Remember me

The Bob Feller Era

February 19, 2019

The Bob Feller Era

We have been reviewing the best pitcher in each league in each season, and we had gotten through the American League in 1936.   I am comparing the best starting pitcher as seen by Baseball Reference WAR, and the best starting pitcher as seen by my own start-by-start examination of the data, which I call Deserved WAR or D-WAR.   If the two methods disagree, then I’ll take a minute to examine the arguments in favor of each pitcher in a little bit more detail.  This series resumes with the National League in 1937.



1936 NL—Carl Hubbell (both systems)

Hubbell was the National League MVP. 


1937 AL—Lefty Gomez (D-WAR) vs. Lefty Grove (R-WAR)

Gomez in 1937 was 21-11 with a 2.33 ERA in 278 innings.  He led the league in Wins, ERA, strikeouts (194) and ERA+ (193!!).   Grove was very good, as always, but he doesn’t match Gomez in any stat—wins, winning percentage, ERA, innings, strikeouts, ERA+. 

Of course, Gomez pitched for a much better team than Grove (the Yankees vs. the Red Sox), and in a much better park for a pitcher (Yankee Stadium vs. Fenway Park.)  I think they’re pretty much even. 


1937 NL—Jim Turner (D-WAR) (both systems)

Turner, a 33-year-old rookie, was 20-11 with a league-leading 2.38 ERA.


1938 AL—Red Ruffing (D-WAR) vs. Thornton Lee (R-WAR)

This one is truly weird.  Thornton Lee was 13-12 with a 3.49 ERA, and gave up 28 un-earned runs.   Ruffing was 21-7 with a 3.31 ERA, only 13 un-earned runs, almost the same innings pitched.  Ruffing’s strikeout/walk data was 131-68; Thornton Lee’s was 77-94. There is a small difference in park effects, but I don’t understand how R-WAR reaches the conclusion that Lee was the best pitcher in the league, and I don’t think it is a credible conclusion.   Frankly, his designation as the most valuable pitcher in the league just seems totally screwy. 

1938 NL—Big Bill Lee (both systems)

1939 AL—Bob Feller (both systems)

1939 NL—Bucky Walters  (both systems)

Bucky went 27-11, and also hit .325, driving in 16 runs with 10 extra-base hits. 

1940 AL—Bob Feller (both systems)

1940 NL—Bucky Walters (D-WAR) vs. Claude Passeau (R-WAR)

Passeau, like Hubbell and Grove in earlier seasons, made 31 starts but also pitched 15 times in relief.  As a reliever he was 4-1, had 5 saves, pitched 35 innings and had a 0.78 ERA (in relief).   My system only evaluates starts.  I accept that Passeau may well have been the best pitcher in the league if you include his relief work. 

1941 AL—Bob Feller (D-WAR) vs. Thornton Lee (R-WAR)

Thornton Lee, who seems like a dart-board selection as best pitcher in the league in 1938, is a more reasonable selection in 1941.  Feller was 25-13 with 260 strikeouts and a 3.15 ERA.  Lee was 22-11 with a league-leading 2.37 ERA.   Feller, working immensely hard, led the league in almost everything else—starts (40), innings pitched (343, but Thornton Lee also pitched 300), wins (25), strikeouts (260), walks (194), batters faced (1,466), shutouts (6).   Since baseball did not yet have relievers pitching 70 games a year, Feller even led the league in game appearances (44).  Within three or four years relievers would start pitching in 70 games a year. Lee did edge Feller in Complete Games, 30-28.

Both Feller and Lee had seasons that we would normally regard as being worthy of a Cy Young Award.   Our system is missing ten starts for Lee, but Feller is shown as so far ahead that it is unlikely that Lee could catch him even with complete data.   R-WAR shows Lee as only a small distance ahead of Feller (8.7 to 8.2, a difference of five runs.)  I really have no insight as to which was the more valuable pitcher for the season.  


