 5th
2006 - 30th
2005 - 28th
2004 - 24th
2003 - 25th

Of course, 2007 was the year the Rockies made the World Series after an improbable run in September and October, and it stands out as an outlier in this list. What jumps out to me, though, is that the Rockies are really the only team routinely this low on the clutch list at Fangraphs. What are they doing wrong? Or is it just luck?

Any regular watcher of the Rockies over the last 10 years will confirm anecdotally what Fangraphs has in numbers: the Rockies routinely fail to produce in key situations. Back-to-back strikeouts with a runner on third and nobody out has been something of a theme for this organization. Again, is this an organizational problem? Is it a hitting coach problem? A manager problem? A personnel problem? Does it have something to do with Coors Field? Or is it just luck?

Obviously in 2009 the Rockies were fine without being a good team in high leverage situations, and Troy Tulowitzki - who has struggled to produce at his normal level in the clutch throughout his career - is still a fine player despite often failing in high leverage situations. There's also no reason to believe that Tulo, or the Rockies in general, will continue to struggle producing when it counts. And yet, as a fan, the trend is disturbing, and as a human being, I'm inclined to attribute the pattern to something.

No comments:

Post a Comment