Posted in MLB, Philadelphia Phillies, Sabermetrics | Print | 8 Comments »
This subject has been covered ad nauseam already, but I’ve had this written for a while now and I don’t want it to go to waste. Originally, it was going to be published in an e-book, but that fell through. As a result, you get to enjoy the content at no cost. If you frequent the blog, the concepts need no introduction, but I originally wrote it for a less Sabermetrically-inclined readership.
Cole’s Curious Conundrum
If you were conscious at any point from the second half of the 2009 regular season through the playoffs, surely you heard the incessant armchair psychoanalysis pertaining to Cole Hamels, the Phillies’ left-handed pitching phenom and MVP of the ’08 World Series. There was something askew with him the whole season. He spent too much time in the off-season posing for magazines, appearing on television shows and lending his persona in advertisements. He got cocky; he wasn’t focused; he was immature; he lacked the mental strength to go through another 162-game grind after reaching the pinnacle of baseball. Right?
The funny thing is, the Cole of ’09 wasn’t much different than the Cole of ’08 — if there is any difference at all. Sure, his ERA was nearly one and a quarter runs higher and his WHIP increased by .2, but aside from superficial statistics like those, Cole appeared to be surprisingly… himself.
I know that is hard to believe after watching him achieve mediocrity in his four starts during the ’09 playoffs, where he never once finished the sixth inning — something he did each and every time in his previous six post-season starts. However, the data analysis that will follow proves that the Cole you saw last year was essentially the same guy that took the bump two years ago. Hopefully, upon reading this article, your Hamels-related fears are assuaged.
Hamels won 10 games and lost 11, the very definition of mediocre. However, we all know by now the fallacy of the won-lost record. Along with the 21 decisions, Hamels had 11 no-decisions. In those ND’s, the team went 6-5 and Hamels allowed 26 runs in 66 and one-third innings (3.53 ERA). The bullpen, meanwhile, allowed 17 total runs (including unearned) in 14 innings (10.93 RA) in the no-decisions that resulted in Phillies losses.
Hamels’ ERA overall was ugly but it is reasonable to think that if the bullpen had been more helpful to him that he would not have been exposed to such harsh public criticism from the media and fans. While the expectations are different, Joe Blanton finished 2009 with 12 wins and an ERA above 4, yet fans were ambivalent about the prospect of trading him to create payroll space during the off-season.
When we analyze pitching data, we have to discriminate between what pitchers can and cannot control. On the statistic BABIP (batting average on balls in play), J.C. Bradbury of the blog Sabernomics writes, “While pitchers may have some ability to prevent hits on balls in play, the effect is small. And any effect a pitcher does have is reflected within DIPS metrics.” (DIPS stands for Defense-Independent Pitching Statistics, which removes the effect of the pitcher’s defense.) Essentially, if a pitcher isn’t any different from one year to the next, but has a significant change in his BABIP, we can attribute that difference to luck, for lack of a better word. This is the BABIP equation, taken from its Wikipedia entry:
Hamels in 2008 enjoyed a .270 BABIP, which is about thirty points under the general average of .300. He finished with a 3.08 ERA and arguably deserved consideration for back-end votes in National League Cy Young balloting. 2008 was very good for Hamels and really, really good for his Phillies.
Last season, that BABIP went way up to .325. For as far below the average it was in ’08, it was nearly equally as far above in ’09. Hamels finished with a 4.32 ERA. Instead of singing his praises, Phillies fans were pondering an appropriate ranking for him among the most disappointing Phillies of ’09. During the off-season, some fans even speculated that the Phillies should consider trading him as a means to both acquire Roy Halladay and keep Cliff Lee.
Is BABIP entirely to blame for Hamels’ failures last year? No, not entirely, but it looks like it had a huge impact on his fortune the past two seasons. That becomes more evident when we utilize ERA estimators such as FIP, xFIP, and tRA. Before the actual numbers, we need to get the equations out of the way. Don’t worry, you won’t need your calculator as FanGraphs.com does all the heavy lifting.
- FIP = (HR*13+(BB+HBP-IBB)*3-K*2)/IP
- xFIP = ((FB*.11)*13+(BB+HBP-IBB)*3-K*2)/IP
- tRA (the FanGraphs version) “involves assigning run and out values to all events under a pitcher’s control and coming up with an expected number of runs allowed and outs generated in a defense and park neutral environment”.
When we account for what Hamels himself is directly accountable for, we see that he has been the same pitcher the past two seasons. In fact, according to FIP, he has been exactly the same. xFIP has him as very slightly worse last season while tRA has him as slightly better.
It’s not enough to simply point to a few ERA estimators and wrap up the analysis. Thanks to the plethora of statistics available at FanGraphs, we can actually look at the characteristics of Hamels from year-to-year. First, a look at his strikeout, walk, and home run rates per nine innings:
Nearly identical numbers. Moving onto his batted ball rates:
Hamels allowed the same amount of fly balls and “turned” 1% of his line drives into ground balls. Additionally, he allowed 0.5% fewer home runs per fly ball. If anything, opposing batters were performing worse against him. It should be noted that it is highly likely that the small changes in those rates are due to random fluctuation, not to anything inherent to Hamels’ skill.
Finally, a look at the characteristics of the batters Hamels faced:
The prefix O- stands for pitches outside of the strike zone and Z- stands for pitches inside the strike zone.
What this table tells us is that opposing batters were swinging less outside the zone 4% less from ’08 to ’09, yet made contact nearly 2.5% more. However, those same hitters swing nearly 2% more at pitches inside the strike zone yet made contact nearly 4% less. Overall, Hamels was in the strike zone 0.6% less in ’09 than in ’08.
The following graphics, using Pitch F/X data, will display each of Hamels’ three pitches from both an overhead and a side view from 2007-09. Click each image to view a larger version.
The above graphic simply combined the three previous graphics into one, so it is easier to spot the differences. Frankly, there’s not a whole lot of difference. There are subtle differences but that would be due to randomness as no one pitcher’s averages would be identical from one year to the next.
The graphic below shows the release point of each of Hamels’ pitches from 2007-09.
Again, not a lot of difference. If anything, Hamels was more consistent with his release points in 2009 than in ’08. Check out the messy release points beyond the positive 1.5 line on the horizontal axis in 2008.
We have looked at BABIP; ERA estimators; strikeout, walk, and home run rates; batted ball rates; and Pitch F/X data. The only significant difference found is his BABIP. It is conclusive that Hamels was just about the same pitcher in his amazing 2008 season as he was in his disappointing ’09.
Does this mean that Hamels will be as good in ’10 as he was in ’08? No, it does not. For as unlucky as Hamels was with BABIP last year, he was about as lucky in ’08. Hamels’ true talent level lies somewhere in between the two seasons. This is echoed by the projections on FanGraphs.com. Bill James puts Hamels on a 3.43 ERA in ’10; CHONE at 3.80; and the fans at 3.51. For comparison, the following pitchers had ERA’s in the 3.40-3.60 range last season: Jon Lester, Justin Verlander (who finished third in AL Cy Young voting), Ubaldo Jimenez, and Joel Pineiro.
Hamels is fine, folks! Don’t give up on him because of his 2009 showing.