Thursday, June 20, 2002
Johnson vs. Koufax
By David Schoenfield
One of the regulars in the ESPN Classic rotation is Game 7 of the 1965 World Series, perhaps the most famous start of Sandy Koufax's career. Dodgers manager Walter Alston skipped over Don Drysdale and pitched Koufax on two days' rest. He blanked the Twins 2-0, on three hits, for his second consecutive shutout (he had pitched a four-hitter in Game 5).
If you've ever seen the game, Koufax mixed in his big curveball with a blistering fastball and struck out 10. The Twins helped him out by swinging at several eye-level pitches (an indication, perhaps, not only of Koufax's explosiveness but also of the higher strike zone intact back then). It was a classic outing and cemented Koufax's legend as one of baseball's greatest pitchers ... and perhaps the most dominating left-hander of all time. In fact, when ESPN broadcast profiles of its 50 greatest athletes of the 20th century in 1999, Koufax was the only pitcher to rank in the top 50.
As great as Koufax was -- he went 26-8 with a 2.04 ERA in 1965 and 27-9 with a 1.73 ERA in 1966 before retiring due to arthritis -- Randy Johnson, another lefty with a devastating fastball and knee-buckling slider, may be ever better. Until last season, however, when he beat the Yankees in Game 6 and then won Game 7 in relief, Johnson's mediocre postseason performance had blemished his legacy. But now that we have an image of a heroic Johnson trudging in from the bullpen in the World Series, his place as one of baseball's greatest pitchers is secured.
Is Johnson better? After all Koufax, won 25, 26 and 27 games and led the National League five straight seasons in ERA, including three years below 2.00. Johnson has a career-high of 21 wins and has never had an ERA under 2.00. But ... Koufax pitched in one of the greatest pitching environments in history -- in the 1960s, and specifically in Dodger Stadium, a tremendous park for pitchers. For example, in 1965, Koufax had a 1.38 ERA at home, 2.72 on the road; in 1964, he was 0.85 at home, 2.93 on the road. Johnson, meanwhile, has pitched in a homer-happy era in which more runs are scored.
In order to compare the two, rather than just simply look at their ERAs and adjust for era and home ballpark, I thought it would be fun to transport Johnson back to the 1960s ... what would his numbers look like if he had pitched in Dodger Stadium in 1965?
Koufax had five dominating seasons before his early retirement; for Johnson, I'll use his five-year run from 1997 to 2001, which leaves out his 1995 Cy Young campaign with Seattle, but includes his three with Arizona. Basically, the premise works like this: compare each pitcher to the league average for each season, and then translate Johnson's numbers to Koufax's era. As an example, in 1965, Koufax fanned 10.24 hitters per nine innings; the league average (minus Koufax) was 5.83 K's per nine. This gives Koufax a ratio of 1.76 times the league average. In 2000, Johnson averaged 12.56 K's per nine; the league average was 6.68, giving Johnson a ratio of 1.88. In transporting Johnson back to 1965, we multiply his ratio of 1.88 times the 1965 average of 5.83 and determine Johnson would have struck out 10.96 batters per nine innings. Once we estimate his innings pitched, it's then easy to figure out his strikeout total.
In order to calculate innings pitched, a couple adjustments need to be made. First of all, since Koufax pitched in a four-man rotation, he started more games (as many as 41 in a season). For Johnson, that means a maximum of six additional starts per season. Pitcher usage patterns have also changed, as relievers are used more now than in the '60s, so another adjustment is needed for Johnson to give him more complete games (and thus more innings). In considering this, I asked: was Johnson as durable in his time as Koufax was in his? The answer is yes. Johnson's percentage of complete games compared to an average pitcher was actually higher than Koufax's (although, since that percentage is so low now -- less than five percent -- it's easier to exceed that total). Also, in their non-complete-game starts, Johnson averaged more innings per start (estimated) than Koufax in three of the five comparable seasons.
So, with that in mind, in order to arrive at an innings pitched total for Johnson, I assumed he would get as many complete games as Koufax did and, in his other starts, average as many innings per outing as he actually did.
In determinining, won-lost records, Koufax had a slightly higher percentage of decisions to starts (about one per year), so we adjust for that. Also, in transforming Johnson's won-lost records from his real totals to 1960s totals, I decided to use support-neutral W-L marks, as constructed by Michael Wolverton of Baseball Prospectus. For example, Johnson went 21-6 last year, but with neutral support would have gone 20-7. In 1999, he was 17-9, but should have gone 21-9, as he suffered from lack of run support that year.
Using the ERA+ number, which adjusts a pitcher's ERA for the league and home ballpark to a common scale, from baseball-reference.com, Johnson's ERAs can be seen in a 1960s window. For example, his ERA+ in 2001 was 184. The NL ERA in 1966 was 3.28. Divide 3.28 by 1.84 and you get 1.78 -- Johnson's 2001 ERA as if he had pitched in 1966 (Koufax was 1.73 that year).
OK, enough with the (boring) particulars, here are Koufax's actual stats compared to Johnson's "transported" stats:
Year Pitcher GS CG IP SO AVG W-L ERA
1962 Koufax 26 11 184 216 .197 14-7 2.54
Johnson 33 12 255 303 .187 19-5 1.83
1963 Koufax 40 20 311 306 .189 25-5 1.88
Johnson (AL) 27 12 202 259 .218 9-12 2.82
Johnson (NL) 13 8 107 127 .179 9-1 0.95
1964 Koufax 28 15 223 223 .191 19-5 1.74
Johnson 41 38 363 418 .197 26-11 1.83
1965 Koufax 41 27 335 382 .179 26-8 2.04
Johnson 41 25 330 402 .211 26-8 1.84
1966 Koufax 41 27 323 317 .205 27-9 1.73
Johnson 40 13 311 381 .200 25-8 1.78
By our calcuations, Johnson finishes with two 400-strikeout seasons, three 25-win seasons and beats Koufax in ERA two seasons and ranks nearly identical in two others. Johnson was more durable, as injuries limited Koufax in two of the five years, and was also a more dominant strikeout pitcher. And keep in mind that the W-L records are support-neutral; with a good team behind him -- like the Dodgers, who were actually an above-average offensive when considering Dodger Stadium -- Johnson would have approached 30 wins.
Of course, we don't really know if Johnson would have been able to pitch 300 innings in four straight seasons; maybe his shoulder would have blown out. On the other hand, if Johnson had pitched in Dodger Stadium in 1965 -- with its famous high mound -- he may have been even better; imagine that 6-foot-10 frame on a mound that was perhaps six inches higher than present rules allow.
Part of Koufax's legend is that he did retire young; a career unfinished, left to speculation. What if he had kept pitching? How many more 300-inning, 25-win seasons could he have pumped out? I'm not sure he could have kept pitching at that extreme; three or four 300-inning seasons appears to be a maximum amount. Among other pitchers from the '60s and '70s, Gaylord Perry had the most 300-inning years at six. Don Drysdale and Jim Palmer had four; Juan Marichal, three; Bob Gibson and Steve Carlton, two each; Tom Seaver never had a 300-inning season. And those were the guys who did stay healthy.
Randy Johnson, meanwhile, is still going strong at age 38. We should be thankful we get to see this amazing pitcher perform his wonders.