I've got an RA+-PythPat record for him of 134-96, with best equivalent seasons in 1929 (18-10), 1931 (15-9), and 1933 (17-9). That's very good but not great effectiveness per inning - guys like Rommel, Luque, Shawkey, and Bender are in the same neighborhood, and each of them has more innings.

Of course, if you look at him that way, he falls well short on career IP compared to everyone who is a serious candidate for us. Is that a fair way to look at him? We haven't yet developed "relief pitcher" as a category or set up standards for it, and of course Marberry wasn't purely a relief pitcher anyway. If we try to bring leverage into it, what difference would that make?

I'll admit that at the moment I have no idea what to do with him.

If a save is roughly 1/3 of a win, he had 101 saves which adds 34 wins, for a total of 182. Not enough!

WARP has him as the best pitcher in the majors in 1929. Win Shares has him tied for 2nd with Red Lucas of Cincinnati (the NL "Cy Young" winner), behind Lefty Grove (who wins the AL Cy). This looks like his only "first-team" All-Star appearance.

Unless one creates a "Relief Pitcher" All-Star slot, in which case he probably will make more appearances, though this would need to be verified.

I admit that closers are not something I've really thought about yet for HOM purposes. I have questions as to how WARP handles them; do they get any extra "high leverage" credit from WARP? I know many people (myself included) thinks WS gives them too much extra credit, at least in relation to other relief pitchers. But it's just as clear to me that a 80 IP relief pitcher is better than 1/3 of a 240 IP starter.

This is an odd spot to bring this, up, but in thinking about potential HOM relief pitchers, I got to thinking about Mariano Rivera. And in thinking about Mariano Rivera, I got to thinking how ludicrous it would be to analyze his HOM credentials without crediting him for his postseason work. Obviously, pre-divisional players had less opportunity to pile up postseason numbers, but just as obviously, these are numbers that should be considered, and I don't think I've given that enough thought.

So which, if any, current HOM candidates have outstanding postseason success on their resume? Given the small sample sizes involved, maybe no one deserves a boost due to this, but I don't want to short a player who may have shined brightly on baseball's biggest stage.

Bill James created an analysis (thru simulation and some judgment) in the Historical Abstract attempted to quantify how much more valuable typical relief innings were than starters throughout different eras. I will try to find the data and post next week; but even back in Firpo's day, relief innings I think were worth maybe 50% more per inning. Which doesn't necessarily mean we bump them all up by that much, since we still must have tother discussions, such as
-- whether pitching a few innings at a time is easier on the ERA, and
-- whether 'replacement level' for a reliever is different; i.e., if Mariano Rivera's usage pattern makes his innings twice as valuable per inning, that doesn't make his WS/WARP twice as high, since Most teams today have a closer who is better than league average.

Having said all of that, I was probably Hoyt Wilhelm's biggest proponent in the Survivor exercise, so I shall make myself knowledgable on Mr. Marberry very soon :)

Having said all of that, I was probably Hoyt Wilhelm's biggest proponent in the Survivor exercise, so I shall make myself knowledgable on Mr. Marberry very soon :)

Having done an analysis of both pitchers, I'm going to be very bullish for Wilhelm when he becomes eligible, while Marberry is in the Very Good pigeonhole. Firpo needed a little more career to bump him onto my ballot.

I agree. At 50% to his IP for "leverage", and he's still only at 3000 IP. Generally speaking, the voters have required at least 4000 innings for a guy with an ERA+ under 120 to earn a vote (Rixey, Welch). 3000 Innings guys are generally need an ERA+ above 120.

Generally speaking, the voters have required at least 4000 innings for a guy with an ERA+ under 120 to earn a vote (Rixey, Welch). 3000 Innings guys are generally need an ERA+ above 120.

I never thought about it in those terms, Matt, but that's about right.

do they get any extra "high leverage" credit from WARP?

I think it's done on a team-by-team basis. Pitchers that have more "decisions" (wins, saves, losses) per inning get additional credit, at the expense of those that have less.

WS gives them too much extra credit, at least in relation to other relief pitchers.

These systems are limited by the fact that more modern stats that indicate increased leverage (holds, blown saves) are not available throughout history. However, these stats do not fully capture leverage either, as "kept it close" and "still tied" can be just as valuable as "held lead".

OTOH, the modern "reliever relay race" does not really exist before the 1980's. Before then, reliever pitching changes tended to be determined by pinch-hitting necessities or ineffectiveness (blown saves might be nice here).

I think the only way to argue for Maryberry as an HOM candidate is to compare him just to "relievers" in his time instead of all pitchers.

Relievers during Marberry's career had ERA's that were .5 to 1.00 HIGHER per season than starter's ERA's. If you use THAT as your baseline, then Maryberry saved quite a few more runs than your average reliever even before factoring in that a lot of other relievers were being used in mopup situations vs. Marberry being used in 'high leverage' situations.

Relievers during Marberry's career had ERA's that were .5 to 1.00 HIGHER per season than starter's ERA's. If you use THAT as your baseline, then Maryberry saved quite a few more runs than your average reliever even before factoring in that a lot of other relievers were being used in mopup situations vs. Marberry being used in 'high leverage' situations.

But weren't the relievers by and large inferior pitchers back then compared to today (which would explain their higher ERAs relative to the starters)?

Who were Marberry's contemporaries? Wilcy Moore? Does Fatty Fitzsimmons count? I know that Lefty Grove pitch some in relief.

Michael Bass asked:

So which, if any, current HOM candidates have outstanding postseason success on their resume?

Since this question seems a little off-topic for the FP thread, I’ve responded on the 1941 Ballot Discussion thread.

There is a lengthy article on Marbarry in the December 3, 1925 issue of TSN. It details his work over the 1924-5 seasons, hi-lighting his contributions in contests that Walter Johnson pitched in 1924, and games that Johnson and Coveleski pitched in 1925.

The article is too long to cover but here is 1924. I have not verified this material.

May 1, 1924. "Blanked the Yankees for three innings after Johnson had been batted from the game, preserving Johnson win.

May 4, 1924: "Shut out the Athletics for 1 1/3 innings without a hit after Johnson batted out. Johnson credited with victory."

May 30, 1924: Stopped the Red Sox on two hits and no runs in 3 2/3 innings. Johnson knocked out, Marberry the winner."

June 13, 1924: "Held the Browns for four innings after Johnson batted out. Johnson the winner."

July 30, 1924: "Stopped the Tigers with no runs and six hits in seven innings. Marberry the winner. Johnson batted for four runs in two innings and score was 4-2 when Marberry entered the game."

August 21, 1924: "Stopped the White Sox for two innings to preserve a Johnson 2-1 win."

August 29, 1924: "Held Yankees for 1 2/3 innings in Johnson victory."

September 20, 1924: "Went in for Johnson after he had been batted for four runs in one inning."

In post #10, I think that KJOK asks one of the most important questions that Marberry's candidacy revolves around: Firpo versus his relief peers.

We know that stud starters were used as ace relievers as well, so should he be compared against Grove and the relief work of stud starters instead of the inferior mop-up pitchers? I think so.

But doing so raises two more questions for me.
1) Did Firpo make so many more relief appearances in his typical season of relief than the stud starters (the Groves) that his role was different enough to weaken the comaparison between himself and the Groves?

and

2) How much of the credit for Firpo's success goes to him as a matter of performance, and how much to Bucky Walters (and later to others) as a matter of elective usage?

This second question is one that we'll have to answer again for Sutter and for Eck when they become candidates.

We know that stud starters were used as ace relievers as well, so should he be compared against Grove and the relief work of stud starters instead of the inferior mop-up pitchers? I think so.

Exactly. Or, better yet:

Firpo Marberry v. Jack Quinn

ERA+: Marberry 116, Quinn 114
IP: Marberry 2067, Quinn 3920
Starts: Marberry 187, Quinn 444
Non-Starts: Marberry 364, Quinn 312

Firpo Marberry is Jack Quinn with 50 more relief appearances and 250 fewer starts, resulting in almost 2000 fewer innings. Jack Quinn isn't classified with the "starters" because he relieved less, but because he started more.

This seems like a place where it would be awesome to have splits on Marberry as a starter vs. as a reliever.... Sadly retrosheet doesn't have boxes for his career.

1923: 2.57 ERA in 14 relief innings.

1924: 2.91 ERA in 96 relief innings. Both led AL.

1925: 3.47 ERA in 93.1 relief innings.

1926: 2.97 ERA in 103 relief innings.

1927: 3.68 ERA in 85.2 relief innings.

1928: 3.71 ERA in 87.1 relief innings.

1929: 3.95 ERA in 41 relief innings.

1930: 4.85 ERA in 26 relief innings.

1931: 4.70 ERA in 44 relief innings.

1932: 3.42 ERA in 81.2 relief innings.

1933: 0.00 ERA in 8.1 relief innings.

1934: 1.86 ERA in 38.2 relief innings.

1935: 9.82 ERA in 3.2 relief innings.

1936: 3.68 ERA in 7.1 relief innings.

His 1924 relief IP and relief ERA are both listed as black ink, the only such times for those numbers in his career. The source for that is almost 40 years old so I do not swear for its accuracy.

As a matter of fact, I would almost have to think that he led the AL in relief innings several other years?

Just eyeballing the numbers for the heart of his career, Marberry certainly seems to have been a better reliever than his overall numbers indicate. He also, however, would necessarily be a worse starter in his 188 starts than his overall numbers indicate. Since for all except two years in the heart of his career he was starting at least 10 games among his relief appearances, I don't think you can add 'extra' relief points without simultaneously subtracting nearly the same amount of starting points.

