Los Angeles Dodgers: Comparing Sandy Koufax and Clayton Kershaw

MINNEAPOLIS, MN - OCTOBER 14: Picher Sandy Koufax of the Los Angeles Dodgers winds up to throw a pitch against the Minnesota Twin in game 7 of the 1965 World Series, October 14, 1965 at Metropolitan Stadium in Minneapolis, Minnesota. The Dodgers won the series 4 games to 3. Koufax was the series MVP and played for the Dodgers from 1955-66. (Photo by Focus on Sport/Getty Images)
MINNEAPOLIS, MN - OCTOBER 14: Picher Sandy Koufax of the Los Angeles Dodgers winds up to throw a pitch against the Minnesota Twin in game 7 of the 1965 World Series, October 14, 1965 at Metropolitan Stadium in Minneapolis, Minnesota. The Dodgers won the series 4 games to 3. Koufax was the series MVP and played for the Dodgers from 1955-66. (Photo by Focus on Sport/Getty Images) /
facebooktwitterreddit

Who was better in their prime for the Los Angeles Dodgers – Clayton Kershaw or Sandy Koufax?

Clayton Kershaw made his 314th career start the other day against the Rockies, an occurrence that created more than a little chatter around the Los Angeles Dodgers because it equalled the number of career starts made by another Dodger left-hander more than a half century ago.

MLB.com went so far as to make a Koufax-Kershaw career comparison its stat of the day, although the comparison was flawed both by its failure to contextualize the data and by its failure to recognize the potential for mischief in any career-based comparison involving Koufax.

To statisticians, Sandy Koufax is the poster boy for the concept of a peak value — as opposed to career value — comparison. Across the first six seasons of his 12-year career, he was a pitching wastrel, somebody the Dodgers always held out hope for, but who never delivered. Coming to the majors in June of 1955, he had a 36-40 record and ERA above 4.10 at the close of 1960, In an era when the average National League pitcher’s ERA was around 4.00, his stats made Koufax a slightly below average arm for the first half of his career.

More from Call to the Pen

He improved to 18-13 in 1961, then in 1962 burst forth as the Koufax we think of today. Winning the ERA title annually from 1962 through 1966, he topped 25 victories three times, won three Cy Young Awards and an MVP, and emerged with a 1.95 ERA for the period. At a time when the average league ERA hovered around 3.50, he was amazing.

The point isn’t to denigrate comparisons of Kershaw to Koufax, but to contextualize them. To do so, you have to look at Kershaw’s best five seasons relative to Koufax’s best five seasons, and consider both sets of numbers against the eras in which they were achieved.

To even do that, however, you have to first decide which set of five seasons actually constitute Kershaw’s peak. Two blocks of seasons contend for the honor: 2011 through 2015 or 2013 through 2017. The younger Kershaw turns out to have been better in some respects, the more veteran Kershaw in others.

In response to the question, “who was better at his peak, Koufax or Kershaw?” three measurements can appropriately be brought into play. One, ERA+, self-contextualizes. The other two, innings pitched and strikeouts, require that adjustments be considered, and perhaps implemented.

ERA+ is a relatively new measurement that adjusts each pitcher’s earned run average relative both to his league and to ballparks, an average performance always equating to 100. It didn’t exist when Koufax pitched, but has been calculated retroactively for him and those of his era. For the five-season period of Sandy’s peak – 1962 to 1966 – his average ERA+ was 168. That’s way out on the good side of his era’s bell curve. But no matter which seasons one chooses as Kershaw’s best, Clayton was better. From 2011 through 2015, Kershaw’s ERA+ averaged 175; from 2013 to 2017 it averaged 196. The latter figure argues that Kershaw was basically twice as good as the average pitcher of his time.

Theoretically, at least, Kershaw might benefit in ERA+ by being very much the best of a bad lot of pitchers, while Koufax’s 168 rating could be argued to be constrained by the presence of Gibson, Marichal, Drysdale, et al. That, of course, is an argument, not a fact; the only facts are the numbers themselves.

A second appropriate data set involves innings of work. Without delving too deeply into the complex subject of leverage, let’s agree that as a general proposition a pitcher working more innings has a greater value to his team than a pitcher working fewer innings. Between 1962 and 1966, Koufax averaged 275 innings per season. Obviously given the construct of the modern game, Kershaw never hit 275 innings even once, much less as an average. His peak, in 2013, was 236 innings. His 2011-15 average was 226 innings; from 2013-17 his average was 199.

Los Angeles Dodgers
Los Angeles Dodgers /

Los Angeles Dodgers

Because a pitcher’s value is directly related to workload, contextualization is less important with respect to innings pitched than it is with respect to either ERA or strikeouts. So it would be entirely appropriate to stop right there and declare Koufax the category winner. But let’s do the math, anyway. Between 1962 and 1965, National League pitchers who worked enough innings to qualify for the ERA title – generally 162 – had an average workload of right around 234 innings. Were we to create a stat called IP+ (for relative innings pitched), Koufax’s 275 average would translate to 117.5 against that 234 average.

Between 2011 and 2015, National League pitchers qualifying for the ERA title averaged 193 innings of work, against which Kershaw’s 122 IP+ is better than Koufax’s, although not by enough to ameliorate the impact of the 40-inning raw difference.

Between 2013 and 2017, National League ERA qualifiers averaged 190 innings, against which Kershaw’s 199 average translates to an IP+ of 104.7. Even setting aside the matter of the 76-inning difference in raw workload, that’s well below Koufax.

The final category is strikeouts, and here the contextual adjustment is critical. But let’s start with the raw numbers. Between 1962 and 1966, Koufax averaged 9.48 strikeouts for every 9 innings of work. Between 2011 and 2015, Kershaw averaged 9.9; between 2013 and 2017, he averaged 10.4.

Now for the context. Between 1962 and 1965, the more strikeout-averse National League hitters fanned an average of 5.78 times every 9 innings. That puts Koufax’s K+ — his relative strikeout rate – at 164.01. Between 2011 and 2015, National League hitters fanned an average of 7.64 times in every 9 innings of play.  Against that norm, Kershaw’s 9.9 average translates to a K+ of 129.58, significantly lower than Koufax’s score. If we use Kershaw’s 2013-17 block of seasons, his K+ does climb, but only to 130.65.

dark. Next. Time to talk about Kershaw

Contextually, then, Kershaw had the superior peak ERA. But Koufax was the superior strikeout ace. And although the two Los Angeles Dodgers aces’ era-contextualized workloads  are close enough for their relative values to be debated, Koufax’s superiority in raw workload is clear and meaningful.