1941 NL—Whitlow Wyatt (both systems)

1942 AL—Tiny Bonham (D-WAR) vs. Tex Hughson (R-WAR)

Bonham, making only 27 starts, was 21-5 with a 2.27 ERA.  Hughson pitched a lot more innings (281) with a similar record, 22-6 with a 2.59 ERA.  Hughson led the league in strikeouts with only 113, less than one-half of what Feller would have had had he not been in the service.   Hughson has some missing data and is close enough to Bonham that he would quite certainly have passed him in D-WAR with complete data, so there is no actual conflict between the systems here.


1942 NL—Mort Cooper  (both systems)

1943 AL—Spud Chandler  (both systems)

1943 NL—Mort Cooper  (both systems)

1944 AL—Dizzy Trout  (both systems)

Trout’s teammate, Hal Newhouser, won the Most Valuable Player Award.  Trout and Newhouser were more dominant in their time than Koufax and Drysdale.   Newhouser was 29-9 with a 2.22 ERA; Trout was 27-14 with a 2.12 ERA.   Newhouser won the MVP Award in both 1944 and 1945, but by only four points over Trout in 1944, and Trout actually had more first-place votes than Newhouser, 10 to 7.  Both systems under consideration here agree that Trout was actually more valuable. 

1944 NL—Bill Voiselle (D-WAR) vs. Rip Sewell (R-WAR)

I’m missing a lot of data for Sewell, so I can’t contribute to his overall evaluation.  Bucky Walters, 23-8 with a 2.40 ERA, and Mort Cooper, 22-7 with a 2.46 ERA, would appear on a superficial level to be the best pitchers in the league.    I can’t resolve the discrepancy because of the relatively high amount of missing data during the war years. 


1945 AL—Roger Wolff (D-WAR) vs. Hal Newhouser (R-WAR)

I have data for all 29 of Wolff’s starts but only 21 of 36 for Newhouser.  I am certain Newhouser would rank ahead of Wolff if I had full data. 

1945 NL—Preacher Roe (both systems)

On a superficial level the best pitcher in the National League in 1945 was Hank Wyse, 22-10 with a 2.68 ERA for the league champion Cubs.   If there had been a Cy Young Award at the time it is 100% certain that it would have gone to either Wyse or Red Barrett (23-12, 3.00 ERA, mostly for St. Louis, traded early in the season.)   Roe was just 14-13 with a 2.87 ERA despite decent offensive support.   But D-WAR and R-WAR agree that he was actually the best pitcher in the league.  


1946 AL—Hal Newhouser (D-WAR) vs. Bob Feller (R-WAR)

1946 NL—Howie Pollet (both systems)

1947 AL—Bob Feller (D-WAR) vs. Hal Newhouser (R-WAR)

Whether Bob Feller was better than Hal Newhouser in 1946 but Newhouser better than Feller in 1947 or the other way around does not seem like a question that we should put a lot of time into.

1947 NL—Ewell Blackwell (D-WAR) vs. Warren Spahn (R-WAR, sort of)

Spahn was 21-10 with a league-leading 2.33 ERA; Blackwell was 22-8 with a 2.47 ERA and a league-leading 193 strikeouts.  They pitched in the two best pitchers’ parks in the league; in fact, Boston (Spahn) and Cincinnati (Blackwell) had park effects so low that all six of the other parks in the league had park factors over 100.

Blackwell whipped the ball sidearm, very hard, and made a tremendous impression on the reporters of the era although he had only a couple of big seasons due to injuries.   Apparently he was very frightening to a right-handed hitter.  Sportswriters in the early 1960s, when I became a baseball fan, still talked about him, and I have no doubt that he would have won the 1947 Cy Young Award had there been such an award.  But I think the matchup, by an analytical evaluation, is too close to call.  See note on Spahn in 1951.   Spahn didn’t actually lead the league in R-WAR for pitchers; I’ll explain later.   


1948 AL—Bob Lemon (D-WAR) vs. Hal Newhouser (R-WAR)

The difference is just missing starts for Newhouser.  I am missing 11 starts for Newhouser, none for Lemon.  I accept that Newhouser might be better if I had full data.

1948 NL—Johnny Sain (both systems)

1949 AL—Mel Parnell (both systems)



Kinder Against Hutchinson

Mel Parnell, the best pitcher in the American League in 1949, went 25-7 with a 2.77 ERA for the Boston Red Sox. Parnell was supported by 48 double play balls that year.   I don’t know what the record is, but that’s a big number, 48 DPs.   Tommy John’s career high was 45, and he was over 40 only twice.   I think 48 is the biggest number that I have seen. 

The Red Sox #2 pitcher was Ellis Kinder, who was 23-6 with a 3.36 ERA.   On May 3, 1949, Kinder started for the Red Sox against Fred Hutchinson of Detroit, in Tiger Stadium.  The Red Sox scored three in the first on a single by Johnny Pesky, a double by Ted Williams, and a home run by Vern Stephens.  They manufactured a run in the second inning to make it 4-0, but the Tigers answered with a solo homer of their own, making it 4-1. 

In the third, however, Detroit got three walks, five singles and a home run, adding them up to 9 runs.   Detroit had a 10-4 lead, and Kinder, of course, was out of the game.   Ted Williams answered in the 4th with a three-run homer, making it 10-7.  In the fifth inning the Tigers got three singles to stretch their lead to 11-7. 

The Red Sox in the sixth got five hits for four runs, tying the score at 11-11, and then got three more runs on three hits in the 7th, making it 14-11.   But in the 9th inning Pat Mullin, pinch hitting for the Tigers with the bases loaded, hit a triple, tying the score at 14-14.

No one scored in the 10th, the 11th or the 12th.   With two out in the bottom of the 13th George Kell grounded to short, the third out potentially, but reached on an error.  The pitcher threw a wild pitch, then issued a walk, then another walk.   The bases were loaded with two out. 

Bob Swift flied to center. 

It was too dark to continue the game.  The game ended in a 14-14 tie.

There are 324 tie games in my data, 648 starts by pitchers which ended in a tie—but that’s the only 14-14 tie.  Three games ended 13-13, thirty-seven games ended 5-5, and seventeen games ended nothing to nothing.  But only one game ended 14 to 14. 


1949 NL—Don Newcombe (D-WAR) vs. Howie Pollet (R-WAR)

Newcombe was 17-8, 3.17 ERA; we have data for all 31 of his starts.   Pollett was 20-9, 2.77 ERA.   Missing data for three starts for Pollett, plus he pitched 11 times in relief and was effective as a reliever.  I accept that Pollett should probably rank ahead in a full analysis. 

1950 AL—Early Wynn (D-WAR) vs. Ned Garver (R-WAR)

Missing a lot of starts for Garver in both 1950 and 1951.   Garver in 1950 was just 13-18 with a 3.39 ERA, so it would be really interesting if he was in fact the best pitcher in the league.  He was pitching with a 96-loss team in a hitter’s park, and there was no pitcher in the league having an obviously tremendous season. 

1950 NL—Ewell Blackwell (D-WAR) vs. Robin Roberts (R-WAR)

Blackwell was 17-15 with a 2.97 ERA in 261 innings; Roberts was 20-11 with a 3.02 ERA in significantly more innings (304).  Both systems have them separated by inches; D-WAR has Blackwell ahead 8.8 to 8.7, while Roberts leads in R-WAR 7.4 to 7.2.   It’s just a difference of one or two runs either way.   The systems are not accurate enough that anyone should treat such a difference as meaningful.


1951 AL—Early Wynn (D-WAR) vs. Ned Garver (R-WAR)

Garver in 1951 won 20 games with a 102-loss team. 


1951 NL—Warren Spahn (D-WAR) vs. Robin Roberts (R-WAR)

Spahn was 22-14 with a 2.98 ERA in 311 innings; Roberts was 21-15 with a 3.03 ERA in 315 innings.  There were other pitchers in the league who had seasons superficially as good or better.   Sal Maglie was 23-6 with a 2.93 ERA, Larry Jansen was 23-11 with a 3.04 ERA, Don Newcombe was 20-9 with a 3.28 ERA, Preacher Roe was 22-3 with a 3.04 ERA.  Murry Dickson won 20 games for a team which won only 64.  There are a long list of viable Cy Young candidates. 

R-WAR likes Roberts, in 1951 and in many other seasons, in part because he pitched enormous numbers of innings, leading the league in innings in 1951, 1952, 1953, 1954 and 1955, and sometimes leading by wide margins.  In 1952, 1953 and 1954 both systems agree that he was the most valuable pitcher in the league. 

But back to Spahn; Spahn, of course, had as many good and great seasons as any pitcher in history; his career won-lost record is about the same as Koufax and Drysdale added together.  He led the league in wins 8 times, in ERA 3 times, and pitched in 17 All-Star games. 

Given that great of a career—17 All Star games?—Spahn does surprisingly poorly in WAR comparisons.   In Baseball Reference WAR (R-WAR) he led the league only once, 1947, and actually has no season ranking #2, only seven seasons even finishing in the top 10.  In 1947 Spahn and Blackwell had 9.5 R-WAR each, nearly tying for the league lead; apparently Blackwell was .0-something ahead, since Blackwell is recognized by the site as the league leader. 

Spahn, however, had a better year with the bat than Blackwell did.  Both Spahn and Blackwell take deductions from their WAR based on their hitting, but after the deductions, Spahn not only passes Blackwell but becomes the #1 player in the league—the Most Valuable Player. 


Wins Above Replacement—all (1947)
































So actually, Spahn NEVER led the league in WAR for pitchers, never.   He did later win a Cy Young Award (1957), and there is no question that he would have won the Cy Young Award in 1953, had there been a vote.   Spahn in 1953 was 23-7, leading the league in ERA (2.10) and ERA+ (188).   Spahn would have won the Cy Young Award, although D-WAR and R-WAR both regard Robin Roberts as the best pitcher in the league.  Roberts was 23-16 with a 2.75 ERA, but pitched 347 innings.   R-WAR regards him not only as the best pitcher in the league in 1953 but as the MVP. 


1952 AL—Bobby Shantz  (both systems)

1952 NL—Robin Roberts  (both systems)

1953 AL—Billy Pierce  (both systems)

1953 NL—Robin Roberts  (both systems)

1954 AL—Early Wynn (D-WAR) vs. Steve Gromek (R-WAR)

Wynn was 23-11 with a 2.73 ERA in 271 innings, led the league in innings pitched.  Gromek was 18-16 with a 2.74 ERA in 253 innings, but led the league in home runs allowed with 26. 

Wynn pitched for the Indians.  The 1954 Cleveland Indians are unique, in that one of their starters (Mike Garcia) led the league in ERA at 2.64, but the TEAM ERA was 2.78.   I don’t think there is anything else like that in baseball history, where the entire staff is basically equal to the best pitcher in the league—and it’s not even a park effect; the Park Factor was 101.  They won 111 games; you don’t win 111 games because you have a pitcher’s park.  They just had a phenomenal pitching staff. 

Baseball Reference WAR has Gromek ahead by a thin margin (5.6 to 5.3), while my new method used here has Wynn well ahead (8.2 to 6.7).   I absolutely believe that Wynn was a better pitcher than Gromek in 1954, and in any other season probably.


1954 NL—Robin Roberts (both systems)

1955 AL—Billy Pierce  (both systems)

1955 NL—Robin Roberts (D-WAR) vs. Bob Friend (R-WAR)

Friend started 20 games, pitched 24 times as a reliever.   As a reliever he pitched 60 innings, was 5-1 with a 1.95 ERA.   A good deal of his value is shielded from this study because we’re not evaluating relievers. 




Suppose that I asked you what was the greatest game pitched by a starting pitcher in the last 98 years.   Many of you would immediately think of the Harvey Haddix game, in which Harvey pitched 12 perfect innings, retiring the first 36 batters before losing the game in the 13th inning.

If I were to state as a premise that the greatest game in my data was pitched by a Pittsburgh Pirate, I would guess that the percentage of you who would assume I was talking about Harvey’s Heartbreaker would go from 30% to something close to 100%.  If I were to further state as a premise that it was a game pitched by a Pittsburgh Pirate in the 1950s, you would have no doubt that this was true, and if I were to assert on top of that that it was a game pitched by a Pittsburgh Pirate in the 1950s against the Milwaukee Braves. .. .well.   And if I were to assert in addition that the opposing starting pitcher was Lew Burdette, that would be more information than anybody needed. 

But actually it isn’t.   Acknowledging that Game Scores may not be perfectly designed to answer this question and may not get this one right, the highest Game Score in my study of 329,988 games, and also the highest Game Score compared to the Target Score, was for Vern Law, against the Milwaukee Braves on July 19, 1955.    Law gave up a home run to Eddie Mathews in the first inning, then gave up an un-earned run in the fourth on a walk to Mathews, a forceout, a strikeout, a wild pitch, and an error by the second baseman.

Not a great beginning, but then, with the help of a double play, he faced only three hitters in the fifth.   With the help of a runner out trying to advance on a passed ball, he faced only three hitters in the sixth.   He now has a 1.50 ERA for the game.   He retired the side in order in the seventh, gave up a meaningless single in the eighth, and retired the side in order in the 9th.   In the bottom of the ninth he personally drove Burdette from the mound.  After a leadoff single to Dick Groat, Law as a batter singled and took second on an error, putting the winning run on third base and sending Burdette to the showers.  An intentional walk loaded the bases, but Gene Freese grounded to third.   The third baseman stepped on third for one out and threw home, trapping the pinch runner for Groat in a rundown.  The game went into extra innings. 

Law gave up singles in the 10th, 11th and 12th, but no runs, dropping his ERA for the game to 0.75.  He retired the side in order in the 13th, walked a batter in the 14th, retired the side in order in the 15th inning—and in the 16th, and in the 17th.   By the end of the 17th he had retired twelve straight hitters.  In the 18th inning he gave up a leadoff double to Joe Adcock but, this being 1955, stayed in the game and retired three tough hitters (Del Crandall, Andy Pafko and Bill Bruton.) 

Bob Friend replaced him in the 19th inning.  Friend gave up a run in the top of the 19th but became the winning pitcher when the Pirates scored two in the bottom of the 19th; you’ve got to love those Winning Pitcher rules.  

Vern Law in the game pitched 18 innings, giving up 9 hits, two runs, one of those un-earned, struck out 12 men and walked two.   His Game Score was 118, highest in my data, although in a sense the Game Score system screwed him.   Think of it this way.   Suppose Law pitched a complete game, 9 hits, 2 runs, 1 earned, 2 walks and no strikeouts; that would be a Game Score of 61.   Suppose that in his next start he pitched a perfect game, striking out 12.   That would be a Game Score of 99.   He’s entitled to 160 points for the two complete games—61 + 99--but he gets only 118 because of the way that Game Scores work.   But 118 is still the highest Game Score in my data. 


COMMENTS (2 Comments, most recent shown first)

@Brock: BB-Ref doesn't use totals from deduced play-by-play (from the logs and newspaper accounts). Retrosheet has him at 48. The missing ones are:

6/4/1949: Johnny "Soup" Lipon grounded into a double play in the 4th with the bases loaded and no one out, with Vic Wertz scoring on the play. (The Detroit Free Press verifies this.) Of more note was Parnell driving in a run on a sacrifice bunt—with runners on first and second. Apparently, it was that kind of game.

6/12/1949 (1st game): Parnell pitched a complete game, with Joe "Glenn" Tipton grounding into a hellion of a double play to end the second and Cass "Elliot Al" Michaels doing the same to end the third.

7/22/1949: Another complete game, with Paul "Tom" Lehner the victim of a twin killing at the bottom of the 2nd. I'll spare the double-checking on these last two games since they're complete games so we know Parnell was the pitcher when the double plays happened.
7:46 PM Feb 25th
Brock Hanke
I took a look on the BB-Ref site Play Index for most GIDPs in a pitcher season. There's a data discrepancy somewhere. They list Parnell, in 1949, as having "only" 44 GIDPs behind him, which is still third all time. Their #1 is Gaylord Perry, with 47 in 1973, which would be one behind Parnell's 48 in your data. I have no idea where the discrepancy comes from.
3:13 AM Feb 24th
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy