D.O.S. (Death of Statistics), Pt. II

Hold up, only writer to rewrite history without a pen
No I.D. on the track, let the story begin, begin, begin

Of all the aspects of baseball, none is more divisive than pitching. There are dozens upon dozens of metrics that range from projected ERA to drop on a curveball. But the issue is not which statistic is most prevalent, the problem is a systematic fissure between old school and new school ways of thinking.

As I began to explain in the first part of this four-part series, there are a lot of people stuck in the back-of-the-baseball-card state of mind. What that means is that we can look at a handful of simple, long-standing stats (batting average, homers, RBIs, runs) and tell how good that player was. But this just doesn’t work. Not only do those stats not tell the whole story, they don’t even tell the story well. Let me explain.

This is the Death of Statistics, moment of silence.

When it comes to pitching statistics, one usually comes to mind first. Sadly, this is the worst pitching stat out there. I’m, of course, talking about the win.

Let me be clear, there is nothing more important in sports than winning the ball game. But the pitcher’s win-loss statistic has no statistical relevance. Ultimately, the stat comes down to how well your teammates play.

I have no issue with Felix Hernandez winning the Cy Young in '10. Maybe Seattle's beyond-anemic offense had something to do with his 13-12 record.

Take, for example, Justin Verlander and Cliff Lee this season. Both had ERAs of 2.40, but Cliff Lee received 4 runs of support per game compared to Verlander’s 4.56 runs of support. All of a sudden, Lee owns a measly 17-8 record compared to Verlander’s “other-wordly” 24-5 record. Did Lee give up any more runs than Verlander? No. His teammates just didn’t play as well.

Wins don’t add anything to the analysis of a pitcher. If you need look at control, look at WHIP or K/BB. If you’re interested in the ability to miss bats, look at Contact Percentage or Line Drive Percentage. But when comparing players, you cannot take the skills of teammates into account, which is the central point of the Pitcher Win-Loss stat.

In the first part of this series, I wrote about the worst stats: the ones that are subjective and the ones you can’t explain. Well, to some extent, the second most recognizable pitching stat, ERA, is subjective. The goal of the statistic is simple: show how many runs the pitcher gives over nine innings. The problem, however, is that errors play a huge role in earned runs, which is the center of the average.

I don’t want to totally dismiss ERA, I think it is still a very useful and mostly representative stat. But since it relies heavily on the error and the play of teammates, it should be taken with a grain of salt. So the best pitching statistic, I’ll say it again, are those that remove the play of teammates.

When pitching, there are only three outcomes that are completely dependent on the pitcher: a strikeout, walk, and home run. The rest are dependent on the skills of the fielders, which cannot be attributed to the pitcher himself. Whether a long fly ball, dribbling ground ball, or liner in the gap, the play is out of the hands of the pitcher once the ball is hit in play. And because of that, a myriad of statistics have been formed to project ERA.

There is FIP, xFIP, SIERA, PECOTA, DIPS, all sorts of advanced defensive metrics. At the core of all of them is taking the skills the pitcher has and turn it into an ERA predictor. My personal favorite is FIP, which is the simplest, and in my opinion, the most effective. By taking the three pitcher controlled stats, these sabermetric stats better illustrate the skill of the pitcher. After all, how can you penalize Mike Pelfrey for playing in front of nine defensive butchers while Big Game James Shields gets to play in front of a cavalcade of elite defenders?

When it comes to the Cy Young Award, I didn’t have much of a problem. Justin Verlander was by far the most dominant pitcher in the AL, although the national attention paid to his 24-5 record blew his dominance slightly out of proportion. Still, CC Sabathia and Jered Weaver put up good fights, but it’s hard to come up with a solid argument over Verlander. He posted an AL-best 2.40 ERA, 4th best 2.99 FIP, and second best 7.1 WAR. Not only that, but he stranded a ridiculous 80.3% of runners, which were few and far between thanks to a 0.92 WHIP and .191 batting average against.

It’s on the NL side, however, that I had a slight problem. Clayton Kershaw, one of the fastest rising pitching stars, won the “pitching triple crown” with the most wins, strikeouts, and lowest ERA. We’ve already be over the (un)validity of the win and the problem with ERA, which rather tempers the excitement over the rare feat. Kershaw may have bested Halladay’s ERA 2.28 to 2.35, but Halladay takes the cake for FIP (2.20 to 2.47) and WAR (8.2 to 6.8).

I’m not terribly upset by the results of the NL Cy Young, but I do hope that voters think again before casting votes, infatuated with out-dated stats like wins and saves (I’ll save this argument for a later date). It’s time to move past the subjective stats and those dependent on teammates. Because the only way you can truly compare players statistically is looking at hard data of those players alone.

This is the Death of Statistics, moment of silence.

La da da da

Hey Hey



Categories: MLB | Leave a comment

Post navigation

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at WordPress.com.

%d bloggers like this: