The 1931 Season as seen through the Prism of Player Won-Lost Records
I have recently updated Player won-lost records to include more recent play-by-play data that was released by Retrosheet this summer. This includes play-by-play data for the majority of games played in three seasons prior to 1940, including the 1931 season.
The 1931 season saw Lefty Grove win 31 games (in only 30 starts - interesting usage pattern by Connie Mack) and lead the 107-win Philadelphia A's to their third consecutive World Series, where they lost in seven games to the St. Louis Cardinals.
So how do Lefty Grove and the rest of 1931 major-league baseball look like through the prism of Player won-lost records?
The Best of 1931
I calculate Player won-lost records two ways: pWins, which tie to team wins and eWins, which control for context and the ability of one's teammates. For players with more pWins than eWins, their Player wins contributed to more team wins than one might expect; for players with more eWins than pWins, just the opposite is true: their Player wins translated into fewer team wins than expected. Or more briefly: a player with more pWins than eWins was better in context, a player with more eWins than pWins was worse in context.
The top 10 players in pWins above Positional Average and Replacement Level were as follows.
The top 10 players in eWins above Positional Average and Replacement Level were as follows.
As of August, 2013, Retrosheet has released play-by-play data for 78% of all major-league games played in 1931. As with the 1927 season, the missing games are not distributed evenly. To attempt to control for this, the next two tables mirror the previous two tables except that Player won-lost records have been scaled to a consistent 154-game season. This scaling is done at the level of the team, so, for example the Player won-lost records of all of the members of the Chicago Cubs are adjusted up based on the 132 games for which I have play-by-play data (i.e., the numbers for all Cubs players are simply multiplied by 154/132 = 7/6); adjustments are not based on the number of games played by individual players.
The top 10 players in normalized pWOPA, pWORL, eWOPA, and eWORL in 1931 are shown in the next four tables.
The top 10 players in eWins above Positional Average and Replacement Level were as follows.
The top four players in eWins over positional average in 1931 were the top four players in (adjusted) pWins over both positional average and replacement level in 1927.
Lefty Grove's 1931 season is the only season for which I have calculated Player won-lost records in which a player has earned at least 20 pWins and fewer than 10 pLosses. This is somewhat misleading, however, as I am missing play-by-play data for 7 of Grove's 1931 appearances and it is all but certain that including these games would push Grove over 10 pLosses for the season.
Even setting aside that distinction, however, Lefty Grove is on a very short list of players who earned at least 20 pWins in a season and had at least twice as many pWins as pLosses. The list of all such players for whom I have calculated Player won-lost records is shown in the next table, sorted by Player winning percentage. No adjustments for missing games or schedule length have been made in this table.
If the adjusted results above are to be believed, Lefty Grove's 1931 season might have been the best season ever as measured by Player won-lost records. For example, his adjusted 10.2 pWORL would be the only season over 10 pWORL for which I have calculated Player won-lost records, besting even such legendary seasons as Babe Ruth's 1927, Barry Bonds's 2001, and Sandy Koufax's 1965.
In fact, however, the adjusted results here are almost certainly not to be believed for Lefty Grove. For the Philadelphia A's, I have play-by-play data for 109 games (out of 153 games they actually played). Based on that, the adjustment factor used to blow up Lefty Grove's numbers here was equal to (154/109), or 1.4. But we have play-by-play data for 34 of the 41 games in which Lefty Grove played: that would imply a player-specific adjustment factor of (41/34), or 1.2, for Grove.
Moreover, while we do not know precisely how Lefty Grove performed play by play in the seven missing games, we do know generally how he performed. For those 7 games, Grove started 5, completing all of them, pitching 50.2 innings with an ERA of 2.49 and a traditional won-lost record of 4-2. Those are excellent numbers, of course.
But in the 34 games for which I have play-by-play data, Grove started 25 games, completed 22, pitched a total of 238 IP with an ERA of 1.97 and a traditional won-lost record of 27-2.
We're missing half of Lefty Grove's (pitcher) losses in 1931! Now, to be fair to Grove, the two losses that we're missing were complete game losses with final scores of 2-1 and 1-0. I'm sure that Lefty Grove's Player won-lost record in the missing seven games was outstanding. I'm also reasonably sure, however, that it was probably at least slightly less outstanding than the "best ever" level of his performance in the 34 games that we have so far.
Rogers Hornsby vs. Frankie Frisch
On December 20, 1926, the New York Giants and St. Louis Cardinals traded second basemen. The Giants received Rogers Hornsby who'd had something of a down year in 1926 batting only .317 after a five-year run from 1921 - 25 where he batted .402. Hornsby rebounded after this trade and batted .361, .387, and .380 over the next three seasons (and while even an "empty" .380 batting average is pretty valuable, Hornsby's batting averages were anything but empty - he led the league in slugging percentage 8 times in the 1920s and OBP 7 times).
What seems curious, in retrospect, though, is that Hornsby did this for three different teams. In 1927, Hornsby led the league in runs scored, walks, on-base percentage, and OPS. Okay, nobody was calculating OPS back then, but he batted .361, had 205 hits, scored 133 runs, and drove in 125. He finished 3rd in MVP voting. And the Giants traded him on January 10, 1928, for two guys that I've never heard of (Shanty Hogan and Jimmy Welsh).
In 1928, Hornsby won his seventh batting title, batting .387 for the Boston Braves. He also led the NL in OBP, SLG, and (of course) OPS. For all of that, the Braves, who had already been terrible in 1927 won 10 fewer games, crossing from bad to godawful (50-103). The Giants, meanwhile, actually won one more game (93) than they had the season they had Hornsby. For whatever reason, the Braves traded Hornsby away a mere ten months after they acquired him, trading him to the Cubs for five (mostly forgettable and forgotten) players.
Hornsby was named the National League MVP in 1929 and managed to stick with the Cubs for four seasons, including 1931.
Meanwhile, the 1927 Cardinals won 3 more games with Frankie Frisch at second base than they had the year before (although finished 2nd instead of 1st) and Frankie Frisch finished 2nd in NL MVP voting (one spot ahead of Hornsby). Frisch remained a Cardinal for the next 10 seasons, managing them the last 4-1/2, helping them win 4 National League pennants and 2 World Series, including 1931. In 1931, Frankie Frisch won the National League MVP award.
One thing that pWins do that full-season statistics can't necessarily evaluate is the extent to which a player's performance actually translates into team wins. Was Frankie Frisch more of a "winner" than Rogers Hornsby? That's a vague and/or loaded question that almost certainly doesn't have a "right" answer. But I thought that comparing the Player won-lost records of Rogers Hornsby and Frankie Frisch in 1931 might be interesting.
As with Grove, we run into the problem of missing data here. The numbers below are calculated based on the play-by-play data which we have, which covers 108 of the 131 games that Frisch actually played in 1931 and 91 of the 100 games played by Hornsby. Their Player won-lost records are then blown up based on the number of games these two actually played. Implicitly, this assumes that Frisch's and Hornsby's performances in missing games were the same as their performance in games for which play-by-play data exists. Obviously, I have no idea how accurate this assumption is. So, take all of this with a grain of salt: it's a fun little exercise, not a formal argument over who should have been the MVP of the National League that year.
Hornsby was the much better hitter - .331/.421/.574 vs. .311/.368/.396. But Frisch has three advantages in the preceding table: he was a better fielder (by about 0.2 wins), he played 31 more games (151 more plate appearances), and his "expected" wins translated into a couple more actual team wins than did Hornsby's.
The 1931 World Series featured two teams that won at least 100 regular-season games for the first time since 1912. It also featured a rematch of the 1930 World Series, matching up the St. Louis Cardinals and Philadelphia A's in what turned out to be the A's last World Series appearance for 40 years (and two franchise relocations).
The Cardinals won the 1931 World Series in seven games. Somewhat curiously, while the World Series went the full seven games, none of the individual games were especially close. There were no one-run games and the winning team had a lead which they would never relinquish by the third inning in six of the seven games. The exception was Game Six, where the A's broke a scoreless tie with four runs in the top of the fifth inning of a game they went on to win 8-1.
Bill Hallahan and Burleigh Grimes earned two pitching wins apiece in the Series for the Cardinals, while Pepper Martin batted .500 with an OPS of 1.330.