Columns:
MdSt -- Number of starts by a median ace starter
MdRl -- Number of relief appearances by a median "ace" reliever
MxRl -- Maximum number of relief appearances (G-GS)

Year  MdSt  MdRl MxRl  Most Relief Appearances (Starts)  
----  ----  ---- ----  --------------------------------
1871   28     1    9   Harry Wright (0)
1872   31     3    8   Cherokee Fisher (11)
1873   46     3    8   Harry Wright (5), Cherokee Fisher (5)
1874   55     3    4   Harry Wright (2), Cherokee Fisher (35), Bill Stearns (18)
1875   41     4   10   Al Spalding (62)
1876   50+    3+  14   Jack Manning (20)
1877   50     5    7   Cal McVey (10)
1878   47+    2    4   Mike Golden (18)
1879   60     2   10   Monte Ward (60)
1880   63     7    8   F,Corey (17), G.Bradley (20), C,Foley (28), L.Richmond (66)
1881   52+    4    5   Charles Radbourn (36)
1882   45+    2+   7   Fred Corey (14)
1883   52+    3+  10   Billy Taylor (9)
1884   57     3+   9   Bill Wise (41)
1885   50     2+   7   Pop Corkhill (1)
1886   56     3    7   Tony Mullane (56)
1887   52     3    5   P.Corkhill (0), M.Tiernan (0), J.Cahill (1), J.Ryan (3), A.Terry (35)
1888   50+    2    6   Jimmy Ryan (2), George Van Haltren (24)
1889   48     6   11   Amos Rusie (22)
1890   46     6   10   George Hemming (12)
1891   51     7   13   Clark Griffith (21)
1892   49     6   10   Gus Weyhing (49)
1893   43     8   10   Tony Mullane (39)
1894   41     8+  12   Pink Hawley (41)
1895   40     8+  14   John Malarkey (8)
1896   42     7+  12   Chauncey Fisher (15)
1897   39     7+  15   Red Ehret (19)
1898   41+    6+  13   Bill Dammann (22)
1899   38+    6+  12   Sam Leever (39)
1900   37     8   11   Harry Howell (10)
1901   37     7   12   Willie Sudhoff (26)
1902   36+    6    8   Charlie Shields (19), Roy Evans (28)
1903   36     6+  10   Harry Howell (15)
1904   42     6+  10   Tom Fisher (21), Ed Walsh (8)
1905   37+   10+  18   Clark Griffith (7)
1906   36    10+  21   George Ferguson (1)
1907   37    12   18   Tex Pruiett (17)
1908   36+   14+  19   Rube Vickers (34)
1909   34    15   22   Doc Crandall (8), Steve Melter (1)
1910   35    18   24   Doc Crandall (18)
1911   36+   18   26   Doc Crandall (15), Mordecai Brown (27)
1912   36+   17   27   Doc Crandall (10)
1913   36    22+  33   Doc Crandall (2)
1914   37    20   27   George Baumgardner (18)
1915   35    22   39   Sad Sam Jones (9)
1916   36    25   37   Bernie Boland (9)
1917   37    24   41   Dave Danforth (9)
1918   31    16   28   Dave Danforth (11)
1919   32    21   31   Jean Dubuc (5)
1920   37    22+  36   Bill Sherdel (7)
1921   36+   29   40   Lou North (0)
1922   34+   27   43   Lou North (10)
1923   36+   29+  47   Allan Russell (5)
1924   34    30   37   Allan Russell (0)
1925   32    30   55   Firpo Marberry (0)
1926   33    30   59   Firpo Marberry (5)
1927   34    29   56   Garland Braxton (2)
1928   32    28   37   Firpo Marberry (11), George Smith (2)
1929   33    28+  41   Wilcy Moore (0)
1930   34    32+  39   Joe Heving (2)
1931   33    27   44   Bump Hadley (11)
1932   33    33+  44   Wilcy Moore (3)
1933   34    35   47   Jack Russell (3)
1934   33    31+  45   Jack Russell (9)
1935   33    34   41   Russ Vanatta (17)

Note: Allan Russell, Firpo Marberry, Garland Braxton, Bump Hadley, Jack Russell all pitched for Washington

Looks like it got clipped. Adonis Terry had 35 starts in 1887; he's the real starter of the bunch.

Bill James wrote an article in the original Historical Baseball Abstract (1984) entitled “A History of Relief Pitching.” For those who don’t have, this is what he said about Marberry:

***
Firpo Marberry was a landmark. Marberry was the first truly outstanding pitcher to be used primarily in relief over a period of several seasons. . . Firpo Marberry was not a good pitcher. He was a great pitcher. Whether or not his career had the longevity that is expected of a Hall of Famer, this I don’t know - but for an eleven-year period of time, 1924 through 1934, Marberry was as valuable to his team as any pitcher in baseball except Lefty Grove. His records are deceptive because, in several of his best years, he split his work between starting and relieving. . . He started 187 times in his career, and his record as a starter was terrific (94-52). How many pitchers in the Hall of Fame do you think won more than half their starts? I don’t know, either, but I know a lot of them didn’t. . . He had a career winning percentage of .623, and that without pitching for either the Yankees or the A’s, the two dominant teams in the American League through most of his career.
***

Strong praise indeed.... but Firpo did not make the top 100 pitchers in the NHBA, indicating his views must have tempered over time.

Here is a link to Mike's baseball rants series where he discusses the evolution of relief pitchers. Very interesting series of articles. Doesn't say much about Marberry in the 1920's section, but he kind of stands out in the tables that follow. You need to scroll down to find the article.

The Halls of Relief