Most Dominant Pitcher At His Peak?

Joe Posnanski is running a poll on his blog about the most dominant pitcher at his peak. The choices given are

  • R. Clemens
  • L. Grove
  • W. Johnson
  • S. Koufax
  • G. Maddux
  • P. Martinez
  • N. Ryan
  • None of the above

My kneejerk reaction would be it’s either Koufax or Pedro. Having looked over brWAR values recently, I know Koufax’s peak doesnt seem quite as high after adjusting for run environment and park effects. So it’s probably Pedro, but maybe W. Johnson? I would have thought Randy Johnson would be a good candidate also. Posnanski threw a curveball there by including Nolan Ryan; again, I learned recently that WAR doesnt show Ryan to have been as valuable as might be commonly assumed. He struck people out and had low batting averages against, but also walked a ton of people, which hurts his value. On the other hand, it’s an interesting point, maybe strikeouts are a necessary component of being “dominant”?

Let’s dig a little deeper.

First I’ll look at peak value, which I’m defining as mean WAA (wins above average) for a 4 year consecutive period, discarding the smallest value of the 4. I am ignoring all the high-innings guys in the 1800’s and early 1900’s; I’m not ready to declare Tommy Bond the most dominant pitcher in history.
By this measure, the candidates rank as,

  • W. Johnson: 9.9
  • P. Martinez: 8.2
  • R. Johnson: 7.9
  • L. Grove: 7.6
  • B. Gibson: 7.6
  • G. Maddux: 7.4
  • R. Clemens 6.9
  • S. Koufax: 6.7
  • N. Ryan: 3.8

So it’s Johnson pretty convincingly. Now, motivated by the inclusion of Ryan, I’ll take a different approach and consider strikeout dominance ahead of WAA. I will also consider opponent batting average and WHIP (walks+hits per inning pitched). To determine what constitutes dominance in these variables, I made a plot of opponent batting average vs WHIP, setting the lower limit on K/9 to (6.5, 8, 10.5). The idea is that Nolan Ryan was a dominant strikeout pitcher, and didnt give up a lot of hits, but he walked a lot of people so his WHIP isnt great; so maybe it’s ok to give up walks if you strike enough people out, and still be considered dominant. As an example, in the 1973 season when Ryan struck out 383 batters, his WHIP was 1.23, which ranks 8th in the league, i.e., pretty good but not historic (Tiant led the league at 1.09). Here is the plot.


It looks like there are plenty of pitchers who were as dominant of strikeout pitchers as Ryan, but had much better WHIPs; so you can’t totally excuse Ryan’s high walk rates. Based on this plot I tentatively define a dominant season to be WHIP<=1.2, opp BA <= 0.21 and K/9>=7.8. I only consider seasons from 1912 or later, and require 15 games started. This gives me 101 seasons, starting with Bob Feller and Hal Newhouser in 1946, up through Felix Hernandez and Ubaldo Jimenez in 2010 (I just realized the database Im using only goes through 2010). One of the fun things about doing this is finding players and seasons you hadn’t known about before. Looking at the players who had one “dominant” season, there are a handful of names I didn’t know, example,

  • Sonny Siebert, 1965! He struck out 9.1 batters per 9, led the league in K/BB and had a WHIP of 0.98.
  • Dave Boswell, 1966. Struck out 9.2 per 9.

Surprisingly, Feller (1946), Seaver (1971), and Carlton (1972) appear only once. There are a lot of other well-known seasons, Tanana 1976, Guidry 1978, Valenzuela 1981, Candelaria 1986, Cone 1988, Rijo 1988, Nomo 1995… Ramon Martinez is on the list for 1998. First-ballot hall-of-famer Dennis Eckersley is on the list for 1976!

The list of players with 2 dominant seasons (only through 2010) generally has more familiar names. Bob Gibson, Luis Tiant, JR Richards, Dwight Gooden, Mike Scott, Jason Schmidt, Johan Santana, Chris Young, and… Juan Guzman (1991 & 1992)! Also Erik Bedard!

The list of players with more than 2 dominant seasons are mostly the players that one would expect; Koufax (5 seasons), Sam McDowell (4), Nolan Ryan (9), Roger Clemens (4), Randy Johnson (7), Pedro (6). The surprising names here are Sid Fernandez with 5 seasons and Jim Maloney with 3! Maloney was especially good in 1963.

I can define dominance a little more strictly, say opp BA<=0.205, K/9>=8.5 and WHIP<=1. This leaves 21 seasons, and the main takeaway is that only Koufax and Pedro have more than one of them, Koufax with 4 and Pedro with 6. Without actually attempting to normalize the different offensive environments for Koufax and Pedro, it seems clear that, by this method, Pedro at his peak was the most dominant pitcher ever.

It is curious that Walter Johnson is the most valuable at his peak according to WAA, but doesnt appear here as having any dominant seasons. Clearly the fact that the strikeout totals were so much lower back then has a lot to do with this. Therefore I will look at these 3 metrics normalized to the league averages. One thing that jumps out immediately is that Dazzy Vance had incredible strikeout numbers for the time, striking out >2.5 times the league average per 9 in 1924, 1925, and 1926. The only other player to do that even once is Rube Waddell in 1902! Surprisingly, Walter Johnson’s strikeout numbers dont stand out that much, he is right around 1.5 to 1.7 times the league average for most of his career. Since his 1912 and 1913 seasons are the motivation for normalizing these numbers to the league, I will use those as a template for defining dominant. This implies I want K/9 about 1.5 times league average and WHIP about 0.75 times league average . This gives me 27 seasons of dominance;

  • P. (G.C.) Alexander, 1915
  • R. Clemens, 1997
  • R. Guidry, 1978
  • R. Johnson, 1995, 1999, 2004
  • W. Johnson, 1912, 1913
  • S. Koufax, 1963, 1965
  • Dutch Leonard, 1914 (era of 0.96!)
  • F. Liriano, 2006
  • Pedro, 1997, 1999, 2000, 2001, 2002
  • C. Mathewson, 1908
  • JR Richard, 1980
  • J. Santana, 2004, 2005, 2006
  • C. Schilling, 2002
  • M. Scott, 1986
  • T. Seaver, 1971
  • B. Sheets, 2004
  • J. Smoltz, 1996

If I relax the WHIP requirement a little (<0.85 of league average) and require real dominance in K/9 (>1.9 compared to league average), I get a slightly different list of seasons,

  • D. Gooden, 1984
  • L. Grove, 1928
  • R. Johnson, 1993, 1995, 1997, 1999, 2000, 2001
  • W. Johnson, 1924
  • S. Koufax, 1962
  • Pedro, 1999, 2000, 2001
  • JR Richard, 1979, 1980
  • N. Ryan, 1987, 1989, 1991
  • F. Tanana, 1975
  • D. Vance, 1924, 1925, 1928
  • R. Waddell, 1902

Tying it all together, I think there is really no rational argument that can be made that Pedro was not the most dominant pitcher in his peak. Behind him, it’s debatable, but R. Johnson, Koufax, W. Johnson all appear to have good cases. In terms of strikeouts, there simply is no one in history who was as dominant as Dazzy Vance; his 1924 might be the 3rd most dominant season ever behind Pedro’s 1999 & 2000. JR Richard and Rube Waddell had some really high, although really short, peaks. Johan Santana’s 2004-2006 ranks really well by historical standards. Sid Fernandez had an amazing peak, which somehow didn’t translate into wins and fame.

So my vote goes to Pedro, but don’t sleep on Dazzy Vance!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your 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