The Cole Hamels Rebound

Cole Hamels turned in his best start of the season on Friday against the San Diego Padres, tossing eight strong innings before deferring to Ryan Madson to close out the game at 3-1. Hamels allowed one run on three hits, striking out ten (tying a season-high) and walking only one. It was a stark contrast to his previous start against the New York Mets on Saturday, when he allowed seven runs and failed to get through the fifth inning.

Although his results were stellar, Hamels was not himself leading up to his start on Friday. In his prior seven starts, he posted a 2.96 ERA but struck out only 33 in 45 and two-thirds innings, a per-nine rate of 6.5. In the 13 starts prior to that selection, he struck out 91 in 90 and two-thirds innings, a per-nine rate of 9.0. Hamels wasn’t being hit hard recently nor did his batted ball profiles change in any significant way; he simply wasn’t missing bats.

What was the problem? Hamels’ velocity was never alarming. Via Joe Lefkowitz’s site (click to enlarge):

FF: Four-seam fastball; FC: Cut fastball; CH: Change-up; CU: Curve

His pitch selection didn’t differ, as the following chart illustrates:

Pitch selection by handedness plus or minus one standard deviation:

RHB

  • Change-up
    • April 5 to June 8: 17.8 +/- 5.3
    • June 14 to July 16: 17.9 +/- 6.4
  • Curve
    • April 5 to June 8: 6.8 +/- 4.3
    • June 14 to July 16: 8.1 +/- 4.3
  • Four-seam fastball
    • April 5 to June 8: 40.8 +/- 10.9
    • June 14 to July 16: 35.7 +/- 12.6
  • Cut fastball
    • April 5 to June 8: 10.4 +/- 3.2
    • June 14 to July 16: 9.6 +/- 6.1

LHB

  • Change-up
    • April 5 to June 8: 4.7 +/- 3.0
    • June 14 to July 16: 5.1 +/- 4.2
  • Curve
    • April 5 to June 8: 3.0 +/- 1.7
    • June 14 to July 16: 3.3 +/- 3.3
  • Four-seam fastball
    • April 5 to June 8: 16.2 +/- 8.1
    • June 14 to July 16: 15.4 +/- 10.0
  • Cut fastball
    • April 5 to June 8: 3.0 +/- 1.7
    • June 14 to July 16: 3.3 +/- 3.3

Finally, although swings and misses were down in the latter selection, there was no particular pitch making the difference.

Swing and miss averages plus or minus one standard deviation:

  • Change-up
  • April 5 to June 8: 6.5 +/- 2.9
  • June 14 to July 16: 5.7 +/- 4.0
  • Curve
    • April 5 to June 8: 1.4 +/- 1.1
    • June 14 to July 16: 1.0 +/- 1.4
  • Four-seam fastball
    • April 5 to June 8: 3.1 +/- 1.6
    • June 14 to July 16: 2.3 +/- 1.4
  • Cut fastball
    • April 5 to June 8: 1.5 +/- 1.7
    • June 14 to July 16: 1.1 +/- 0.7

    Is it possible that it was all random? A pitcher whose K/9 was 9.1 last year, sits at 8.4 this year, and is 8.5 for his career — could he simply have gone down to 6.5 for a span of seven starts just by sheer randomness? It appears that that is exactly the case. In fact, it is a great illustration of exactly why small sample sizes are met with such skepticism among the statistically-minded. When we notice a trend, we have been trained to think that there must be an underlying cause; we need a reason to fit our narrative, and randomness is such a boring plot line.

    The good news is that Hamels righted his ship in his last outing against the Padres, notching his 18th double-digit strikeout performance of his young career. He has matched last year’s Cy Young winner, Roy Halladay, pitch-for-pitch all season long. Not even an unfortunate bout with the baseball gods could sidetrack him from helping lead his team to the post-season.

    Leave a Reply

    *

    3 comments

    1. The Citizens Bankers

      July 25, 2011 07:16 AM

      That was some really, really tremendous work. Had to take a lot to go through all of those stats, probably looking for a reason why he wasn’t striking out as many only to arrive at a conclusion that contrasts with what most people wanted to hear. I loved the point about how “randomness” doesn’t fit our narrative. It’s the truth, but randomness is exactly what happened.

    2. COAL HAMLETS

      July 25, 2011 10:13 AM

      It kinda reminds me how Cliff Lee had a stretch where he was striking out a ridiculous amount of batters, then went through his crazy June where he struck out relatively few, before settling down somewhere in the middle since then.

      There didn’t seem to be much noticeable difference with what Cliff Lee was doing, but the results differed. Cole Hamels seemed to have had the same thing happen to him, although a little less extreme.

    3. sean

      July 25, 2011 10:22 AM

      it was a hot game to sit through, great outing by cole though

    Next ArticlePhillies Pitching Dominant, Including the Bullpen