August 20, 2002

Greatest Living Pitcher

Running the Numbers to Settle the Argument

by Keith Woolner and Jonah Keri

Few baseball arguments elicit more debates than comparing players across eras.

No matter how much the numbers may suggest otherwise, a fan growing up in the 60s may never be persuaded that Barry Bonds deserves to be ranked alongside Hank Aaron. Likewise, a fan growing up in the 90s may never fully grasp the greatness of Stan Musial.

But bringing Aaron and Bonds fans together is child's play compared to settling pitcher debates.

Deciding on the Greatest Living Pitcher, for one, requires a wide range of metrics and analysis.

While changes in ballparks, equipment and training methods have made hitters' arguments more heated, hitter usage has remained effectively constant for nearly 100 years. Sure, batters now get 162 games instead of 154, but an everyday player is still an everyday player, blessed with four or five at-bats a game to make his case. Metrics like VORP and EqA help adjust for most other variables.

Not so easy with pitchers. Changing usage patterns have yielded an increased specialization of pitchers' roles over the last 30 years. Gone are four-man workhorse rotations, with starters pushed to complete games nearly every other start. Now, every team carries a five-man starting staff, with six, sometimes seven relievers behind it.

Pulling two great pitchers from different eras out of a hat, how then do we compare Bob Gibson and Roger Clemens? Sure, we can adjust for league averages to compare ERAs across eras. But a 2.50 ERA carries a lot more value when spread over 300 innings instead of 200. So we'll also have to adjust for innings pitched by era to find our man.

Let's start with career numbers. Here's a chart to get us going. (See key following chart for details)

Career (150+ IP seasons only)

NAME                    IP   NRA RAPLUS   RAA   RAR    RANK   VALUE YEARS
------------------ ------- ----- ------ ----- ----- ------- ------- ---------
YOUNG,CY            7082.1  3.14  1.432   706  1879   0.849  25.181 1891-1910
JOHNSON,WALTER      5553.0  2.98  1.511   650  1560   0.867  24.960 1908-1926
SPAHN,WARREN        5102.5  3.31  1.361   394  1208   0.866  23.492 1947-1965
MADDUX,GREG         3520.0  2.82  1.596   447   983   0.821  21.656 1987-2001
CLEMENS,ROGER       3515.4  2.64  1.704   513  1052   0.790  21.382 1986-2001
SEAVER,TOM          4671.3  3.03  1.487   496  1225   0.712  20.998 1967-1986
PERRY,GAYLORD       5231.5  3.48  1.291   300  1119   0.739  19.963 1964-1983
CARLTON,STEVE       5038.6  3.50  1.284   274  1061   0.724  19.875 1967-1987
ALEXANDER,PETE      5010.2  3.19  1.412   462  1277   0.788  19.475 1911-1928
NIEKRO,PHIL         4986.3  3.51  1.281   268  1049   0.779  19.370 1967-1986
BLYLEVEN,BERT       4682.6  3.29  1.370   368  1096   0.715  19.087 1970-1989
GROVE,LEFTY         3697.3  2.74  1.642   529  1127   0.724  18.641 1925-1940
SUTTON,DON          5195.0  3.62  1.243   215  1018   0.662  18.521 1966-1987
MATHEWSON,CHRISTY   4672.4  3.30  1.364   379  1148   0.818  17.758 1901-1915
ROBERTS,ROBIN       4313.0  3.38  1.333   301   983   0.813  17.648 1949-1965
NICHOLS,KID         5045.3  3.14  1.432   505  1343   0.801  16.595 1890-1905
JENKINS,FERGIE      4382.1  3.42  1.316   282   973   0.722  16.507 1966-1983
GIBSON,BOB          3613.0  2.99  1.505   402   974   0.735  16.384 1961-1974
RYAN,NOLAN          4812.7  3.67  1.226   176   920   0.607  16.049 1971-1992
WYNN,EARLY          4231.3  3.91  1.152    47   717   0.703  15.581 1942-1962
JOHNSON,RANDY       2661.0  2.88  1.561   321   726   0.718  14.971 1989-2001
BUNNING,JIM         3545.9  3.39  1.329   236   793   0.758  14.902 1957-1970
PALMER,JIM          3585.3  3.07  1.465   368   929   0.701  14.895 1966-1982
GLAVINE,TOM         3069.5  3.28  1.374   237   703   0.722  14.728 1988-2001
RUFFING,RED         4119.5  3.71  1.212   132   799   0.637  14.719 1925-1942
HUBBELL,CARL        3400.4  3.05  1.477   367   913   0.661  14.679 1929-1942
FELLER,BOB          3263.4  3.26  1.382   273   797   0.810  14.270 1938-1953
MORRIS,JACK         3531.1  3.69  1.220   121   668   0.738  14.154 1979-1993
NEWSOM,BOBO         3621.5  3.85  1.168    63   640   0.835  14.111 1934-1947
PLANK,EDDIE         4364.8  3.38  1.330   309  1028   0.652  14.068 1901-1916
DRYSDALE,DON        3270.3  3.34  1.345   237   749   0.853  14.012 1957-1968
MARICHAL,JUAN       3362.6  3.36  1.341   241   771   0.716  13.410 1961-1973
TANANA,FRANK        3930.2  3.78  1.191   100   703   0.585  13.176 1974-1993
DERRINGER,PAUL      3645.1  3.82  1.179    75   662   0.706  12.941 1931-1945
BROWN,KEVIN         2632.3  3.14  1.433   243   645   0.690  12.894 1989-2000
KAAT,JIM            3857.6  3.83  1.174    74   670   0.655  12.777 1961-1977
LYONS,TED           3994.7  3.66  1.230   152   799   0.559  12.727 1924-1942
WILLIS,VIC          3996.1  3.67  1.226   152   814   0.751  12.696 1898-1910
FORD,WHITEY         2812.1  3.08  1.459   285   725   0.701  12.532 1953-1965
VIOLA,FRANK         2634.8  3.43  1.310   160   563   0.818  12.484 1983-1993
FRIEND,BOB          3330.7  3.68  1.222   116   638   0.692  12.286 1952-1965
PIERCE,BILLY        3093.2  3.27  1.375   249   741   0.604  12.247 1949-1962
KOOSMAN,JERRY       3596.4  3.63  1.241   148   706   0.581  12.141 1968-1984
JACKSON,LARRY       3177.3  3.56  1.266   156   656   0.726  12.132 1955-1968
COVELESKI,STAN      2988.6  3.33  1.353   230   711   0.781  12.044 1916-1926
REUSCHEL,RICK       3208.8  3.48  1.292   179   672   0.642  11.884 1973-1989
RIXEY,EPPA          4058.9  3.86  1.166    63   718   0.580  11.841 1912-1930

NRA: RA (Runs Allowed) normalized for league and park. Measured on a scale like ERA, with 4.50 being average in any given year.

RAPLUS: RA+, which is to NRA as PRO+ is to OPS. 100 is average, any score above 100 is better.

RAA: Runs Above Average, as in the number of runs a pitcher saves his team in his career above a league-average counterpart.

RAR: Runs Above Replacement, as in the number of runs a pitcher saves a team in his career above a replacement-level counterpart.

RANK: Average percentile rank in innings pitched of the pitchers' five peak seasons. Suppose that 100 pitchers put up seasons of 150 or more innings pitched every year in a given five-year period. The pitcher with the most IP gets a RANK of 1.00, the second-highest gets a RANK of .99, third-highest gets a RANK of .98, and so on. The lowest IP total of the 100 would get a zero.

VALUE: The career value represented by the number of seasons leading the league in IP with a league average RA+. Assigns a pitcher 1.00 for leading the league in IP with a league average RA+ in one season (RANK = 1.00 * RA+ = 1.00 = 1.00 VALUE). For instance, leading the league in IP with an RA half of league average is worth 2.00 VALUE. Young's career is like 25 heavy-workload seasons of league-average performance, or 12.5 heavy-workload seasons at half the league RA.

OK, we've got some perspective now. All-time greats like Cy Young and Walter Johnson dominate this list, with future Hall of Famers like Clemens and Greg Maddux hovering near the top.

Let's examine what these numbers mean.

Our first chart ranks every pitcher, living or deceased, by VALUE, derived from RANK. We're trying to find the Greatest Living Pitcher, but this gives you a feel for where pitchers like Maddux and Clemens rank compared to their pitching ancestors.

After the late Young and Johnson, Warren Spahn leads all living pitchers with a VALUE of 23.492. By adjusting his "Spahn and Sain and pray for rain" burden against the average workload of his era, we can more easily compare Spahn's performance to modern aces like Maddux and Clemens. Maddux barely edges Clemens for the #2 spot on the active pitchers' list, a margin slightly widened this year.

Of course, there's a problem here. Spahn's career lasted 19 full seasons. It might have lasted longer had World War II not truncated the start of his career. Maddux, meanwhile, had pitched only 15 full seasons heading into this year.

Here we must consider what makes a pitcher great. Should longevity be a factor? Absolutely.

But peak value should also play a starring role. If not for his six-year stretch of dominance, Sandy Koufax would have as much claim to a Hall of Fame plaque as Oil Can Boyd. Maybe less.

Let's run the peak numbers and see what we get.

Top 5 years (not necessarily consecutive)

NAME                    IP   NRA RAPLUS   RAA   RAR    RANK   VALUE PEAK5
------------------ ------- ----- ------ ----- ----- ------- ------- ---------
JOHNSON,WALTER      1749.4  2.24  2.006   354   641   0.976  10.056 1912-1918
CLEMENS,ROGER       1262.7  2.13  2.110   253   447   0.924   9.933 1986-1998
MADDUX,GREG         1228.4  2.14  2.100   245   433   0.909   9.794 1992-1998
MARTINEZ,PEDRO      1122.0  2.21  2.039   216   386   0.832   9.365 1996-2000
JOHNSON,RANDY       1197.4  2.22  2.027   229   411   0.891   9.070 1995-2001
GROVE,LEFTY         1386.7  2.32  1.937   265   490   0.914   8.893 1930-1937
HUBBELL,CARL        1512.4  2.55  1.764   248   492   0.971   8.802 1932-1936
GIBSON,BOB          1480.3  2.59  1.740   233   467   0.945   8.536 1964-1970
BROWN,KEVIN         1209.6  2.56  1.760   187   371   0.941   8.440 1996-2000
SEAVER,TOM          1391.2  2.41  1.868   242   462   0.894   8.409 1969-1977
FELLER,BOB          1630.3  2.69  1.671   240   503   0.986   8.302 1939-1947
NEWHOUSER,HAL       1495.3  2.70  1.670   221   463   0.976   8.300 1944-1949
YOUNG,CY            1958.7  2.59  1.740   322   646   0.948   8.274 1892-1904
KOUFAX,SANDY        1448.4  2.51  1.790   238   468   0.889   8.139 1961-1966
ALEXANDER,PETE      1883.6  2.78  1.621   262   570   0.991   8.104 1911-1920
MARICHAL,JUAN       1492.6  2.70  1.664   216   451   0.950   7.973 1963-1969
PALMER,JIM          1539.3  2.73  1.646   215   457   0.962   7.964 1970-1978
SPAHN,WARREN        1437.4  2.78  1.619   196   427   0.960   7.935 1947-1958
CARLTON,STEVE       1512.7  2.85  1.577   190   426   0.980   7.902 1972-1983
ROBERTS,ROBIN       1632.7  2.83  1.588   213   475   0.988   7.856 1950-1954
WALSH,ED            2017.7  2.91  1.548   255   588   0.992   7.711 1907-1912
NICHOLS,KID         1984.7  2.76  1.631   288   617   0.922   7.602 1890-1898
PERRY,GAYLORD       1612.0  2.93  1.534   190   447   0.971   7.590 1967-1974
GLAVINE,TOM         1192.3  2.79  1.613   154   334   0.931   7.551 1991-2000
MATHEWSON,CHRISTY   1724.0  2.84  1.583   230   513   0.944   7.543 1903-1912
MUSSINA,MIKE        1153.8  2.69  1.674   161   337   0.889   7.523 1992-2001
BUNNING,JIM         1458.9  2.85  1.578   182   413   0.949   7.510 1957-1967
VIOLA,FRANK         1252.4  2.77  1.626   165   357   0.912   7.472 1984-1992
STIEB,DAVE          1341.0  2.89  1.554   161   370   0.942   7.442 1980-1985
VANCE,DAZZY         1385.9  2.79  1.611   188   414   0.902   7.388 1924-1930
WALTERS,BUCKY       1464.7  2.95  1.525   173   409   0.950   7.344 1939-1944
DEAN,DIZZY          1531.0  3.01  1.494   174   423   0.974   7.316 1932-1936
SCHILLING,CURT      1216.4  2.83  1.591   154   341   0.893   7.134 1992-2001
BLYLEVEN,BERT       1402.3  2.92  1.542   167   386   0.918   7.117 1973-1989
JENKINS,FERGIE      1585.6  3.07  1.467   164   415   0.968   7.116 1968-1974
WARNEKE,LON         1363.3  2.90  1.551   169   389   0.895   7.026 1932-1941
COVELESKI,STAN      1525.3  2.97  1.517   180   428   0.918   7.011 1917-1921
DRYSDALE,DON        1483.6  3.13  1.439   144   377   0.961   7.001 1959-1965

Here, RANK is expressed as an average score for each pitcher's top five seasons. VALUE is the value of each pitcher's top five seasons (RANK*RAPLUS). PEAK5 is the range of years that contain each pitcher's top five seasons.

You can argue the Big Train took advantage of baseball's lily-white landscape in imposing his will on the league - the above metrics don't account for pre- and post-integration performance. You can also make a great case for Walter Johnson as the greatest pitcher of all time.

Johnson, however, is ineligible to participate in a discussion of living pitchers, so the question remains: Who is the king of all living pitchers? Let's whittle out the pretenders first.

It's not Randy Johnson or Pedro Martinez. Though each can claim one of the top peaks in history, neither has shown enough longevity to rival the very best. Check back in 2005 or so and we'll see how far they've progressed.

It's not Steve Carlton or Tom Seaver. Both fare well in both career and peak rankings, just not well enough.

Guys like Phil Niekro and Don Sutton lag behind the leaders by a fair margin.

Nolan Ryan apologists have little to stand on. Ryan's noted longevity amounted to only the 19th all-time rank in career VALUE. You won't find him anywhere in the top peaks list.

We're down to Spahn, Maddux and Clemens.

We'll let Spahn duel with Carlton for Best Living Lefty status. Spahn's 18th rank in peak VALUE negates his top spot in career VALUE.

So, Maddux or Clemens?

Clemens' peak VALUE wins by 0.139. Maddux's career VALUE wins by 0.274. Too close to call.

However, Clemens edges Maddux in career RA+, RAA and RAR, falling short only in innings totals, which netted him the lower career VALUE total.

So by the slimmest of margins, we'll take Clemens.

But with Maddux three and a half years Clemens' junior and the Rocket's durability finally taking a hit this year, Maddux's future ranking looks a little brighter. He could pass Clemens as early as next season, or toward the end of the duo's careers.

Stay tuned.

Keith Woolner and Jonah Keri are authors of Baseball Prospectus. Contact Keith by clicking here. Contact Jonah by clicking here.

Keith Woolner is an author of Baseball Prospectus. 
Click here to see Keith's other articles. You can contact Keith by clicking here
Jonah Keri is an author of Baseball Prospectus. 
Click here to see Jonah's other articles. You can contact Jonah by clicking here