I vote to align with HOF as well.

Yes to aligning with HoF eligibility

Yes to aligning with HoF eligibility for me too

Either arrangement is fine with me, but I think it's important that we have a clear policy, so I vote to align with HoF eligibility as well.

At this point it's easier to align, avoids potential headaches on deciding eligibility. So add me as Yes on aligning.

I'd prefer to keep the old rule, but I am very clearly outnumbered.

I count 13 votes to align so far.

If anyone's holding out, let me point out that if we align HoM eligibility with the HoF, there'll be plenty of real-time commentary in mainstream sports media on new eligibles. More information is never a bad thing.

I vote for going with HOF eligibility.

78. Brent Posted: January 11, 2022 at 10:05 AM (#6060591)
Bleed, Yes, if you have updated DRA data, I'd appreciate receiving it. Please let me know how you'd like to transmit it. Thanks

My email is marchand man 34 @ yahoo. com , send me a message, the file is large but small enough to send through regular email, thanks!

Align with HoF

I'll pop in to also vote to align our eligibility with the Hall of Fame

In terms of syncing with the HoF eligibility requirements I just want to make sure a few things are clear that might be assumed or otherwise overlooked.

The election eligibility rules for the HoF:

3. Eligible Candidates -- Candidates to be eligible must meet the following requirements:

A. A baseball player must have been active as a player in the Major Leagues at some time during a period beginning fifteen (15) years before and ending five (5) years prior to election.

B. Player must have played in each of ten (10) Major League championship seasons, some part of which must have been within the period described in 3(A).

C. Player shall have ceased to be an active player in the Major Leagues at least five (5) calendar years preceding the election but may be otherwise connected with baseball.

D. In case of the death of an active player or a player who has been retired for less than five (5) full years, a candidate who is otherwise eligible shall be eligible in the next regular election held at least six (6) months after the date of death or after the end of the five (5) year period, whichever occurs first.

E. Any player on Baseball's ineligible list shall not be an eligible candidate.

Relative consistency with the Hall of Merit's existing eligibility requirements would suggest we're only syncing with HoF eligibility rules 3C and 3D. We're not tossing aside VC players, nor players who played fewer than 10 MLB seasons, nor players on the ineligible list. All of them have been eligible, and would continue to meet eligibility, for the Hall of Merit. We're just aligning with how the HoF handles "no longer playing". Correct?

That is correct.

Just aligning on the first year of eligibility, yeah. So I agree with you, whoever would qualify under C and D and ignoring the other stipulations.

Not that it's likely to happen again, but I'd add that in the case of a Gehrig/Clemente sort of special election, that player would just be eligible during the next HoM vote.

good catch, Jaack, and yes I think at this point we even align there as well.

If 16 votes means the motion carries, then I've finally contributed something to the project! Time to pack it in and take my leave while I'm ahead, it's all downhill from here.

OK, I've made two updates: one, I have cut 3B position adjustment in half for the 60s + 70s, and two, I have zeroed out negative WAA values for seasons age 22 or younger and 38 or older. The biggest gainer from this adjustment was Lou Brock, who moved up 19 spots. The adjustments only had small effects on my ballot (no one moved on or off, just a few trading positions). I don't plan on incorporating any more adjustments, so this is probably my ultimate prelim.

1. Carlos Beltrán (0)
2. Buddy Bell (+1)
3. Sal Bando (-1)
4. Tommy John (0): the WAA adjustment almost got him above the 3B - it's silly (to me) that people will use the lower WAA against him as if his 760 IP and 4.5 WAR after age 40 made his career less meritorious.
5. Bobby Bonds (0)
6. Lance Berkman (0)
7. Roy Oswalt (0)
8. Brian Giles (+1)
9. Robin Ventura (-1)
10. Kevin Appier (0)
11. Chuck Finley (+1)
12. Bob Johnson (-1)
13. John Olerud (+1)
14. Ron Cey (-1)
15. Chet Lemon (0): am not going to incorporate clutch scores. I did for Sosa because his value was like a six sigma outlier, but I am not sure clutch measures anything other than random variation

16. David Wright (0)
17. Mickey Lolich (0)
18. Jerry Koosman (+1)
19. Cliff Lee (-1)
20. Mark Buehrle (0)

Zeroing out negative WAA values makes absolutely no sense for Tommy John or anyone else.

Rob, with all due respect, I think that depends on how you use WAA in your rankings. Here's a thought experiment I've posed before, but illustrates why I don't agree with you.

Player A: 8000 PAs age 23-37, 60 WAR, 30 WAA, that's his full career
Player B: 8000 PAs age 23-37, 60 WAR, 30 WAA, has 2000 additional PAs from two years at the beginning and two years at the end, 6 WAR and -3 WAA in those years

Now, any analysis that says Player B had a less valuable career than Player A is wrong, full stop. But raw WAA says this. I would go further and say that any analysis that says these two players were equals is also wrong, but don't care to have that battle right now. Zeroing out negative WAA seasons can make sense depending on how WAA is incorporated into a ranking. If you don't agree that that is the case in this hypothetical comparison, then we will never agree on the issue at all. Do you object to zeroing out negative WAA for player B in the head-to-heard comparison I invented, and if so, why specifically? If WAA is my primary method of ranking, I would suggest that the wrong thing is to NOT zero them out (under reasonable constraints).

Brooks Robinson had 823 PA in his final three seasons - and an OPS+ a shade under 60 in that span.

me, I just throw that out. so his career OPS+ of 105 gets a modest boost - but I'm not crediting him with the 150 hits, either (in his case, that only drops him to 2,698, so doesn't matter much) or 10 HR (so down to 258). for a player who needed "filler" to get to 2,000 hits, the debits can show a lack of ballast.

that might be a roundabout way of saying I agree more with kcgard (the sequel).

Since it's relevant to John, I think of his 1988 season in particular. He wasn't above average, but he certainly was more useful to the Yankees than Ron Guidry was that year, because he actually was available to pitch and wasn't horrible. Does 1988 helps John's case? YMMV. Does it hurt it? I can't see why it should.

Since the 2022 election concluded, I've been working on a fairly in-depth study of Negro-League position players, trying to develop a fresh and informative angle that would enable us to examine the Negro-League players meaningfully in relation to each other prior to looking at them through the lens of MLEs. The document for the study has grown quite long, such that I'm not sure what the best place is to post it: it's a comparison of a bunch of players, not an in-depth focus on any. I'll post the first parts soon in the 2023 ballot discussion thread and look for feedback about where else (if anywhere) the study as a whole should go.

As a preview for the study, I thought I would look at how the Negro-League players as a whole fared in the 2022 election. By my count, 8 players who spent time in the Negro Leagues received votes:

Ben Taylor: 12 votes, 156 points
Heavy Johnson: 6 votes, 126 points
Don Newcombe: 6 votes, 74 points
Luke Easter: 3 votes, 60 points
Hurley McNair: 3 votes, 51 points
George Scales: 1 vote, 22 points
Hilton Smith: 1 vote, 22 points
Charlie (Chino) Smith: 1 vote, 15 points

NeL-linked candidates received 33 votes overall, a bit more that 1 vote per voter. However, only 20 members of the electorate gave a ballot slot to an NeL-linked player. Since all the candidates elected in 2022 received at least 21 votes, it is not clear whether there is enough support across the electorate to bring about the election of another Negro-League player, even if there were agreement among everyone voting for NeL players about who merited a ballot spot. Therefore, if any more NeL players are to be elected, a stronger consensus would need to develop among the electorate around the worthiness of at least one candidate. I am looking at my project as seeking to help move us toward a clearer sense of the standing of NeL candidates relative to each other. That may then facilitate another round of focused analysis of comparison of the remaining top NeL candidates to the rest of the candidate pool.

Negro-League Stars: 12-Year & 5-Year Rate Studies for Position Players Using Seamheads Data

Overview. This study is an effort to advance the electorate’s analysis of the remaining Negro League position-player candidates relative to one another by looking more directly at the data from Negro-League play available via the Seamheads’ Negro-League database. It’s intended to supplement MLEs by giving a closer look at the data that lies behind them and by doing analysis at a more general level so that many players can be considered together. (I am also looking at pitchers, but their situation is more complicated, so I am not trying to integrate the two studies.) This broad analysis seems to me to be important as we consider the cases for players that we may think are among the last players from the Negro Leagues who are serious candidates for election. We may well have identified the best remaining players, but our knowledge of the Negro Leagues is still quite incomplete, so it’s worth checking that assumption, and the cases of NeL players currently attracting support such as Ben Taylor, Heavy Johnson, and Luke Easter may be clarified by examining their records in the context of the leagues in which they played as well as translated into the context of their National and American league contemporaries.

I’ve focused the study on position player rates as a way to deal with the problem of highly varied season lengths in the NeL data. MLEs address this problem by finding ways of projecting players’ documented performance into standard 154-game schedules, but process requires a lot of estimates, and as we try to make the fine distinctions in value that are necessary to rank the large group of players on the HoM borderline, the validity of those estimates needs careful checking. Rate stats deal with the problem of variability by aggregating short seasons and comparing players across multiple seasons of production by a consistent segment of measure. By comparing equal segments of players’ careers and looking at their rates of production, we can reduce the influence of schedule variability on apparent differences between players. (That influence is not eliminated, of course, only reduced.)

This study focuses on two rates: a 12-year “prime rate” and a 5-season “peak rate.” My review of the data suggests that the 12-year rate is the more meaningful measure, for three reasons.

• A 12-year prime includes a length of career that is sufficient for a player to accrue enough value to be a serious candidate for the Hall of Merit. If a player’s 12-year prime rate is of certain level or higher, then the player is a serious candidate, regardless of what the rest of their career looks like. A five-year rate, on the other hand, may contribute to a player’s HoM case, but it will very rarely include sufficient value in itself for a player to be elected. Therefore, after looking at peak, one always has to look at the rest of the career to determine if a player is a serious candidate or not. With the 12-year period, one can look at a player’s prime rate, and only their prime rate, and ascertain fairly definitely whether the player is a serious candidate for the HoM or not.

• The 5-season peak rate is much more influenced by variability of season-length. If a player has a single great year with a large number of games, surrounded by seasons with a small number of games, their five-year rate will be largely based on a single season’s production. The twelve-year rate is much less subject to single-season influence, although imbalances in playing time across the 12-year period are still possible and would need to be accounted for in an in-depth assessment of a player’s accomplishments.

• Strong primes are rarer and somewhat easier to identify that strong peaks. I am thus reasonably confident that I have found all the NeL position players over the 40-year period of 1909-48 who achieved a 12-year prime rate of 5.0 or better in play documented in the Seamheads database. (Conveniently, there are 40 players for the 40 years.) I don’t have the same confidence that all of the 5-season peaks above a certain level are included

Here are some factors to keep in mind when interpreting the rates presented here.

In this study, 12-year primes are measured purely chronologically. Even if players miss seasons within that period, their prime is still considered as only including 12 years. (There are two exceptions to this standard. If adding a thirteenth season would raise the player’s rate, it would be included. If a player did not play in the NeL during what would be the 12th year of their prime, their first year back in the NeL is included in the measured prime.)

5-season peaks, on the other hand, are constructed in terms of seasons played. That is, no seasons in which a player does not play are counted as part of the five-season peak. Missed seasons are skipped so that every five-season peak has five seasons in it. Some of those seasons may include very few games, but they are not empty.

For both 5-season and 12-year rates, winter ball statistics are considered to be part of the summer season preceding them, not as a separate season. So if a player played in the NeL in 1921 and then played in the CWL in 1921-22, those are counted as a single year or season.

Heavy Johnson provides a salient example of how the counting works. His 12-year prime is considered to be 1916-27, even though he has no playing data for 1917-19 or 1921. His 5-season peak, on the other hand, is considered to be 1916, 1920, and 1922-24. His play in the Cuban Winter League in 1923-24 is considered to be part of his 1923 season. When evaluating the rate data, one needs to keep in mind what it contains, but I have tried as much as possible to avoid making judgment calls about which years to use from a player’s career. I have calculated their rates for each 5-season and 12-year career segment, and the segments with the highest rates in each set are recorded as the player’s peak and prime.

The Seamheads’ data itself differs in a couple of ways from the bWAR and fWAR that we are used to working with. Since MLEs are designed to project players into bWAR, players may look somewhat different in the Seamheads data than in the MLEs because of these differences. First, fielding runs and wins in Seamheads WAR are derived from DRA, not from the total-zone-based measure used by Baseball Reference and Fangraphs for the National and American leagues during the segregation era. DRA has a larger range of values, so fielding can have greater impact in the Seamhead stats than it does in the MLES. Second, the Seamheads’ data includes multiple different league and competitive contexts without competition adjustments, including playoffs and exhibitions against major-league teams. The MLEs exclude play-off and exhibition data and may introduce different competition adjustments for different leagues. In working directly with the Seamheads data, I have used that data without introducing any competition adjustments and without filtering out playoffs and exhibitions. Awareness of those factors may be important when interpreting players’ prime and peak rates.

This study is not aiming to provide a definitive evaluation of the Negro-League players but to provide a data set that can help to clarify some of the evaluative issues and perhaps, at best, to help distinguish players who should be scrutinized in depth as serious HoM candidates from players who were obviously very good but who are equally obviously short of HoM standards.

With the analytical context explained, let’s start looking at the players!


Part 1. 12-Year Prime: The Top 10 Players

This first part is going to be the shortest and the least significant for the Hall of Merit at this point in time of all the parts of this study, but it has value in establishing that the prime rate measure is fairly effective at identifying the top NeL players. These are the top 10 position players by 12-year consecutive prime, as calculated using Seamheads WAR:

                  12-yr      Prime     Prime  Prime
HOM? Player       Prime Rate Seasons   Age    Games
HOM  Gibson, Josh  11.16     1933-44   21-32  665
HOM Charleston, O. 10.49     1916-27   19-30  936
HOM Lloyd, Pop     10.30     1909-20   25-36  527
HOM Torriente, C.  10.15     1913-24   23-34  892
HOM Brown, Willard  9.08     1936-48   21-33  512
HOM Wilson, Jud     8.81     1922-33   26-37  631
HOM Wells, Willie   8.59     1927-38   22-33  694
HOM Stearnes, T.    8.43     1923-34   22-33  809
HOM Hill, Pete      8.37     1908-19   25-36  666
HOM Leonard, Buck   8.17     1934-46   26-38  577

All of the top 10 players by this measure are in the HoM. They are also the only position players of the 1909-48 period to achieve a 12-year prime rate of better than 8.0 WAR/162 games. Moreover, their primes are all well documented, by NeL standards: all of these players’ games played totals are at or above the median games played for the 40 top primes. (Willard Brown’s 512 games is the 21st highest out of the 40 players with prime rates above 5. His is the only prime in the top 10 group with seasons missing, as he spent 1943-45 in military service.) The fact that players with higher numbers of games played also have higher rates suggests that the teams’ interest in fielding the best players possible generally outweighs the tendency of shorter seasons to make higher rates of production possible, at least when the period surveyed is as long as 12 years.

The fact that the top players by this measure are all in the HoM does not mean that this measure is always a sufficient criterion for evaluating NeL candidates, of course. It does confirm, though, that the 12-year prime is one good measure among others for revealing value in the tricky context of NeL data. Part II will look at the 11th to 25th ranked players by 12-year prime, which includes a number of current, serios NeL candidates for the HoM.

Part 2. 12-Year Prime: Players 11-25

Since I’ve introduced the framework for this study, I'll go right to the list of players:

                  12-yr      Prime     Prime  Prime
HOM? Player       Prime Rate Seasons   Age    Games
     Pettus, Bill   7.73     1909-20   24-35  285
HOM  Beckwith, John 7.43     1920-31   20-31  584
     McNair, Hurley 7.37     1912-23   23-34  466
     Taylor, Ben    7.18     1913-24   24-35  706
     Allen, Newt    7.00     1926-37   25-36  590
     Strong, Ted    6.97     1936-48   19-31  372
HOM  Suttles, Mule  6.74     1925-36   24-35  697
     Wright, Bill   6.71     1934-45   20-31  693
HOM  Lundy, Dick    6.66     1918-32   19-33  832
     Dixon, Rap     6.61     1926-37   23-34  473
HOM  Trouppe, Q.    6.48     1933-44   20-31  430
     Johnson, Heavy 6.47     1916-27   25-36  446
     Scales, George 6.39     1923-34   22-33  502
HOM  Oms, Alejandro 6.34     1921-32   25-36  406
HOM  Santop, Louis  6.31     1913-24   24-35  421

This group of 15 includes 6 players elected to the HoM and 9 unelected, including 3 of the NeL position players currently getting votes: Hurley McNair, Ben Taylor, and Heavy Johnson. This, then, is the stratum where the in-out line is being drawn. Since it’s a key area for analysis, I’ll provide contextual details and some analytical observations for all of these players.

Let’s start with the HoM players in this group. John Beckwith is near the top of the group, just slightly below the Big Ten that head the list. Almost all of his value is in his 12-year prime, so he’s a good indicator of where the in-out line sits for prime value alone. Mule Suttles and Dick Lundy are in the middle of this group: their primes are less strong than Beckwith’s, but both have significant career value outside of their primes which helps to lift them over the in-out line. Both are well-documented players with strong reputations, which also helped them to election. Trouppe and Santop are catchers, for whom standards appear to be a bit different. Trouppe has the better rate stats, but a lot of his play was in the Mexican League, so he might need a downward competition adjustment. (Biz Mackey’s prime rate is even lower, so he is not in this group.) The other HoMer in the group, Alejandro Oms, is the one who could most plausibly be viewed as a mistake. A question for Oms is what to do with his play in Cuba prior to 1921. He has no documented play for his age 22-24 seasons prior to the start of his 1925 prime. If those seasons are projected strongly, that could lift him toward a solid career case. Still, if catchers get an adjustment, Oms is the weakest player in this group by 12-year prime, and he does not have as much documented career as Lundy or Suttles.

Then next group of players to consider is the closely related set that appears to include the top unelected candidates (by this measure): the early trio of Bill Pettus, Hurley McNair, and Ben Taylor. Dr. C’s MLEs discount pre-1920 play, so a competition discount could drop these players versus their later competition in the MLEs. By raw Seamheads’ WAR, though, they are on top. It’s worth noting that the level of documentation for these three players is quite uneven. Although Taylor has the lowest prime rate of the three, he has the best documented prime. Pettus played his entire prime in the east before the formal organization of the Eastern Colored League, so his stats are based on an average of only 24 games per season, versus Taylor’s 59-game average. McNair lands between the two with an average of 39 games per season. His career is strongly bifurcated, with very little data for his pre-1920 years but lots of games after 1920 with the Kansas City Monarchs. From 1912-19, he played in only 90 documented games, 59 of which came in 1915, his one season with the Chicago American Giants, but from 1920-23, he played in 376. (1918 and 1919 are missing due to military service.) This means that his prime rate is very heavily weighted toward these four years, his age 31-34 seasons. If he peaked earlier, then his prime rate in the Seamheads’ data may underrate him. An earlier prime would be typical, but a number of sluggers who were in their early 30s when the ball got livelier in 1920 had a late peak then—Ben Taylor, for example—so the evidence about McNair’s prime rate being lowered by being weighted to the latter third of his prime is somewhat ambiguous. (The 5-season peak part of the study will revisit this issue.) The presence of these three players at the top of the unelected group confirms that Taylor is a player worth considering, but it complicates any claim that he is the last, best viable Negro-League candidate.

The largest group here is outfielders with middling career length: Ted Strong, Wild Bill Wright, Rap Dixon, and Heavy Johnson. Their careers weren’t short, since they all played long enough to have their twelve-year prime be measured, but none of them have much value outside that prime. If John Beckwith represents the lower bound of prime rate that will get a player elected on 12 years of play alone, then this group of players may be clearly on the “out” side of the in-out line. Thus, we need to examine their cases to see if there is evidence of more value in their careers than is revealed in the 12-year prime measurement.

Ted Strong’s prime looks the best of this group, and he was a great all-around athlete, playing professionally for the Harlem Globetrotters in addition to his Negro League career. He didn’t stay in organized baseball after the dissolution of the Negro Leagues, however, as far as I can discover, except for a short stint in the Mandak league in 1950. With a lower prime rate than the early trio, and lacking their additional career value, it’s hard to see how Strong could place ahead of them or ahead of the top unelected players of the 1940s from the National and American Leagues. In assessing Strong’s prime, there are a couple of other factors to consider. One is that he missed the 1943-45 seasons due to military service, and he played the 1940 season and part of the 1941 season in Mexico. Since his performance is similarly strong on both sides of his stint in the military and he wasn’t in Mexico very long, I don’t think either of these factors affect his prime rate very much. If Strong has a post-1948 playing career deserving of MLE credit, then he could be a serious HoM candidate, as his prime is a bit better than those of Suttles and Lundy. (I’ve found that there’s an excellent biography of Strong, Ted Strong, Jr.: The Untold Story of an Original Harlem Globetrotter and Negro Leagues All-Star, by Sherman Jenkins, that probably goes into what Strong did in professional sports after 1948, but I haven’t looked at it yet.)

Wild Bill Wright’s case is tricky because his play in the Mexican League is such a large part of his value. His 362 games in the MxL in 1940-41 and 1943-44 constitute over half of his prime games (and most of his peak games), and he played in the MxL exclusively after 1945. With his time in the MxL fully considered, his career isn’t actually short, but his post-1945 MxL seasons aren’t included in the Seamheads database, so there’s a question of how much value to assign to him for his later career. Since the majority of his prime come in Mexico, there’s also a question of competition discounts. If one credits Wright’s full Mexican League career without discounts, then he looks a lot like the elected Mule Suttles and Dick Lundy, but if the MxL is discounted below the NeL, then it looks like he would fall below them on both prime and career.

With Rap Dixon, there’s no question of competition levels, as his career came during the period in which competition levels in the Negro Leagues may well have been at their highest. Dixon’s career is fairly short, however, and he has little value outside his prime. Unless competition levels drop the candidates above him, it’s hard to see how he could match Pettus, McNair, or Taylor, who have stronger primes and stronger careers. And, I should add, if those players’ values are lowered due to competition-level considerations, that doesn’t bring Dixon any closer to John Beckwith, whom he trails in prime rate by over 1 win per year: that’s a lot of value over 12 seasons. (If it’s judged that the remaining 1910s stars aren’t as good as they appear in the Seamheads data, it’s possible that there are no remaining serious Negro-League candidates from the 1909-48 period. Though see Newt Allen below.)

For Heavy Johnson to be a serious candidate, significant adjustments to the Seamheads’ data are needed, as he’s behind quite a number of other players by prime rate, and he has little value beyond his prime. The current MLEs for Johnson make two such adjustments. The negative impact of his fielding on his value is substantially reduced by the current process of norming NeL fielding values to Rfield, and the value of his peak and prime are increased by filling in his missing seasons playing for the Army Wreckers with seasons that approach his two top seasons in 1922 and 1923. If these adjustments are reasonable and appropriate, then Johnson comes up toward Beckwith’s level.

The last group in this candidate set to consider is the second-base pair, Newt Allen and George Scales. Using only the Seamheads data, it’s hard to see how Scales could rank ahead of Allen. Allen’s prime rate is both higher and better documented than Scale’s prime rate. As the two were almost exact contemporaries during a period in which there are few questions about competition levels in the NeL, there aren’t any competition-level factors to raise Scale over Allen. In addition, both Allen and Scales have long, modestly productive careers beyond their primes, so Scales doesn’t gain anything in career against Allen. The only way to bring Scales ahead of Allen, I think, is to discount fielding value more heavily than batting value in converting the Seamheads’ data into MLEs. Scales is a bat-first infielder, while Allen is glove-first. If the MLEs show them as very close in value, the reduction of Allen’s fielding advantage in the MLEs is likely the reason. The twelve-year prime measure shows Allen as a very strong candidate. He is not far behind the top trio, even though some fielding value is missing from his prime (there is no fielding value for his 1927, 1929, and 1930 seasons, except for the positional adjustment). Where he stands as a candidate depends on how much trust one places in the fielding evaluation. If one takes it at full value, then he’s very close to Pettus, McNair, and Taylor in prime. As he has a Suttles/Lundy type career with a better prime than theirs, he looks potentially above the in-out line by the Seamheads numbers. It’s harder to see a case for Scales in the Seamheads data, however, as he trails both Suttles and Lundy in prime.

Here's a summary review of what I think a prime rate study shows about this set of players:

Pettus—Serious candidate. Top by prime rate, but there are questions about competition levels and also about the thinness of the data for his whole prime.
McNair—Serious candidate. Questions about competition levels and the thinness of data for 1912-17.
Taylor—Serious candidate. Questions about competition levels.
Strong—Needs post-1948 credit to be a serious candidate.
Wright—Needs post-1945 credit to be a serious candidate. MxL competition levels are also a key consideration.
Dixon—Appears to fall a bit short, and there are no sources of underdocumented value to bring him up.
Johnson—Falls short by the Seamheads data; adjustments to fielding value and filling in missing seasons could make him a serious candidate.
Allen—Serious candidate. Questions about fielding values from Seamheads’ use of DRA.
Scales—Falls short by the Seamheads data.

Part 3. 12-year Prime: Players 26-40

This post completes a review of the top 40 NeL position players by 12-year prime rate, 1909-48. (The next parts will look at 5-season peak.)

                  12-yr      Prime     Prime  Prime
HOM? Player       Prime Rate Seasons   Age    Games
     Parnell, Roy   6.28     1927-38   23-34  374
     Robinson, Neil 6.09     1936-47   28-39  242
     Brooks, Irvin  5.92     1920-31   29-40  213
     DeMoss, Bingo  5.91     1912-23   22-33  595
     Poles, Spotts. 5.88     1910-21   22-33  372
     Lyons, Jimmie  5.87     1913-24   23-34  478
HOM  Mackey, Biz    5.74     1922-33   24-35  638
     Pierce, Bill   5.56     1913-24   23-34  300
   Davis, Steel Arm 5.56     1920-32   24-36  417
     Patterson, Pat 5.55     1934-46   22-34  391
     Malarcher, D.  5.36     1923-34   28-39  505
HOM Bell, Cool Papa 5.20     1931-42   28-39  573
     Bankhead, Sam  5.15     1932-43   21-32  616
     Cannady, Rev   5.15     1922-33   20-31  497
     Harris, Vic    5.01     1926-37   21-32  586

Since the analysis of the 11-25 group has strongly suggested that the borderline falls in that stratum of the 12-year prime rankings, and all of these players are below that, there’s less need to analyze this group closely, but it’s useful to see that they are here and to confirm that this group of players is consistently below the 11-25 group in value.

Let’s start with the two members of the HoM in this group. Biz Mackey’s prime rate is (relatively) low, but given other factors, he’s probably not a “mistake.” As we’ve already seen, rate standards for catchers are a bit different from those of other position players. In addition, Mackey has a considerably longer career outside his prime than either Trouppe or Santop, and he has a very strong reputation. Cool Papa Bell, on the other hand and like Alejandro Oms, seems like he probably falls into the “mistake” category. He has a long career outside his prime, but his prime rate is a full win lower even than Oms (and a half win below Mackey), and it is brought up above 5.0 only by a very big year in the MxL. Bell’s reputation helped him get elected, but as the data for his career has become more and more complete, it hasn’t corroborated his reputation.

For the other players, it’s worth noting that, generally, they have fewer games in their primes than the players with higher prime rates, and a number of the players here have gaps in their record. Roy Parnell is missing four years out of his prime—years in which he was playing in the Texas Negro League or the Negro Southern League instead of one of the NeL major leagues. Neil Robinson didn’t play consistently in the NeL major leagues until his age 28 season, but had bounced between NeL teams, semi-pro teams and the Negro “minor” leagues since his age 20 season. It looks like he stuck in the NeL once he really learned to hit for power, but there could be missing value in the 8 years prior to the start of his late prime. Similarly, Irvin Brooks doesn’t show up with top NeL teams before his age 26 season. I have no information about what he was doing prior to his sudden appearance in the 1917-18 Florida Hotel League, so who knows what is missing there? Brooks is very obscure now, but Cumberland Posey named Brooks to his all-time NeL all-star team. Steel Arm Davis is missing several seasons from his prime as well. He stayed in the Texas Negro League until a fairly advanced age, and then he moved back and forth between NeL teams and barnstorming teams, where he was a headliner in part because of his skill in clowning, which might have made him a more marketable asset in that context than in the Black majors, although he was a top hitter for several seasons. Pat Patterson missed what would have been the last three years of his 12-year prime to military service in WW2. The histories of players like Parnell, Robinson, and Davis especially suggest that, for players with very good talent but who were still a clear step below HoM tier, being the top player in a context of lesser competition may sometimes have been a more attractive financial option than playing a supporting role in the NeL majors. I don’t think that their sticking with the NeL majors would have added the 1.5-2 WAR/162 needed to make their rate of production competitive with the top remaining NeL candidates, but these are players for which a gap-filling MLE might show more strength that this rate study does.

For the rest, the story is simpler: they were very good players. Poles and Lyons were medium-career-length outfielders who were a win or so per 162 games below players like Ted Strong and Rap Dixon. Bingo DeMoss is sort of Newt Allen-lite. When he hit at league average or better during his short peak, he was outstanding, but outside that peak, when he didn’t hit much at all, he was merely good. Bill Pierce is a first-base contemporary of Ben Taylor and Bill Pettus. Dave Malarcher, Sam Bankhead, and Rev Cannady were very good infielders. Bankhead is probably the best of this trio because of his long career, but their primes are all pretty close in value. Vic Harris, who rounds out the group of 40 top primes from the Negro Leagues, was a fixture in the outfield of the Homestead Grays for many years (his full playing career was 25 years!), and he, like Dave Malarcher, was a highly successful manager as well as a star player. He was very good but is pretty distant from the level of performance needed to be a strong candidate for the HoM.

To sum up what I see as the findings from this group: the top three players in it—Parnell, Robinson, and Brooks—could be worth more investigation through MLEs or historical research that could help to fill in the gaps in their performance record. Their situation is similar to Heavy Johnson’s. Although the prime rate platform from which they begin is a bit lower than his, they are not so very far behind: .2 WAR/162 for Parnell, .4 WAR/162 for Robinson, and .55 WAR/162 for Brooks. None of the other players in the 26-40 group have similar prospects for additional value, so they are situated clearly below the in-out line.

Outstanding stuff, Chris.

Agreed, great stuff. Since I usually induct more pitchers PHoM than the larger group I'm more interested in the marginal pitchers than the marginal position players but I agree, great analysis.

A few other things I gleaned from that chart

Bill Pettus and Ben Taylor are first basemen with similar rate production during their prime who also had overlapping seasons. They're direct contemporaries who are ripe for a head-to-head comparison.

More evidence that we overrated Mule Suttles at the time of his election. There were a lot of questions regarding Cool Papa Bell's offense (or lack thereof) when we elected him but the surprise here is how mediocre his fielding rates considering his speed. FWIW - Mule and Cool Papa are evidence that nicknames help.

If Newt Allen was a SS we probably would have elected him already.

I need to re-assess Willard Brown to see if he makes my PHoM. IIRC his big seasons are during the Negro American League expansion years.

2023 Prelim

You will see some particular areas where I am going to differ from group consensus - I think pitchers are quite under-represented in the Hall of Merit. I give war credit and also credit for minor league seasons played at a level that indicates the player would have been above average in the major leagues. I think the consensus right now is to NOT give war credit which is a shame. I use Dan R's standard deviation adjusted WAR (when available). I like the positional average and positional replacement value calculations in his spreadsheet as well. I also believe a season is a season whether 60 games or 162 and adjust shortened seasons to a 162 game baseline. I think we have enough pre-1890 players and generally identified the correct ones.

I look at two main things - value above replacement and value above average. I'm not as interested in a (theoretical) 65 WAR player with 0 WAA as I am with a 55 WAR 25 WAA player. Therefore I zero out seasons at the beginning and end of a player's career where WAA + WAR < 0. I don't care about peak (consecutive or non-consecutive) and it doesn't really enter into my calculations.

I haven't ever boycotted due to steroid usage. I have penalized players who gambled by applying the lifetime ban for the offense to immediately after the time when it occurred.

1) Tommy Bridges - I give war credit for two seasons at the level he was pitching 1941-1945. That gives him roughly 60 WAR and 30 WAA. PHoM 1958.

2) Carlos Beltran - very similar value to Jim Edmonds with Beltran's superior postseason record breaking that tie. PHoM 2023

3) Mark Buehrle - My biases explained above are going to like a guy like this. FIP WAR is going to miss his plus defensive contributions. PHoM 2021.

4) Bob Johnson - Giving him 1.5 seasons of credit for minor league performance as he was performing at a level above MLB average. I've been voting for Bob Johnson every single year I have participated in
the project. PHoM 1986.

5) Phil Rizzuto - gets 3 full seasons of WWII credit at the level of his average performance from the surrounding 9 seasons. That doesn't give him any extra credit for playing 1946 with malaria. Fantastic fielder with just enough bat. PHoM 1967

6) Urban Shocker - giving him 1/2 season credit for WWI. Good bat for a pitcher helped him in my recalculations. PHoM 1968.

7) Bert Campaneris - If you look at wins above positional average instead of BBREF wins above average you will like 1970s SS more than 1970s 3B. Good fielder and great baserunner - his bat is basically average but his baserunning makes his offensive contribution a positive. PHoM 1991.

8) Kevin Appier - Another pitcher with solid performance above average. PHoM 2009

9) Dave Bancroft - another mostly glove SS with extra credit (50% for 1919). PHoM 1976

10) Tommy John - a compiler with only a few top seasons but does well with STDEV adjustments, postseason bonus. PHoM 1995

11) Roy Oswalt - contemporary of Johan Santana who was a similar high peak, short career pitcher. PHoM 2019

12) Brian Giles - Perhaps the modern day version of Bob Johnson - he was an average contributor right away and provided a ton of value from 2000-2005. Zero out his last season and he's 30WAA. Has an argument for minor league credit but I'm not giving it to him on this placement. PHoM 2020.

13) Bucky Walters - overlooked pitcher from WWII era. Very good hitter for a pitcher which helped him in my re-evaluation. PHoM 1972

14) Ben Taylor - moves down after latest MLE adjustments. He's Rafael Palmeiro of the deadball era. Also compares well to Keith Hernandez. Better than Mule Suttles. The last obvious Negro League candidate. PHoM 1973.

15) Tim Hudson - another pitcher that fits my system well. PHoM 2021

16) Norm Cash - Terrific fielding 1B with a plus bat and one monster season. PHoM 1997
17) Bus Clarkson - Needs credit for NGL, Mexican League and time missed due to the war as well as minor league credit for integration quotas. He's likely an average fielding 3B or below average at SS but the bat is a plus at either position. PHoM 1967
18) Frank Tanana - will probably pull the trigger and make him and Kenny Lofton my other two PHoM choices in 2023.
19) Johnny Pesky - 3 seasons WWII credit. PHoM 2005
20) Gavy Cravath - 4 seasons minor league credit. PHoM 1927
21) Jorge Posada - PHoM 2022 - even with the glovework lacking there is too much to like. Postseason bonus helps.
22) Wally Schang - PHoM 1987
23) Don Newcombe - gets 2-1/2 years of war credit for serving 1952-54. Gets 1/2 year of minor league credit for being held back due to integration quotas. Good bat. PHoM 2004.
XX Kenny Lofton - PHoM likely this year
24) Dave Concepcion
25) Tommy Leach - PHoM 2009

34) David Ortiz - Borderline choice. Unlikely to make my PHoM soon with 15 players ahead of him in line including Ducky Medwick, Will Clark and Vlad Guerrero.
57) Bobby Bonds
66) Buddy Bell - not much value above 3B positional average
74) Lance Berkman - not quite as good as Chuck Klein. Similar to Jack Clark.
78) Thurman Munson - career cut short hurts a lot in these rankings when players are bunched together this tightly. There is not much separation between 40th and 80th.
91) Vic Willis - adjusted for standard deviations he's 46 PWAR and 21 WAA. Terrible hitter.
141) Sal Bando - Concepcion and Campaneris are preferred for 1970s infielders

Here's my "clump of off-ballot NGL pitchers I really can't tell apart"

Webster McDonald
Roy Welmaker
Juan Padron
Leroy Matlock
Dave Barnhill
Bill Holland
Ramon Bragana
Leon Day
Bill Byrd
William Bell
Hilton Smith

Then there's the multipositional Lazaro Salazar who is even harder to place.

Some Replies to DL from MN, who wrote:

Bill Pettus and Ben Taylor are first basemen with similar rate production during their prime who also had overlapping seasons. They're direct contemporaries who are ripe for a head-to-head comparison.

Yes, this is sort of study I was hoping that the broad-based survey of rates would turn up. I think that Hurley McNair, as a corner outfielder who is also a direct contemporary, could be included in the direct comparison as well, although there is the key and uncertain detail of position adjustments.

More evidence that we overrated Mule Suttles at the time of his election. There were a lot of questions regarding Cool Papa Bell's offense (or lack thereof) when we elected him but the surprise here is how mediocre his fielding rates considering his speed. FWIW - Mule and Cool Papa are evidence that nicknames help.

I don't think the rate study is clear evidence that Suttles was overrated, unless you mean that when he was elected he was viewed as one of the all-time great NeL power hitters on account of the home run record? He's not at the bottom of the elected group, and it's far from definite that any of the players who rank above him by prime rate in the Seamheads data are in fact better than he is. Also, his prime rate suggests that the conversion factor from NeL to white majors would have to be quite steep -- say below .75 -- before Suttles wouldn't be a good comp for the lower-tier long-career HoM electees from his time period. (I'll be sharing more data on NL/AL 12-year prime rates as part of this study.)

I agree that nicknames help. (Amusing aside on nicknames: the Seamheads page for Steel Arm Davis indicates that his nickname was "Lefty.")

If Newt Allen was a SS we probably would have elected him already.

Maybe. He certainly would have been assigned more fielding value in the first-generation MLEs. However, his weak career hitting (106 career OPS+) is below any of the elected shortstops (Lundy OPS+ 124, Wells OPS+ 139, Moore OPS+ 149, Lloyd OPS+ 141, HR Johnson OPS+ 158), and it's been a question for his candidacy as to whether his bat would have played in the NL/AL. A study based just on NeL data doesn't engage questions of that kind. Some would argue that such questions are irrelevant, but I think a lot of voters see them as important.

I need to re-assess Willard Brown to see if he makes my PHoM. IIRC his big seasons are during the Negro American League expansion years.

As you'll see from peak rate data when I share it, his big seasons were later: his 5-season peak rate occurred 1941-43, 46-47.

hey DL, where do you have Pettitte?

hey DL, where do you have Pettitte?

Pettitte was elected.

of course, already forgot :)

NeL Rate Study, Continued.
Part 4. 5-Season Peak: Players above 10.0 WAR/162

Now, let’s turn from prime rates to peak rates. As I discussed earlier, this measure tells less in itself, but it does provide cross-check on the prime rate study and enable us to identify some more players worthy of further investigation. The top peak group is larger and less neat than the top prime group, so I’m going to look at them in cohorts segmented by rate, with cut-offs at 10, 8, and 6 WAR/162.

There are 11 players in the 10.0+ WAR/162 group:

                  5-yr       Peak      Peak   Peak
HOM? Player       Peak Rate  Seasons   Age    Games
HOM  Gibson, Josh  13.47     1936-40   24-28  196
HOM  Torriente, C. 13.31     1917-21   23-27  286
HOM  Lloyd, Pop    12.92     1916-20   32-36  208
HOM Charleston, O. 12.06     1918-22   21-25  404
HOM  Moore, Doby   10.97     1920-24   24-28  449
HOM  Brown, Wil.   10.55  1941-3, 46-7 26-32  218
HOM  Wilson, Jud   10.37     1927-31   31-36  300
HOM  Leonard, Buck 10.32     1937-41   29-33  223
     Smith, Charl. 10.23     1925-29   23-27  201
HOM  Wells, Willie 10.12     1927-31   22-26  418
     Wilson, Artie 10.11     1944-48   23-27  133

This group is very similar to the top 10 by 12-year prime rate. 8 players appear on both lists, but the peak rate list drops Pete Hill and Turkey Stearnes and adds Dobie Moore, Charlie (Chino) Smith, and Artie Wilson. Hill and Stearnes are in the next group, with rates in the 9-10 range, which is still very strong. The three players with a peak rate of 10+ but who don’t appear on the 8+ prime rate group suggest both the importance and limitations of peak rate in the NeL as an indicator of merit.

Dobie Moore’s HoM case is mainly based on his 5-season peak, but it is not limited to that. His peak level of production extended beyond his five seasons until, as is well known, his career was cut short by a catastrophic injury. With a reasonable amount of additional value projected from his pre-1920 years with the Army Wreckers, he fits in easily with the rest of the NeL HoM contingent. Moore also benefits from having the best-documented peak in all the top 10. The fact that this peak came for a top NeL team additionally reinforces his case: the certainty that he was a truly outstanding player is quite high.

Charlie (Chino) Smith’s case is less clear than Moore’s. His career was cut short by his untimely death due to illness. He has one full season of play after his 5-season peak, so his career is much shorter than Moore’s and would be shorter than any elected player’s. Also, Smith spent much of his career with a relatively marginal team—the Brooklyn Royal Giants—during a period in which the Eastern Colored League was coming apart. Thus, the number of league or league-equivalent games he played was small—about 40 per season. Because of this playing context, there’s also some uncertainty about whether he was really as good at his peak as contemporary players like Jud Wilson and Willie Wells, who have similar peak rates but in many more games.

That uncertainty is heightened by Artie Wilson’s presence in the 10+ peak rate group. Wilson’s peak rate is based on the fewest games of any in the group, and it comes in the last five seasons of the NeL. The meaningfulness of Wilson’s peak rate is undercut by the fact that he was not able to transition to a successful career in the National or American League after the dissolution of the Negro Leagues. An extreme slap hitter with great speed but no power whatsoever and only moderate plate discipline, his batting style brought him great success in the Negro Leagues, where the lower quality of fielding and the suppression of power accentuated the value of putting the ball in play. The game in the American and National Leagues, however, rewarded power and plate discipline and took away many of the small-ball hits the Wilson was able to accrue in the NeL. Wilson enjoyed a long and successful career in the PCL, but not in the integrated majors. The apparent height of his peak in the Negro Leagues raises questions about the meaning of Negro-League success for Hall of Merit evaluation. Such questions create uncertainty around a player like Charlie Smith and have even provoked some doubts about Willard Brown’s case, as he, like Wilson, was not able to continue his NeL stardom in the major leagues. Brown’s and Wilson’s cases are significantly different, however. Brown was asked to jump straight from the NeL to the American League, the team environment he entered was not supportive, and he was at an age at which declining quality of play would be likely. Wilson was still in his 20s when he got his shot at the majors, he had 1.5 seasons in the PCL to ease his transition, and he played for a team that had already given at least minimally sufficient support to other Black players as they had integrated the team. Wilson’s case highlights that the Negro Leagues were different from the American and National leagues in contextual features as well as in competitive quality, such that transfer of quality from one context to the other was neither straightforward nor guaranteed.

In sum, the cases of Charlie Smith and Artie Wilson show the limitations of 5-season peak rate as a stand-alone indicator of merit. Although they had great peaks in the NeL, neither of them looks like a serious candidate for the HoM, though voters who assess peak value only may support Charlie Smith.

Part 5. 5-Season Peak: 8-9.99 WAR/162

In addition to this range, I’m including the next two players below 8.0 because they are benchmark players from the prime list.

                  5-yr       Peak      Peak   Peak
HOM? Player       Peak Rate  Seasons   Age    Games
     McNair, Hurley 9.72     1912-16   23-27   85
HOM  Irvin, Monte   9.64  1940-2, 45-6 21-27  220
HOM  Hill, Pete     9.59     1915-19   32-36  282
     Brooks, Irvin  9.54     1920-24   29-33   90
     Pettus, Bill   9.41     1916-20   31-35  124
HOM  Stearnes, Tur. 9.00     1932-36   31-35  234
HOM  Beckwith, John 8.83     1921-25   21-25  277
     Strong, Ted    8.37     1938-42   21-25  236
     Wright, Bill   8.30     1940-44   26-30  414
HOM  Lundy, Dick    8.20     1918-22   19-23  156
     Dixon, Rap     8.10     1927-31   24-28  256
     Jethroe, Sam   8.10     1944-48   27-31  112
     Parnell, Roy   8.01 1927-8,32,34-5 23-31 259
     Taylor, Ben    8.00     1919-23   30-34  340
HOM  Suttles, Mule  7.93     1926-30   25-29  378
HOM  Oms, Alejan.   7.85     1921-25   25-29  225

This is a complicated set of players. Most of the names here are familiar from the list of the top 25 players in 12-year prime. There are some shifts in the order that need consideration, and, as with the players in the 10+ WAR/162 peak rate group, there are a couple of new arrivals: Monte Irvin and Sam Jethroe. Let’s start with them, before checking in with the returning HoM players in the list, and then considering the unelected players.

Irvin and Jethroe are on this list but not on the 12-year prime list because segregation in major-league baseball ended before they had twelve years of playing time in the Negro Leagues. Irvin’s 5-year peak rate is 13th overall, which is a level of production that definitely supports his selection to the HoM. Jethroe’s peak is a bit lower. On the prime rate list, we saw that Dick Lundy and Mule Suttles fall below several unelected players, and that their election thus depends on significant career value beyond their prime. Lundy and Suttles fall in a similar place on the peak list. If Jethroe’s multi-league prime is also similar to theirs (and I expect it would be, given his very good but not great seasons with the Braves in his mid-30s), then he would probably also need a career case for the HoM. For that case to be good, it would depend either on his post-majors play in the minor leagues or on his four years of semi-professional play before he joined the NeL in 1942. Although some voters have been giving his semi-pro teammate Luke Easter credit for these years, to my knowledge no one has tried to build a comparable case for Jethroe. That’s unfortunate, given that Jethroe’s documented play in the Negro Leagues and in the NL/AL is better than Easter’s. I think Jethroe’s case is one that merits further investigation, but the height of his documented peak suggests that he would need to have (1) the equivalent of several solidly above average seasons during his semi-pro years of 1938-41, (2) several solidly above-major-league average seasons in the minors after 1952, or (3) some combination of the two to have a serious HoM case. If his above average play is limited to the ten-year period between his first full year in the NeL—1942—and his second strong year in the major leagues—1951—then he won’t have enough career value. Instead, he would be a close comp for Rap Dixon, with whom he shares a peak rate of 8.1, and who lacks the career value needed to be a serious candidate. I think that’s most likely where Jethroe lands, but I don’t see his case as fully closed, if one is basing their analysis only on what the data from Seamheads and from his stint in the National League show.

As my comments on Jethroe suggests, the presence of the elected HoM players in this group who have also appeared in similar positions in the 12-year prime rankings provides a rough basis for benchmarking the strength of unelected players’ candidacies. As he did in the prime rate list, Beckwith serves as a kind of lower bound for shorter-career HoM players. Hill and Stearnes, who were above Beckwith in the prime rankings, remain above him in the peak rankings as well. Lundy, Suttles, and Oms are again close together and below Beckwith, with Lundy and Suttles suggesting the height of peak needed for a long-career NeL player to be a strong HoM candidate. That’s the context I used to assess Monte Irvin and Sam Jethroe, and so it can be the starting point also for consideration of the unelected players who also appeared in the prime rate rankings. I’ll take those in two groups: those above the Beckwith line and those between the Beckwith line and the Suttles line.

The Above-the-Beckwith-Line group includes the two players who were the top unelected players by prime rate--Hurley McNair and Bill Pettus—and a more surprising third member, Irvin Brooks, who was in the prime rate top 40 but was much farther from the top, placing 28th. His having a peak rate above 9.0/162 but a prime rate below 6.0/162 is an unusually large gap between the two rates, and its causes are worth considering. The first observation to make about this whole group, however, is that it could be subtitled “the small sample size group.” McNair and Brooks are two of only three players in the peak rate set whose rate is calculated from less than 100 games, and Pettus’s rate is based on only 124 games, which is also very low. It is perhaps obvious, but the reliability of a rate of wins above replacement per 162 games drops very considerably when it is based on a number of games that is considerably less than 162! The fact that all of the non-HoM players with peak rates above 9.0 WAR/162 games have a small number of games played suggests that these rates have not been taken at face value (when they have not been overlooked), and perhaps they shouldn’t be. Another factor to consider in the case of McNair is that he was doing a certain amount of pitching early in his career, and so pitching adds to his value during his 1912-16 peak. Having value as a pitcher is, of course, a perfectly legitimate thing for a Negro-League player to have, even though in most cases it probably wouldn’t have carried over into play at the National or American League level of competition, but since pitchers accrue value at different rates per game than do position players, the inclusion of a bit of pitching value in a small set of position player games can significantly affect the player’s rate. If Hurley McNair’s pitching WAR (about 0.4) is removed from his 1912-16 value, his peak rate drops from 9.72 WAR/162 to 8.96 WAR. That gives a sense of the volatility of these rates when they are based on so few games. The fact that McNair and Pettus have such apparently strong peaks doesn’t hurt their status as serious candidates, but with their peaks coming in a small number of games in the 1910s, serious questions about quality of competition and the reliability of the sample need to be asked. Irvin Brooks’ peak is later, but in a very small number of games, and the majority of it comes from seasons before the ECL was organized. It might be worth noting that Brooks and Charlie Smith both had their monster peaks playing for the Brooklyn Royal Giants, so it’s possible there are park effects that the Seamheads system isn’t accounting for sufficiently, leading both players to appear a bit stronger than they actually were. Brooks’ prime isn’t of a level to make him a serious candidate, so I’m not sure that a peak rate based on a problematically small number of games is enough to change that. But he’s still a player turned up by this study about whom I knew nothing before I started, so I want to learn more about him and see if there’s more to his career and his playing quality than is revealed in the Seamheads data.

