Knuckleballs

Unpredictable, rare, and occasionally effective…but always entertaining.

Cole Hamels

leave a comment »


In 2008, Cole Hamels was the best pitcher on the World Series champions and was just 24 years old.  Including the postseason, he pitched 262 innings, going 18-10 with a 2.91 ERA and 226 strikeouts against only 62 walks.  In 2009, Hamels pitched 213 innings, going 11-13 with a 4.61 ERA and 183 strikeouts against only 47 walks.   What happened?  Well, just as we’ve seen with other pitchers to this point, these numbers don’t tell the whole story.

W L ERA GS IP H R ER HR BB IBB SO
2008 Reg. Season 14 10 3.09 33 227.33 193 89 78 28 53 7 196
Playoffs 4 0 1.80 5 35 23 7 7 2 9 0 30
Total 18 10 2.92 38 262.33 216 96 85 30 62 7 226
HR/9 = 1.03 K/9 = 7.75
K/BB = 3.65
2009 Reg. Season 10 11 4.32 32 193.67 206 95 93 24 43 4 168
Playoffs 1 2 7.58 4 19 25 16 16 7 4 0 15
Total 11 13 4.61 36 212.67 231 111 109 31 47 4 183
HR/9 = 1.31 K/9 = 7.74
K/BB = 3.89

Hamels struggled mightily in the 2009 postseason, possibly due to fatigue, having thrown almost 500 innings in two seasons.  Removing that from the equation, he gave up home runs at the same rate he did in 2008, struck out the same number per nine innings and walked the same.  His batted ball types stayed within one percent of their previous norms, resulting in the same number of home runs per fly ball as well.  All of his numbers look remarkably the same between the two seasons, except for wins and losses and ERA.  First I examined our old friends BABIP and strand rate to see if these were the major culprits.

Last year, Hamels had a BABIP of 0.325, 22 points above the league average.  It was roughly the same with or without runners on base.  Same goes for 2008, when Hamels BABIP was 0.270, 34 points below the league average.  In 2008, he stranded 76 percent of runners, a number that actually went down to 72 percent (the league average) in 2009.  These numbers tell us that a lot more ground balls and fly balls were going for hits than they did in 2008, and that Hamels wasn’t particularly lucky or unlucky with runners on base in either year, with respect to what happened without runners on base.

I have talked extensively about BABIP to this point, but haven’t really made it clear why I care about it so much.  BABIP (batting average on balls in play) takes away all the things a pitcher has direct control over – walks, strikeouts, and home runs.  Based on extensive studies, BABIP is mostly based on the defense behind a pitcher…and blind luck.   Even if the pitcher does everything he wants, someone still has to field the ball and record the out;  BABIP is more susceptible to statistical noise because of this.  From year-to-year, a pitcher has a different defense behind him.  A pitcher’s home park will also have an effect on his BABIP.

It is possible to sustain higher or lower BABIPs as a hitter or a pitcher (definitely easier as a hitter), but very difficult to do.  In a season’s worth of plate appearances against a pitcher, BABIP has a great deal of randomness.  Over several years though, it can provide a more sure number from which to base projections and evaluate performances.  It’s not always luck though; sometimes, pitchers give up more hits because they are getting hit harder, but this is usually shown by their batted ball types.  To further investigate, let’s look at the following table, which shows the career BABIPs in different situations for Hamels.

Season All Situations Bases Empty Men On Base Runners in Scoring Position
2006 0.291 0.269 0.324 0.253
2007 0.282 0.289 0.271 0.236
2008 0.262 0.261 0.265 0.270
2009 0.321 0.323 0.319 0.333
Career 0.288 0.285 0.294 0.277

As I looked closer at these numbers, I saw a fairly disturbing trend:

Season Innings 1-2 Innings 3-4 Innings 5-6 Innings 7+
2008 0.249 0.297 0.228 0.261
2009 0.277 0.338 0.369 0.283
Career 0.270 0.309 0.288 0.279

Obviously paring down the sample size even further increases the randomness of the statistics.  That said, even if they provide no predictive value, they do tell us what happened.  I usually ignore what happens in the 7th inning and later, because one of two things is occurring (or both): either the pitcher is tiring (probably having thrown a lot of pitches), or he’s locked in (hence why he’s still in the game so late).  But seeing that in turn 2 and 3 through the  batting order Hamels was knocked around in 2009, I wondered why.  I took a look at what pitches he was throwing to see if maybe he was overusing a pitch or maybe if one of his pitches just sucked.  The data:

Season Fastballs Curveballs Changeups Unknown
2006 61.1% (91.3) 11.3% (76.2) 27.7% (81.6) 2.10%
2007 54.3% (90.4) 10.6% (76.5) 35.2% (81.4) 1.90%
2008 54.8% (90.4) 13.7% (75.6) 31.5% (79.9) 1.30%
2009 59.1% (90.2) 10.5% (75.6) 30.3% (80.3) 1.30%
Total 57.0% (90.5) 11.6% (75.8) 31.4% (80.6) 1.60%

The first number in each column is the proportion of total pitches thrown with that pitch, and the second is the speed.  It seems as though nothing has changed with regard to his “stuff,” but he throws two pitches almost 90 percent of the time.  I wonder if after seeing his repertoire for four years now, National League hitters are just not as surprised as they used to be.  His changeup is his best pitch, working off his fastball, which is a full 10 miles per hour faster.  His curve has always been more of a token curve, shown for the sake of having a third pitch.

I think Cole Hamels will be just fine in 2010, mostly due to the BABIPs.  His ERA will come down, his secondary statistics will remain the same, and he’ll still be pitching for a good team.  But, there’s a chance that what happened in 2009 wasn’t an accident.  If he could figure out how to throw that curveball though to get hitters out, instead of just to show it, I’d feel a lot better about the first sentence in this paragraph.  He could become even better than he has already been in his short-spanning but excellent career.  Pretty impressive stuff for a guy entering his age-26 season.

Fastballs Curveballs Changeup Unknown
2006 61.1% (91.3) 11.3% (76.2) 27.7% (81.6) 2.10%
2007 54.3% (90.4) 10.6% (76.5) 35.2% (81.4) 1.90%
2008 54.8% (90.4) 13.7% (75.6) 31.5% (79.9) 1.30%
2009 59.1% (90.2) 10.5% (75.6) 30.3% (80.3) 1.30%
Total 57.0% (90.5) 11.6% (75.8) 31.4% (80.6) 1.60%
Advertisements

Written by Dan Hennessey

March 16, 2010 at 8:20 PM

Posted in Uncategorized

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: