HOUSTON - Pitcher J.A. Happ #30 of the Houston Astros throws against the Milwaukee Brewers. (Photo by Bob Levey/Getty Images)
One of the most appealing features of sabermetrics is when it tells you that something that you believed really isn't true. Sabermetrics became popular in the Bill James era because it challenged certain "truths" about baseball.
Suppose sabermetrics told you that J.A. Happ may turn out to be a pretty good starting pitcher this year?
Based on comments in the game threads, I think it's fair to say that many Astros' fans don't think that J.A. Happ is a good pitcher---or at the least, they don't trust his pitching. I admit that I probably shared those views. His ERA is almost 5, which means that his results have been bad. But, when I began looking at the Astros' pitchers whose performance is better than his results, I was surprised to find that J.A. Happ fits that description. And it's no contest. He is pitching a LOT better than his results.
Besides the results, I think Happ has some aesthetic reasons for fans' negative opinions. He can slow down to a snail's pace when he gets in a jam. Sometimes he doesn't throw enough strikes. He sometimes accumulates high pitch counts. He is prone to walks. And, well, yes, the walks are more than an aesthetic issue. But aesthetics don't win games, and they can make a pitcher look worse than he is.
Sabermetric Pitching Stats
HAPP 2012 PITCHING STATS
4.96 ERA 4.59 FIP 3.91 x-FIP 3.87 SIERA
Let's start with the fact that Happ's ERA is much higher than all of the sabermetric stats. This tell us that Happ's results probably aren't reflective of his pitching. But I will focus on the two predictive stats, x-FIP and SIERA. SIERA is a complex statistic that takes into account a number of factors, including batted ball type, but is similar in objective to x-FIP. x-FIP is a defense independent pitching stat that normalizes HR rate at the league average.
Despite its complexity, I like the SIERA statistic, which is more predictive of a player's future performance than ERA, FIP, x-FIP, and even some projection systems. SIERA indicates that Happ has been pitching quite well. Happ's SIERA is virtually the same as Wandy's SIERA (3.87 vs. 3.85). Yet Wandy's ERA is much lower than Happ's ERA (2.24 vs. 4.96). Again, Happ's earned run results may not indicate how well he has been pitching this year.
Some people don't like SIERA, given its complexity. So we check it with x-FIP, which is intended to be more predictive than FIP and ERA. And x-FIP gives us a similar result to SIERA. Happ's x-FIP is based on the assumption that his HR per fly ball rate (15.1%) will regress toward league average (10.5%). And Happ's HR/ fly ball rate isn't an artifact of the Crawford Boxes. Happ's HR/fly rate actually is league average at MMP. The liklihood is that Happ's HR/fly rate is a product of luck.
Batting Average on Balls in Play (BABIP) is used to evaluate whether pitchers are likely to undergo regression in their allowed hits. The use of BABIP for such evaluation evolved from Defense Independent Pitching System theory (DIPS), a fundamental contribution of sabermetrics to understanding pitching. DIPS says that most major league pitchers have limited control over batted balls, with the exception of inducing more or less groundballs, HRs, or fly balls. As a result, most pitchers' sustainable BABIP falls within 20 - 25 points of league average BABIP.
Happ's BABIP is .323, which is considerably higher than the average BABIP for LHPs in the National League, .290. His current 2012 BABIP falls outside the normal range around the mean for BABIP. Happ has a career average BABIP of .282, and his BABIP this season is the highest of his career. This strongly suggests that Happ's BABIP is likely to regress downward.
It is possible that an excessive BABIP may reflect poor defense. But, if that's true, it's likely bad luck on defense, rather than a systemic failure. The Astros' team UZR and team DER are both well above average.
Did you realize that J.A. Happ has a 8.5 K/9? One of the knocks on Happ from his Phillies' years is that he achieved good results with a mediocre strike out rate. But his K rate has been increasing each year with the Astros. His current K rate is very good. Happ is ranked 18th among NL starters in K/9. Happ still walks a few too many batters, but he has shown considerable improvement this year. His current 3.57 BB/9 rate is noticeably lower than his career BB/9 rate of 4.05. Are the improvements in Happ's peripherals an indication that he has made real progress as a pitcher? It's early in the season; so we can't really answer that question. But the improved peripherals have to be a good sign.
J.A. Happ has pitched much better than his ERA. If his pitching results regress to his SIERA, and he pitches 170 innings this year, Happ would end the season with a 4.15 ERA. That's pretty darned good---better than the No. 5 rotation pitcher label usually placed on Happ. Perhaps more intriguing is the possibility that he is showing fundamental improvement in his K/9 and BB/9, which might allow him to make a leap forward. This could make Happ a viable trade candidate, who can bring back valuable prospects in the off-season, or, alternatively, make him the best option to replace Wandy's as the veteran LHP in the rotation. Previously I had wondered whether Happ would be worth an arb eligible contract next year. This analysis changes my view of that question.