Best of 1931 by Factor and Position
Next, let's look at the top players in (context-neutral, teammate-adjusted) eWins over Positional Average in various aspects of the game. The numbers in this section have all been normalized to 154-game seasons. As above, this adjustment is based on team games for which I have play-by-play data, not individual player games.
Best by Factor: Batting, Baserunning, Pitching, Fielding
There are four basic factors for which players earn Player won-lost records: Batting, Baserunning, Pitching, and Fielding. The top players in 1931 in eWOPA by factor were as follows.
Positional Average excludes pitcher offense
Positional Average excludes pitcher offense
Best by Position
Next, we look at 1931 Major-League leaders in eWOPA by position. The figures shown here only include Player decisions earned while playing this particular position, and include no contextual adjustments (expected or actual).
The best major-league player at all 8 non-pitcher positions in wins over positional average in 1931 is now in the Hall of Fame.
For relief pitchers, context-neutral records may not be the best measure of how good they are, as context can matter a great deal, depending on how a pitcher is used. Here are the top relief pitchers of 1931 in context, in terms of pWins and pWOPA.
Relief pitcher is the first position listed here for whom the best player in wins over positional average is not now in the Hall of Fame. This seems to me to be an accurate representation of the importance of relief pitching in the 1930s. Bump Hadley had a fairly respectable 16-year major-league career. Between 1927 and 1939, this was the only season in which Hadley pitched more in relief than as a starter: for his career, he started 355 of the 528 games in which he pitched.
Jack Quinn turned 48 years old on July 1, 1931.
Finally, here are the best at three oft-forgotten positions that can nevertheless matter: pitcher offense, pinch hitting, and pinch running.
Wes Ferrell is considered by many to be the best hitting pitcher perhaps ever and 1931 was probably his best hitting season ever with career highs in home runs (9) and OPS (.994). He also batted in 48 games while only pitching in 40 games (the numbers here are only for games when he was the pitcher).
Ferrell's 1931 season leads all players in offensive wins for a pitcher for seasons for which I have calculated Player won-lost records and is one of only four seasons in which a pitcher was at least one full win over positional average in pitcher offense. Those seasons are shown below (Ferrell's season is not normalized to 154 games here).
I don't know about you, but I love a list of old-timey baseball players that includes the first names Smead, Bibb, and Footsie.
The Great Teams of 1931
The 1931 season saw the Philadelphia A's win their third straight American League pennant. This "three-peat" had been immediately preceded by three straight American League pennants won by the New York Yankees (including their famous 1927 Murderers' Row).
A couple of interesting notes about the 1931 A's and Yankees. In 1931, the A's won 107 games and the Yankees finished a distant second with 94 wins. In 1932, the two teams switched places completely, with the Yankees winning 107 games and the A's finishing a distant second with 94 victories. Second, although the A's finished 13.5 games ahead of the Yankees in 1931, the Yankees actually had the slightly better Pythagorean record (based on run differential), beating the A's 100 - 97 in "Pythagorean wins".
The next table compares the 1931 Philadelphia A's and New York Yankees. I also added the 101-win World Champion St. Louis Cardinals who were in the midst of a stretch of five pennants in nine years between 1926 and 1934.
As with earlier tables in this article, I have adjusted the results here for missing games. The 1931 Yankees had a somewhat better record (21-10-1) in games for which I do not have play-by-play data than games for which I do have play-by-play data (73-49-1), so the results here probably under-rate them. The A's and Cardinals, on the other hand, have better records in games for which I do have play-by-play data (80-28-1 and 77-37, respectively) than in missing games (27-17 and 24-16, respectively), so the results here could slightly over-rate these two teams.
|Player Won-Lost Records by Team, 1931
|(adjusted to tie to actual games)|
Controlling for context, the Yankees were the best of the three - consistent with their "Pythagorean" record - and had by far the best offense of the three teams. But both the A's and Cardinals converted eWins to pWins at a much greater rate than the Yankees (it's actually very unusual for a team as good as the Yankees to have more eWORL than pWORL) and the result was that the A's got to play in the World Series instead of the Yankees.
All articles are written so that they pull data directly from the most recent version of the Player won-lost database. Hence, any numbers cited within these articles should automatically incorporate the most recent update to Player won-lost records. In some cases, however, the accompanying text may have been written based on previous versions of Player won-lost records. I apologize if this results in non-sensical text in any cases.
List of Articles