The players with peaks between Beckwith and Suttles--Strong, Wright, Dixon, Parnell, and Taylor—do not all call for extensive comment. Strong, Wright, and Dixon were part of the same cohort in the prime section. By prime, all of them seemed a bit short of being serious candidates because they probably lacked the needed career value given the quality of their primes; their similar placement in the peak rankings doesn’t change that. Strong and Wright have the possibility of longer, currently undocumented careers after their NeL careers ended: that’s what needs investigation in their cases. There are no indications of possible hidden value for Dixon.

In Roy Parnell’s case, on the other hand, a look at his five-year peak rate opens up a couple of lines of inquiry. In 12-year prime, he was a bit behind the Strong, Wright, Dixon group, but in peak he closes that gap, and the shape of his peak suggests that he might have significant hidden value. For one thing, his five-season peak is actually spread across nine years: he did not play in the NeL in 1929-31 or 1933. If one fills in those four years with a rate of production equal to his peak rate, that probably would raise his prime value significantly as well. It would be interesting to see how MLEs would project his missing seasons or what data exists for them (since several were in organized ball in southern Negro Leagues). If his prime rate were up over 7.0, he would have a much better case to be considered as a serious candidate. He is also a player for whom credit for years before the NeL might be considered. He played in his first NeL season at age 23, and, since his OPS+ in 92 games was 201(!), it doesn’t look like he struggled with the higher level of competition, suggesting he might have been an outstanding hitter prior to his NeL debut. If complete information were available for his full career, he might look a lot like Hurley McNair or Heavy Johnson, as Johnson is projected by MLEs.

In Ben Taylor’s case, although he appears to lose ground relative to McNair and Pettus because his peak is quite a bit lower than theirs, the fact that his peak can be taken pretty much at face value—due to the number of games it includes and the seasons in which it occurred—may still help his case. Taylor’s peak, at face value, is a bit higher than that of Mule Suttles, so he is ahead of Suttles by both peak and prime rates, and has a similarly long career. If Suttles is not a mistake selection, then the Seamheads data suggest Taylor has a sound case, as long as competition-level issues from the 1910s don’t reduce his value. Of course, a broader look at the unelected NeL players calls into question Taylor’s claim to be the best remaining unelected NeL position player, so the Seamheads rate study is by no means all good news for Ben Taylor.

To sum up this group quickly, Pettus, McNair, and Taylor remain strong candidates. Irvin Brooks and Roy Parnell could possibly be added to this group, depending on the significance of Brooks’ peak and how Parnell’s missing seasons are filled in or projected. Sam Jethroe, like Ted Strong and Wild Bill Wright, could possibly be a strong candidate if there is undocumented career value in his favor. If there isn’t, Jethroe is probably a close comp for Rap Dixon.

Part 6. 5-Season Peak WAR: 6-7.99 WAR

This is a much larger group, and one that I doubt is complete. Still, let’s take a look:

                  5-yr       Peak      Peak   Peak
HOM? Player       Peak Rate  Seasons   Age    Games
HOM  Suttles, Mule  7.93     1926-30   25-29  378
HOM  Oms, Alejandro 7.85     1921-25   25-29  225
     Lyons, Jimmie  7.79     1917-21   27-31  212
     Clarkson, Bus  7.76     1938-42   23-27  219
     Scales, George 7.75     1928-32   27-31  232
     Allen, Newt    7.71     1933-37   32-36  103
     Johnson, Heavy 7.66 1916,20, 22-4`21-29  279
     DeMoss, Bingo  7.31     1919-23   29-33  339
HOM  Santop, Louis  7.28     1914-18   25-29  187
HOM  Trouppe, Qu.   7.16     1942-46   29-33  251
     Robinson, Neil 7.10     1938-42   30-34  121
     Russ, Pythias  6.98     1925-29   21-25  332
     Pierce, Bill   6.97     1917-21   27-31   79
HOM Bell, Cool Papa 6.89 1935-7, 40-1  32-38  308
     Poles, Spotts. 6.72     1912-16   24-28  217
     Davis, Piper   6.72     1944-48   26-30  118
     Shively, Geo.  6.63     1912-16   19-23  153
     Hughes, Sammy  6.56     1935-39   24-28  153
HOM  Campanella, R. 6.56     1941-45   19-23  269
     Austin, Frank  6.51     1944-48   27-31  259
HOM  Mackey, Biz    6.48     1922-26   24-28  385
     Cannady, Rev   6.37   1922, 24-7  20-25  221
     Young, Tom     6.30     1928-32   25-29  216
     Patterson, Pat 6.12     1936-40   24-28  180

A few HoMers here, but not many. Suttles and Oms mark the general borderline at the top, the catchers are scattered around, with Roy Campanella joining the full-NeL-career catchers, and Cool Papa Bell. As a whole, this group calls for less discussion, but there are a couple of groups of players worth examining.

The first is the group of players in the 7-7.99 range.

Jimmy Lyons, at the top of this group, ranks 28th on five-year peak, as compared to a rank of 31 on 12-year prime. He doesn’t have much career value outside his prime. He’s similar to but not as good as Ted Strong and Rap Dixon, so he’s probably not a serious candidate.

In contrast, the cluster of infielders right below Lyons—Bus Clarkson, George Scales, Newt Allen, and Bingo DeMoss—includes at least one serious candidate. This group is also interesting because the three top career second basement in NeL history are all clustered together here. By peak rate, there isn’t a significant difference between Scales and Allen. DeMoss trails Allen and Scales by peak rate, just as he does by prime rate. When prime and career are considered as well as peak, Allen remains the top candidate of the group, very similar in the Seamheads data to both the elected Mule Suttles and the unelected Ben Taylor. (You may notice that I haven't called attention to the small number of games in Allen's peak as a concern. That's because his actual peak is probably earlier, from 1928-32. His rate for those five seasons is strong, 7.28, and in a robust 314 games. That's the period that is missing fieldoing data for three seasons, and I think the missing fielding value would bring his peak there above his 1933-37 rate. But by the numbers I have now, the 1933-37 peak rate is higher.) Bus Clarkson is an interesting addition to this group. He doesn’t have a 12-year prime rate because he only had six seasons in leagues tracked by Seamheads: 1938-40 NeL, 1940-41 MxL, 1942 & 1946 NeL; he missed 1943-45 for military service. After 1946, he returned to Mexico in 1947, played in the Provincial League in 1948, was ??? in 1949, and starred for Milwaukee in the American Association 1950-52 before getting a belated cup of coffee with the Braves during the 1952 season. Then he went down to the Texas League for 1953 and 1954 and to the PCL for 1955 before sliding out of organized ball in 1956, going from the PCL to the Texas League to the single-A Western League in the course of the season. That history would give him an MLE career of up to 15 years (1938-52) with a 12-year prime of 1938-49. But does that put him over Newt Allen? Clarkson’s peak includes one MxL season, so there are competition-level questions there as well as for the various stints that would constitute the back end of his prime. I think Clarkson still merits a look, but it’s also the case that the unelected (and, as far as I know, never supported) Sam Jethroe has a better NeL peak and a stronger record after the demise of the Negro Leagues.

Finally, there’s sluggers Heavy Johnson and Neil Robinson. Johnson is 32nd by 5-year peak, versus 23rd by 12-year prime. As with 12-year prime, the handling of his missing seasons and his fielding value would very significantly affect an assessment of his viability as a candidate. His Seamheads rate for 1922-23 (including Cuban play) is 8.70 WAR/162 g. If his 1919 and 1921 seasons are filled in at that level, his peak is very close to John Beckwith ‘s, which is at 8.83 WAR/162. If the projection is made more conservatively, he would be short of Beckwith. Robinson is 36th by peak rate, versus 27th by 12-year prime. In the prime analysis, Robinson was grouped with Irvin Brooks and Roy Parnell, but while both of them jumped into higher cohorts by peak rate, Robinson stays about the same or drops a little bit: his peak rate doesn’t add anything to his case.

If most of the 7.0-7.99 WAR/162 peak rate players aren’t serious candidates, none of the players with peak rates between 6.0 and 6.99 are going to be (except, of course, for the elected catchers). Most of these players were in the bottom tier of the prime rate rankings, so they don’t call for comment, but some are new and should be noted. Pythias Russ, at the top of the group, was a star shortstop in the late 1920s for six years, but he contracted tuberculosis and died at the age of 26 in 1930. He was an outstanding all-around athlete. His peak rate suggests he probably wouldn’t have had an HoM career if he had lived, but as he was only 25 in his last big-league season, he might not yet have reached his peak. Piper Davis was a star first baseman and manager for the Birmingham Black Barons who didn’t move into organized baseball after the end of the Negro Leagues. He was an important mentor to Willie Mays. Sammy Hughes was a second basemen with a strong reputation: he was a star player in the Negro Leagues of the late 1930s. If DRA found his defense to be strong, he might be close in value to the Allen, Scales, DeMoss trio, but his poor fielding values keeps his peak in the 6-7 WAR/162 range, and his prime rate doesn’t reach 5.0 WAR/162. George Shively was a star outfielder in the early teens, but he trailed off quickly, so his prime rate is below 5.0 also. Frank Austin was sort of Artie Wilson lite. Like Wilson, his NeL career lasted only from 1944 to 1948, and he followed it with a long, successful career in the Pacific Coast League. Unlike Wilson, he never appeared in the majors. He began in the NeL at age 27, but I don’t know what he was doing before that. Ted Young was a catcher for the Kansas City Monarchs with a nice peak but not much value outside of it. He played for quite a long time, but his prime rate was well below 5.0.

To sum up the 6.0-8.0 5-season peak rate list, there are three players here who show potential to be serious candidates:

• Newt Allen – a serious candidate
• Bus Clarkson -- might be a serious candidate, depending on the quality of his post-NeL Play
• Heavy Johnson – status depends on how fielding value and missing seasons are projected

That completes the 12-year prime and 5-season peak rate studies. There’s still one group of players left to look at in comparison to the prime and peak rate groups. These are star players, all from the 1940s, who played in the Negro Leagues less than five seasons. Seeing them in the context of the rest of the top position players from 1909-48 may help contextualize their Negro-League accomplishments.

One other possible plus for Pettus is that he probably has some level of major league caliber play from before any data exists since it appears he was catching in the southwest in his early 20s (and based on a reference in his wikipedia article, endorsed by a personal favorite in Babe Adams).

I think Taylor's amount of documentation probably is working against him - it seems unlikely he was substantially better than Pettus or McNair, but it seems more plausible that they have more hidden value. It's definitely enough that I'm comfortable continuing to pump the breaks on Taylor. He looks to have the least potential of the strong Negro League candidates to be substantially above the backlog.

Newt Allen and Heavy Johnson are still the two hitters who intigue me the most as possible ballot choices, but I'm not confident enough in either to pull the trigger yet. Both have pretty high ranges of plausible careers.

Should I open any new player threads (Hurley McNair and Sam Jethroe)?

Jethroe has a page here.

@Chris Cobb

Thanks for all the work. Amazing.


I had not really done all my homework last year on the new Heavy Johnson MLEs, and when he started picking up votes late in the voting process I was considering him for the first time really. There is just so little data, but the data that is there is so impressive, I was in a quandary. But I kept thinking about Levi Meyerle, and the we have passed on him.

With Heavy not having a strong Prime or Peak argument I don't see him on my ballot.

The point is the prime and peak argument would depend heavily on the missing data. If those years are as strong as his documented peak years, then he's viable. As dicsussed the MLEs from Dr. Chaleeko give very large credit for these missing seasons.

The negative impact of his fielding on his value is substantially reduced by the current process of norming NeL fielding values to Rfield, and the value of his peak and prime are increased by filling in his missing seasons playing for the Army Wreckers with seasons that approach his two top seasons in 1922 and 1923.

We are discussing alterations to the fielding value in the heavy Johnson thread currently (it probably won't be favorable to Johnson's MLE). So the open question is the seasons that are missing from Seamheads. If they are similar to his peak seasons, he is in fact a good candidate. If they fall moderately short of his huge peak seasons, then he's not.

one thing I have learned about our Hall of Merit voting, honestly, is how dull the voting is - relatively speaking - when you take out the obvious deserving candidates. so for "keep the public talking baseball" idea, this was a home run year again.

We at HOM previously had elected 2 thru 6 in the rankings of this latest 2022 HOF voting: Bonds Clemens Rolen Schilling Helton, as well as 8 AJones, 9 Sheffield, 11 Kent, 12 Manny

so our top HOF holdovers by their rankings pre-ballot were
7 Wagner
13 Vizquel
14 Sosa
15 Pettitte
17 Abreu
18 Buehrle
19 THunter

yikes.
this was a "pick 4 year," too, for HOM

absent PED issues, A-Rod was a unanimous selection of course.
after that, it was choosing 3 more between the above flawed candidates and two "rookies" of note (with 2022 HOF vote ranking): 1 Ortiz and 15 Rollins.

as you know, we collectively passed over Ortiz, Wagner and Vizquel as well as Rollins, settling for Sosa, Pettitte, and Abreu.

Papi only grabbed a shore of 9th with us along with BOBBY Bonds - bigger DH discount than I expected, plus he got postseason credit from most voters for kicking ass at it, but not "ended the curse," "rejuvenated the grieving city after a bombing" and so on.

Wagner, Vizquel, and THunter have never been serious HOM candidates, but Buerhle claimed 26th in HOM voting in 2022.

so if 2022 HOF voters had been stuck with this collection of players and PED was not a thing (thus removing the same candidates as HOM already had), A-Rod is in easily, Ortiz maybe some drama over whether he's first ballot, and... that's it.

one solitary player for people to debate - and with zero issue of that 2003 NYT Times story in our parallel universe, Ortiz probably would have been more expected to make it - and perhaps done better on the public ballots to minimize even that solitary drama.


HOMers have these "leftovers" next year, based on 2022 HOF voting rank:

1 Ortiz
7 Wagner
12 Vizquel
15 Rollins
18 Buehrle
19 THunter

again, we have no interest in Wagner, Vizquel, Rollins, or THunter.
that just leaves Ortiz (9th in HOM vote) and longshot Buehrle (26th) among HOF guys with 5 percent.

guh. how about some new blood, like:
Beltran - yes, we will love him and it's likely he finishes first. no one else will get a sniff out of the newcomers.
after that, it looks like Ortiz battling the guys who surrounded him this year:
5 Berkman, 6 Buddy Bell, 7 Munson, 8 Bando, 9T Bobby Bonds, 11-12-13 old-timers BTaylor, VWillis, and BJohnson, plus 14 Tommy John, 15 old-timer Schang, and 16 Tim Hudson.

with a year of politicking to go, it looks like any 2 of those 12 names above could get in with the "pick 3" year of 2023....

hopefully this is post 145 as I can't edit 144:

when I say "dull," I mean for an average baseball fan. minus PEDs, they'd know A-Rod is a lock and most would go for Ortiz as well. but there wouldn't be most of those other guys in the top half to debate about....

hopefully this is post 147 as I can't edit 145

fixed

#sigh

16 Tim Hudson.

with a year of politicking to go, it looks like any 2 of those 12 names above could get in with the "pick 3" year of 2023....

Lets right BBWAA's whiff and get Hudson in 2023?!

I'll add my thanks to Chris Cobb for the excellent deep dive into NeL candidates and the Seamheads data.

I noticed that Chris didn't mention Dihigo, which means he probably evaluated him as a pitcher. (On career WAR, slightly more than 50% of Dihigo's value -- 33.6 of 66.0 -- comes from pitching, so that's reasonable.) But while Dihigo had so many gaps in his record that he's tough to evaluate, I think during his prime years, which I will consider to be 1925 to 1936 (ages 20 to 31), he was more position player than pitcher. During most of that period, he was someone who would fill in as a pitcher when needed, but seldom one of his team's top starters. Only beginning in about 1935 did he emerge as an ace pitcher, and later, as his batting skills faded, he went on to have some outstanding seasons as a pitcher in Mexico in his mid to late 30s.

As Chris noted, the measure of WAR/162 is problematic for comparing pitchers with position players, so I will drop his pitching WAR and see how he compares just based on batting, fielding, and position adjustment. If my math is correct, he averages 7.75 WAR/162 over the 1925-36 seasons when excluding pitching WAR. (It would be 10.97 if I had included pitching WAR.) That suggests that there is a strong case that Dihigo qualifies for the HoM even without crediting him with his excellence as a pitcher.

Since I was doing this for Dihigo, I though I'd look at Rogan too. Rogan's career was different, in that he was primarily a pitching ace and secondarily a position player from 1916 through 1928 (ages 23 to 34). At age 35 and 36 he switched to full-time outfielder. After that, the record becomes skimpy.

As a pitcher, I would place his prime years as 1916-28, but as a batter, I will go with 1918-29 to get in his age 35 season, when he hit well but didn't pitch. Rogan's WAR/162 excluding pitching for 1918-29 was 6.28, which is quite a bit lower than Dihigo's.

I found this to be surprising, as Rogan's career OPS+ was 156, compared to 140 for Dihigo. So I had expected Rogan's non-pitching WAR to be ahead of Dihigo's. Part of the explanation is that while Dihigo averaged 4.2 PA/G during his prime, Rogan averaged only 3.5 PA/G. Rogan was used much more often as a pinch hitter, and presumably played more partial games than Dihigo. I tried switching the rate metric from WAR/162G to WAR/650PA, and that raised Rogan to 7.19 and lowered Dihigo to 7.34. Dihigo's remaining lead is due to stronger fielding numbers and positional adjustments, as he fielded well and played quite a bit as an infielder, including 178 G at 3B and 128 at SS.

A final comment. After reading Cobb's analysis suggesting that Oms may have been below the in/out line, I went back to review the Oms thread, and there's a lot of good stuff there, including quite a bit of information about seasons in Cuba and Venezuela that don't appear in the Seamheads database. (Seamheads has very good coverage of Cuba for the early years through 1923/24, but very little thereafter.) Oms always played better in Cuba than he did in the States, and one reason may have been that he played for the Cuban Stars that were strictly road teams and never had home field advantage. In contrast, Torriente and Dihigo played for America-based teams that didn't have that disadvantage. While I'll agree that Oms is probably a borderline HoMer, I'm not ready to concede that he was a mistake. There's more to his record than what is shown on Seamheads.

Brent, you are correct that Rogan and Dihigo are omitted because pitching is a very significant part of both of their values. It's hard to extract their batting/fielding value from their pitching value, and since they are both pretty clearly top players, it didn't seem worth the time to do the work to calculate position-player rates for them. They are included in my pitching set but not my hitting set at this point, as I think it's easier compare their value as pitchers to other pitchers. That approach will underrate them as players, but it will help to show how pitching worked in the Negro Leagues.

Thanks for the reminders about Oms' additional play in Cuba and the value it adds to his career! I hedged a bit in my comments about Oms, but probably not as much as I should have. The Seamheads data is a tremendous resource, but it's still far from a complete playing record for so many of the players who spent significant time in the Negro Leagues. I believe the Seamheads rates data can help us get a big-picture, comparative view of the NeL players of a kind that we haven't been able to obtain before, but for the actual ranking of any NeL player, a detailed examination of the player's full career is needed, including filling in as much of data missing from Seamheads as possible.

DL from MN wrote:

Should I open any new player threads (Hurley McNair and Sam Jethroe)?

I don't know if threads for a pair of players are still a thing, but it might be good to have a page for Hurley McNair and Bill Pettus, or else a page for each one of them. Alternatively, we could agree to use the Ben Taylor page as the space for McNair and Pettus, since Taylor is the player to which they'll need to be directly compared.

I think having a McNair/Pettus page is probably a good thing. At the very least, it would be good to have a hub to point to should either gets elected as they would be among the most anonymous HoMers.

Congratulations, Eric, on your ESPN interview!

https://www.espn.co.uk/mlb/story/_/id/33194902/top-100-mlb-players-all-oscar-charleston-best-baseball-player-all-why-important-try-find-out

FYI, of possible interest to folks here: Retrosheet released game logs and (mostly partial) box scores for the 1946, 1947, and 1948 Negro Leagues over the weekend.

This is my baby, so let me know if you have any feedback - comments, questions, suggestions, bad links, etc.

Congratulations again to Eric for the duiscussion of his methods in yesterday's "The Negro Leagues Are Major Leagues" panel.

{BTF is really being a PiTA and not letting me post in the Heavy Johnson thread. I will attempt to cross-post this there when the site lets me. But let's not let that stop me from sharing this.}

Finally getting around to the no-data season question. Sorry it took so long, but you are going to love why it took so long when I can pull the curtain back on it.

So, I'm looking at Heavy via the Marcel technique Brent suggested. There are two places in the process where this could be performed.
a) Before: I solve for the Marcel z-score of a hitter's NeL wOBA, and then make all the usual adjustments.
b) After: Just use the MLE Rbat outputs as the weighted Marcel inputs.

These come out with fairly different results. Here's how the calculations look.

BEFORE---solving for 1921 with actual z-scores
1922: 1.663 z-score, 263 PA
1923: 1.487 z-score, 489 PA
1924: 0.482 z-score, 302 PA
League average assumed to be 0 z-score

Weighted z-scores: ((1.663 *5)+(1.487*4)+(0.482*3))/12=1.313
Weighted PA: (263*5)+(489*4)+(302*3) = 4177
r: 4711 / (4177+1200) = 0.776
Regression: (1.313 * 0.776) + ((1-0.876)*0) = 1.019
Age adjustment: (0.991 * 1.150) = 1.001

When I run this through the sausage mill, I get 28 Rbat and, eventually, 4.4 WAR

AFTER---solving for 1921 with MLE Rbat/PA
1922: 77.5 Rbat, 590 PA
1923: 40.4 Rbat, 640 PA
1924: 65.3 Rbat, 600 PA
League average assumed to be 0 Rbat

Weighted Rbat: ((77.5*5)+(40.4*4)+(65.3*3))/12 = 61.7
Weighted PA: (590*5)+(640*4)+(600*3)=7310
r = 7310 / (7310+1200) = 0.859
Regression: (0.859 * 61.7) +((1-.0859)*0) = 53.0
Age adjustment: (0.991 * 53.0) = 52.5 Rbat

When I plug this into the existing MLE I get 6.8 WAR.

So as a reminder, this is what we're looking at:

BEFORE method
1921 (26): 28 Rbat, 4.4 WAR
1922 (27): 77 Rbat, 8.6 WAR
1923 (28): 40 Rbat, 5.1 WAR
1924 (29): 65 Rbat, 7.7 WAR
1925 (30): 3 Rbat 1.3 WAR
1926 (31): 38 Rbat (leads NL), 5.2 WAR
1927 (32): 48 Rbat, 6.2 WAR

AFTER method
1921 (26): 52 Rbat, 6.8 WAR
1922 (27): 77 Rbat, 8.6 WAR
1923 (28): 40 Rbat, 5.1 WAR
1924 (29): 65 Rbat, 7.7 WAR
1925 (30): 3 Rbat 1.3 WAR
1926 (31): 38 Rbat (leads NL), 5.2 WAR
1927 (32): 48 Rbat, 6.2 WAR

I greatly appreciate any feedback anyone wants to provide re which method makes more sense, Before or After. Thanks!

BTW: I relied on this post at Triples Alley. It's more specific than Tango's original instructions.

The after past the smell test, based on age and performance in the following years. But maybe just take the average of the two?

I just finished cataloging info about ballot history, and here's some (maybe) interesting tidbits or trivia:

- Only three players have ever been elected to HOM without receiving votes in every year they were eligible. All three were NgL players (Dobie Moore, Alejandro Oms, Dick Lundy). Man, Charley Jones and John McGraw came awfully close, though (I still don't know how Charley Jones of all candidates randomly became a HOMer given his vote history).
- We easily could have (should have?) about twice as many unanimous #1 electees as we do. There've been some very eccentric votes over the years.
- Hugh Duffy has received the most raw votes of all time (28,183) but last year received only a single vote.
- Were ties officially disallowed at some point? It was frequent for voters to let 3-4 guys share the last ballot spot or two for the first 30+ elections or so.
- Lance Berkman is the candidate appearing on the highest percentage of possible ballots (just under 44%) who is not yet inducted. Seems like a safe bet this year, IMO.
- Mickey Welch is the only player to have appeared on a ballot every year of voting, but like Duffy his support has gradually tapered and last year appeared on only two ballots.
- Vlad Guerrero is the only player to rank the same in two different years but get elected in one of them (quite trivial, I know). Alternatively, the only player who was eligible multiple years whose lowest vote rank and highest vote rank are the same.
- 61 players who finished 10th or worse in an election eventually got inducted. 16 who finished 30th or worse have been inducted.
- Candy Cummings has the lowest points share in HOM history of all players to receive a vote. He shared the final spot of a voter's ballot with 5 other players, giving him 1 total point, and he has been eligible since the very first year.
- 274 non-inducted players have received votes in history, very close to the 281 actually inducted.

Also, voting inertia (or stubbornness, depending on your generosity level) is very real!

Interesting stuff, thanks for summarizing

Top unelected players by MMP points per ballot (15 points is equivalent to unanimous MMP)

Mike Trout 111.07
Albert Pujols 105.53
Chase Utley 50.12
Robinson Cano 47.22
Miguel Cabrera 44.21
Mookie Betts 41.75
Clayton Kershaw 40.46
Josh Donaldson 40.06
Joey Votto 38.16
Jason Giambi 36.75
Andrew McCutchen 34.81
Dale Murphy 33.80
Justin Verlander 32.12
Albert Belle 31.29
Vic Willis 31.19
Ted Breitenstein 29.10
Jose Ramirez 29.04
Adrian Beltre 28.24
Jose Altuve 27.25
Joe Mauer 27.07
David Wright 27.00
Snuffy Stirnweiss 25.95
Zack Greinke 25.41
Dave Parker 24.71
Bucky Walters 24.43
Eddie Cicotte 24.29
Jose Bautista 24.15
Alex Bregman 24.13
Max Scherzer 23.53
Fred Lynn 22.58
Ryan Braun 22.55
Pedro Guerrero 22.20
Noodles Hahn 22.06
Carlos Beltran 21.77
Buster Posey 21.67
Bryce Harper 20.86
Chuck Klein 20.71
Vida Blue 20.62
George Foster 20.39
Giancarlo Stanton 20.38
Jose Canseco 20.12
John Olerud 20.00

MMP points for top returnees:

Lance Berkman 19.39
Buddy Bell 6.83
Thurman Munson 6.11
Sal Bando 8.44
Bobby Bonds 9.66
David Ortiz 18.28
Ben Taylor 1.88
Vic Willis 31.19
Bob Johnson 6.57
Tommy John 1.58

Andrew McCutchen 34.81
Dale Murphy 33.80

I hadn't thought of this before but those two careers are quite similar

The top 7 unelected by MMP points are locks, I think safe to say. Josh Donaldson may be an interesting case, so 40 MMP points looks like shorthand for a bright line delineation. Strength of the general player pool around you will make a difference. Beltre, Mauer, Scherzer, Verlander, Ramirez, Greinke, Beltran, and probably Votto are locks also, so career MMP points is a distinct thing from HOM candidacy (to state the obvious).

Interesting little cohort of pitchers there in Willis, Breitenstein, and Hahn. Breitenstein is the biggest standout relative to his complete lack of support for the HoM - does anyone have him even close to ballot territory?

You are right, Breitenstein has never received a HOM vote (this astonishes me). Belle has received votes in 9 years, topping out at #25 in his best year of voting. I thought those two would probably be similar in HOM voting but not really. My best guess is that Breitenstein looks poor by win shares, which is what most voters were using back when the field of candidates was small enough for someone like him to have been in main consideration. MMP votes were much more influenced by WAR, thus the discrepancy?

DL, if I interpret your numbers right, active players are likely to see their MMP points per ballot decrease. For example, once MMP voting happens for this year, and Donaldson gets no love, his total will drop below 40 as the average points per ballot will now include a "ballot" (a year) with few or no points gained. So per ballot you really mean per eligible year?

If you appear on a ballot you get points. If you don't appear you get 0 points for that year. 0/# of voters = 0. If everyone puts you in first place that year you get 15 points.

A technical question for the group:

Wouldn't OPS+ and ERA+ be more accurate if we subtracted a player's individual numbers from his league's numbers before calculating? Otherwise, the player himself influences the league average.

As an example, leaving aside park factors for the moment, the accepted MLB calculation for OPS+ is:

(player's OPS / league OPS) x 100

Wouldn't this be more accurate?

(player's OPS / (league OPS - player's OPS) ) x 100

Two things about this. First, that's not actually the OPS+ formula; the OPS+ formula is:

(PlOBP/LgOBP + PlSLG/LgSLG) x 100 - 100

I believe the rationale is that it scales more proportionately with run production. This is why some players (typically pitchers) will have a negative OPS+.

Second, I can see the argument for removing the player from league average, but that mostly seems like a lot of work for not a ton of difference. If you remove Ed Walsh in 1908 (464 innings at a 164 ERA+ in an 8-team league) from the league average, the league average changes by 1.7%, so Walsh's ERA+ would go from 164 to 166-167. And that's about the highest-volume, smallest-league case you can find outside of the 19th century guys. (If you want a more extreme ERA+ example, Bob Gibson's 1968 goes from 258 to 261-2; league ERA moves by 1.3%.)

OPS+ changes should be relatively similar; Babe Ruth in 1920 increases the AL average OBP by 2.5 points (0.7%) and league average SLG by a bit over 5 points (1.3%). I get something like a 4-point increase in OPS+ by removing him from the averages, and again, that's an all-time great having his best season in an 8-team league. (OPS+ uses non-pitcher averages and is park adjusted, I'm not going to mess with either of those things for a quick-and-dirty calculation.)

Short version, the change would be pretty minor and would significantly increase the computational load to produce something that's not exactly the most precise metric to begin with.

Eric,

Thanks very much for your reply.

Answering your two points:

1) Just FYI, I derived my definition of OPS+ (minus park effects) directly from MLB. That's not to say your definition isn't correct - just letting you know the source of mine.

2) Any concern about stat calculation "significantly increasing the computational load" is meaningless in the Excel era. It's just a matter of writing the formula and copying it. I agree that such a change would be a pain for a quick and-dirty calculation - but my purpose is manipulation of thousands of data points.

We live in an era in whch PITCHf/x and FIELDf/x tracks every play to the millimeter. I think we've moved past the time when any improvement to the accuracy of historical performance can be disregarded as "pretty minor."

I agree that OPS+ is "not exactly the most precise metric to begin with" - but I'm thinking that removing the individual player from the team component helps ameliorate that deficiency. Therefore, my question is not whether it's difficult to do, but whether it's beneficial to the accuracy of the metric.

P.S.: The same question (improvement, or not, of the accuracy of the metric) applies to ERA+ also.

I'm open to all opinions.

Practically speaking, the effect is not only tiny, but also pretty uniform. The best of the best seasons see the same small bump, and everything else is a rounding error. But there's no new or missing data that changes the evaluation. You'd just be slightly emphasizing big years a little bit more.

I can't think of a disadvantage to calculating OPS+ that way, but I can't think of any advantage either. There's no new signal and it's well within the margin of error.

The best of the best seasons see the same small bump... You'd just be slightly emphasizing big years a little bit more.

An interesting and valid point. A possible analogous question might be, do big years need the bump to accurately reflect their true impact vis-a-vis league context?

1) Just FYI, I derived my definition of OPS+ (minus park effects) directly from MLB. That's not to say your definition isn't correct - just letting you know the source of mine.

Note that the definition provided by MLB in that link says that a 150 OPS+ is "50 percent better than league average," not "50 percent higher." Cabrera's OPS in 2014 (the example cited in the link as a 150 OPS+) was .895; the AL average OPS was .706, and the Tigers played in a basically neutral park (101 park factor). 50% higher than .706 would be 1.059.

I think we've moved past the time when any improvement to the accuracy of historical performance can be disregarded as "pretty minor."

Regardless of what you do with OPS+, you'll never get the level of accuracy that you can with a relatively basic linear weights or BaseRuns estimation that can be applied to any era. (For instance, OPS+ inherently ignores base stealing, and inherently treats unintentional and intentional walks as identical events.) And given that, as 172 points out, the effects are going to be basically uniform (with some allowances for league size - the effect will be bigger in a smaller league, which is why I went to Ruth and Walsh for the sample computations), I think there are better places to look. (For instance, I would far prefer to see RA+ offered alongside ERA+ than to see minor changes in how ERA+ is calculated.)

It's kind of a philosophical question, but in reality it doesn't really matter (except for NgL stats, arguably). Taking an individual player out of the league will not materially affect the league OPS for the vast, vast majority of league seasons. It would also be a bit strange that Mike Trout is compared to a different league baseline value than, I dunno, Max Stassi is.

The disadvantages come in several forms. Imagine the calculation burden if you want to do OPS+ for any kind of splits. Or combinations of splits. At that point results get both computationally intensive and likely pretty wonky. Different players are compared to different baselines (not inherently wrong, but difficult to justify all this extra work and complexity of explanation for no new signal as others have pointed out). OPS+ no longer represents player performance relative to league average, because no player's OPS+ could be computed by comparing OPS to league OPS. How do park factors work? Do you calculate a park factor absent the player in question or with the player in question included? What if he's a pull-happy lefty in a stadium that plays very small (or large) for lefties or vice versa? Again, results get weird and reflective of something that's some degree of removal from reality. If I were Sean Foreman, I would not consider doing this.

Do we ignore this concept for defense? Defensive chances can already be quite small, especially in boundary zones. Is it right to say that the probability of a certain catch is 0% when evaluating player X because player X is the only player in the league to have made such a catch? Even though we know 0% is not right? How do you assign run value to a successful catch with 0% probability (I guess you'd give it 100% run value credit to the fielder, actually)?

If it were a league of four teams or what have you, then maybe. But I wouldn't introduce the complications for microscopic gains that are questionable to even call gains.

Exccellent and thoughful comments - many thanks to all who have responded.

Regardless of what you do with OPS+, you'll never get the level of accuracy that you can with a relatively basic linear weights or BaseRuns estimation that can be applied to any era. (For instance, OPS+ inherently ignores base stealing, and inherently treats unintentional and intentional walks as identical events.)

Agreed, agreed. However:

It's kind of a philosophical question, but in reality it doesn't really matter (except for NgL stats, arguably).

Aha! You have identified my purpose. A lot of data points are missing for NgL play - including, almost invariably, caught stealing and intentional walks. This necessitates a search for other comparative metrics that can be employed within a context of severe data-set limitations.

If it were a league of four teams or what have you, then maybe.

I agree that the size of the data set strongly influences the importance of the adjustment. Some of the seasons I'm working with run as low as 5 teams' worth of data - which is why I broached the question.

I would far prefer to see RA+ offered alongside ERA+ than to see minor changes in how ERA+ is calculated.

This is in fact something I'm doing. Numerous data points are often missing for NgL pitchers who pitched <45 innings in a season - including earned runs, which makes RA (and RA+) the only available point of comparison between all pitchers in a given league.

Different players are compared to different baselines... OPS+ no longer represents player performance relative to league average, because no player's OPS+ could be computed by comparing OPS to league OPS.

This is, I think, the best argument I've seen against subtracting individual players from the league component: it results in each player being measured against a different baseline.

progrockfan’s 2023 Preliminary Ballot

I’ve gone back to the drawing board for this year’s ballot, visiting not just the top returnees and new eligibles, but also dozens of players throughout history who intrigue me.

I do take cognizance of voters who encouraged me, during last year’s debate, to look more closely at modern players. At the same time, I feel obliged to follow the HoM’s mandate to consider players from all eras, and there are a few new “old guys” on my ballot. (And some more recent guys too.)

Frankly, though I didn’t vote for him, I’m relieved that Slammin’ Sammy was elected last year. I have a particular and strong dislike for clubhouse cancers, and were I a manager, I’d do anything I could to avoid having players like Sosa on my roster. I remain convinced that the man sowed dissension among his teammates, and that this was bound to have a negative impact on his teams’ play on the field. But I can’t prove it, or assign a numerical value to it – and this means, I think, that my decision to place him outside my top 15 last year was unconstitutional. The man’s in now, and I’m glad I’m not faced with the same choice this time around.

And so, my 2023 prelim.

Elect-me slots:

1. Luke Easter. My book Black Stats Matter has a chapter devoted to Easter’s Hall of Fame credentials, titled “The Toughest Case to Prove.” This chapter was motivated by a HoM voter challenging me to defend his case systematically. I won’t enumerate my arguments here, ‘cos I want you to buy my damn book – but I will say that I project Easter at 500-550 home runs with excellent plate discipline. While Easter’s exclusion to this point in time is fully understandable given the fragmentary nature of the proofs for his greatness, I consider him the #1 omission from the ranks of the HoM.

2. Ben Taylor. Best NgL glove ever at first base – and this in a ground ball-rich offensive environment. .339 career hitter with four titles in walks. By my analysis, an obvious elect-me player and serious HoM omission.

3. Bobby Bonds. The more I contemplate this guy, the more I think he might just deserve induction. His ten 20-20 HR/SB seasons have few historical parallels. I think he, and not Charlie Hustle, was the rightful 1973 NL MVP: he led the league in runs scored and total bases, hit 39 homers, stole 43 bases at a 72% clip, and led NL rightfielders in putouts, double plays, RF/G, RF/9 and TZR, essentially matching Rose’s performance in left. Only Bobby’s strikeouts, I think, kept him from the contemporary recognition that should rightfully have been his. If Taylor weren’t such an egregious omission, I’d rank Bobby #2.

The rest of the top 15:

4. Jim Kaat. Pettitte's election implies a standard for HoM starting pitchers well below the one I thought existed - and if you’re gonna vote for Pettitte, it’s hard to see how you can overlook Kaat, who to my mind has a much stronger case. He’s got a longer career, a little bit of black ink (Pettitte has none), and is a spectacular glove – I always thought Maddux was better, but Kaat has to be considered one of the best-fielding pitchers of all time. Kaat also has at least one season (1966) where he was arguably the best pitcher in his league; Pettitte has nothing like that. To my eye, Kaat is clearly qualified for the HoM. (How come a different Jim, and not this guy, gets to be called “Kaat-fish”?)

5. Tommy John. Another Pettitte-type innings eater, but lacking Kaat’s spectacular defense. A good solid postseason pitcher. As with Kaat, I rate John far above Pettitte.

6. Junior Gilliam. A true all-around player, master of multiple defensive positions, possibly the most valuable utility man ever (I'd love to see some discussion on this point) – better even, I think, than Gil McDougald. This is a guy I could see in the HoM.

7. Bucky Walters. A rare pure-peak vote from me. Not sure how I missed this guy before. In some respects an analogue to the now-enshrined Johan Santana; Santana has the superior peak, but Walters has more career – not a long career, mind, but damn! Two ERA titles, three 20-win seasons, three straight years leading in IP, three titles in H/9. Add to that the 1939 MVP and a blistering 1940 World Series, and we have ourselves a Pitcher. One SO title, but generally not a power hurler, which may hurt him in the eyes of some voters; but for me, as long as he got the job done – which he did – then he merits a high slot on my ballot.

8. David Ortiz. I was wrong about him last year, I think. I’ve long regarded Edgar Martinez as the best-hitting DH ever, but Ortiz’s post-season numbers are staggering, which really blurs the line for me. I hesitate to put a pure DH in my top ten – but MLB recognizes DH as a legitimate position, and so, therefore, must I.

9. Heavy Johnson. His rate stats are off the charts – but his career is so damn short! Also, he had only one truly great season, 1923, plus one other very high-quality season, 1922, and 100% of his black ink is compressed into those two years. It’s easy to see why a pure peak voter would give him an elect-me slot; I tend to lean more towards career (though there are plenty of peak guys on my ballot), and so, while he makes my ballot, he doesn’t merit an elect-me place.

10. Hugh Duffy. A trio of “Old Guys” here… The greatest defensive outfielder of his time; holds two presumably unbreakable hitting records, and had the finest single postseason of the 19th century in 1892.

11. George Van Haltren. Check out those runs scored totals in consecutive years – 126, 84 (in 92 games), 136, 115, 109, 113, 136, 119, 129, 118, 114. Add to that a .316 average on 2544 hits, a 122 OPS+, a bit of black ink, and a 40-31 pitching record, and you have a guy who will always make my ballot.

12. Ed Williamson. Repeatedly singled out by contemporaries as the great player in 19th century baseball. Not a dominant hitter, but clearly a fabulous defensive third baseman.

13. Bob Johnson. Still think he could’ve been a 3000-hit man with the right opportunities. I’ve downgraded him somewhat due to WWII-level competition over the latter portion of his career.

14. Monte Ward. A true two-way player as few men have ever been: 2107 hits and 164 wins. A decent second and third baseman with devastating basepath speed, and briefly a monster pitcher, one of the best in baseball in 1878 and probably the best in ‘79.

15. Kirby Puckett. Still sufficiently sold on his credentials to give him a lower-rung ballot place: five 200-hit seasons, five 420+-putout years in center, a rare right-handed batting title, spectacular postseason play. It’s interesting that Minnesota had two players as similar as Puckett and Tony Oliva for the whole of their MLB careers.

Mandatory disclosures and noteworthy players from my research:

Carlos Beltran. I’m deeply unimpressed by this year’s crop of new eligibles, but Beltran's good enough to warrant a comment. Excellent defense – but not historic, and zero black ink plus a very good but not great 119 OPS+ tells me that he’s simply not on the level of the enshrinees here.

Sal Bando and Buddy Bell. Combine Bando’s offense and Bell’s defense into one package and you get an elite, HoM-worthy third baseman. Unfortunately Nature doesn’t work like that, and separately they each fall short for me.

Lance Berkman. Good offense, short career, <2000 hits, never played a key defensive position. With the exception of Dick Allen, his comps are all HoVG-type guys.

Bob Caruthers. One Hell of a pitcher for a few years, but the timeline is not his friend.

Addie Joss. Dominant, but not Koufax dominant; his career is much too short, and that dominance is heavily tempered by the timeline.

Dolf Luque and Luis Tiant. Possibly worthy, but no room for them on this year’s ballot.

Rabbit Maranville and Bill Mazeroski. Just not enough offense to go with those superhuman gloves. As with Luque and Tiant, it wouldn’t offend me if they were eventually elected.

Tony Mullane. A long and interesting career, but not enough dominance for me vis-à-vis his era.

Thurman Munson. Give him four-five more seasons like 1975-76-77 and he makes my ballot. I strongly suspect a Big Apple-induced illusion with respect to his historical reputation.

Phil Rizzuto. I’ve been overrating the guy, I think, in past elections. Great defense at short, sure, but lots of guys can claim that, and his 1950 MVP probably should’ve gone to Larry Doby. <1600 hits.

Wally Schang. A very high OBP for a catcher, good longevity, good postseason play, but again a total lack of black ink. He was a good catcher, but I’m holding out for Mauer.

Vern Stephens. One of the great RBI men at third base. If he’d bounced his hits off anything but the Green Monster, his 1948-49-50 run would probably sway me to place him on ballot.

Vic Willis. Someone’s going to have to explain to me why he rates so highly with some voters. The low-offense context of the 1900s has to be figured into those nice ERAs.

14. Monte Ward. A true two-way player as few men have ever been: 2107 hits and 164 wins. A decent second and third baseman with devastating basepath speed, and briefly a monster pitcher, one of the best in baseball in 1878 and probably the best in ‘79.

Elected in the 1900 election

Luis Tiant. Possibly worthy

Elected in the 2020 election

(sigh)

Thanks for the catches, DL.

I missed Ward because he's listed as "John Ward" on the Plaque Room page rather than Monte or John Montgomery - I don't recall ever seeing him listed as just plain "John" in baseball literature. May I humbly suggest that this be changed? Tiant, by contrast, is just plain absent from the page.

I'll remove Ward from my top 15 but keep the same order otherwise, and bump Dolf Luque to #15 with an appropriate comment.

On a separate note: I can't remember how to insert tables into this site; the standard HTML <table> doesn't seem to work...? It would be nice to have a radial button for this purpose - but for now, can someone please just remind me of the appropriate code?

Bob Caruthers is also in the HOM. John Ward is how Ward is listed at baseball-reference, FWIW. The plaque room here at BBTF is incomplete for several reasons, but one of them apparently being page size limits that have cut off a number of inductees from the listings there (that is my understanding). DL linked my HOM site the other day in the MMP trophy room (thanks DL) which has all the inductees/info.

I like any ballot with high Tommy John rating. Curious why Kaat ahead considering by WAR they're not particularly close to each other in either system. Looks like John adds what you might consider some hidden value based on unearned runs allowed compared to Kaat, if WAR is not explicitly part of your evaluation (and apparently played in front of worse defenses, too). Things to consider.

I'm also glad we don't have to debate with you about Sammy Sosa's omission anymore :)

Are you aware of seamheads.com Negro Leagues Database? It has more complete NgL statistics than b-ref does. Actually I don't know where you're finding the .339 career BA for Taylor, b-ref has .337, Seamheads has .332.

If you downgrade Bob Johnson for WWII competition (rightly so IMO), are you aware you could also bump him up for minor league entrapment, if you will? He was in the minors long after it was clear he could have been a real contributor in MLB. Just another thing to consider.

Beltran is also the most efficient basestealer of all time and not on too low of volume. I have many things you might consider (if you haven't already considered them). Beltran's bat+baserunning were roughly equal to Jeter's bat+baserunning (better by rate but slightly less volume), still not impressive? Beltran wasn't a SS, but then Jeter was a pure butcher there while Beltran was an outstanding defender in CF.

You might also be surprised to learn that, despite having almost 2300 fewer PAs, Berkman has more career runs created from hitting + baserunning than David Ortiz does (422 to 408). Of course, while Berkman spent most of his career in the corner outfield/first base, he did also have a season in CF (and of course Ortiz was a pure DH). Berkman also had a .950 OPS in 224 postseason PAs, almost identical to Ortiz's postseason OPS, though Ortiz got 369 PAs.

Beltran is also the most efficient basestealer of all time

I was going to note this but I'm 10 minutes too late!

it's a factor in his evaluation

Carlos Beltran. I’m deeply unimpressed by this year’s crop of new eligibles, but Beltran's good enough to warrant a comment. Excellent defense – but not historic, and zero black ink plus a very good but not great 119 OPS+ tells me that he’s simply not on the level of the enshrinees here.

Black ink is (a) a mediocre-at-best measure, heavily favoring specialists over all-around talents (like Beltran), and (b) inherently biased against modern players who have bigger leagues to compete with. I would strongly encourage you to use something more comprehensive.

Also, if you're considering postseason for Ortiz, how about Beltran's 1.021 postseason OPS in 250 PA (plus 11 SB/0 CS) as a quality CF?

"Bob Caruthers is also in the HOM."

...That's just a pure swing and miss on my part. Thanks for the catch. (I adore awful puns.)

"Curious why Kaat ahead considering by WAR they're not particularly close to each other in either system. Looks like John adds what you might consider some hidden value based on unearned runs allowed compared to Kaat."

Both good points, and you may be able to persuade me to flip their ballot positions. WAR is one of many metrics I consider, but no single metric ever decides me one way or the other. For now, I give Kaat the edge because:

* I don't see John as ever having a truly great season like Kaat's 1966.

* Kaat has a lower career WHIP (a more important metric than ERA in my eyes).

* Kaat struck out more batters, and walked fewer.

* Kaat was a much better hitter.

* Kaat was one of the truly great defenders at pitcher.

(I know, I know - "defense is factored into his pitching stats already," I hear that all the time - but great defense is a major component of a complete player, and when other factors are equal, I tend to favor complete players.)

John has his advantages too, including 180 more innings, more career WAR (as you mentoned), fewer H/9, a lower career ERA, and a superior ERA+. I'm not saying the debate is cut and dried - there is after all a reason I listed them back-to-back - but for now my answer is 4. Kaat, 5. John.

"Are you aware of seamheads.com Negro Leagues Database? It has more complete NgL statistics than b-ref does. Actually I don't know where you're finding the .339 career BA for Taylor, b-ref has .337, Seamheads has .332."

I actually worked closely with key Seamheads personnel while writing Black Stats Matter. Kevin Johnson wrote an essay for the book and peer-reviewed the manuscript, and Gary Ashwill kindly granted me permission to run Seamheads numbers within its pages. I even found a correction for Seamheads during my research.

Bear in mind that my .339 includes pre-1920 NgL play. Also, it's crucial, when citing Seamheads, to separate inter-league play from all other forms - East-West Games, Negro World Series, winter ball, etc. The NgL regular-season-only summary in my book shows .339, so that's what I went with.

"If you downgrade Bob Johnson for WWII competition (rightly so IMO), are you aware you could also bump him up for minor league entrapment, if you will?"

Oh, absolutely. One of the reasons he makes my ballot in the first place is because he had extensive pre-and post-MLB careers in the minors - I mentioned that in last year's ballot, thought about doing so again this year, then decided it was repetitive. (Then decided it was repetitive.)

"Beltran is also the most efficient basestealer of all time and not on too low of volume."

This point I'll debate. Beltran's on 86.4% with 367 attempts, which obviously counts strongly in his favor - but Tim Raines is on 84.6% with 954 attempts, over 2 1/2 times more, and won four stolen base titles. To argue that Beltran is more efficient than Raines is a bit like arguing that Mickey Mantle (80.1% on 191 tries) is more efficient than Willie Mays (76.6% on 441 tries with 4 stolen base crowns) - I mean, it's technically correct, but I don't think it passes the smell test. Raines is a pretty clear choice as the second-greatest base stealer of all time next to Rickey!, while Beltran is nowhere in that discussion. (No disrespect to the man intended; just stating the fact.)

"Beltran's bat+baserunning were roughly equal to Jeter's bat+baserunning (better by rate but slightly less volume), still not impressive?"

Slightly less volume? 3465 hits from short, which has exactly one historical precedent (Honus), is orders of magnitude more impressive than 2725 hits from center, which has numerous historical analogues.

If, instead, you framed the argument as, "Carlos Beltran was a greater all-around player than Derek Jeter," I'd agree with you without hesitation - but the greatness of Jeter was never about him being a complete player; it was about 3400 hits from a friggin' shortstop.

I dunno, I may well be underrating Beltran, perhaps severely; he's clearly the rose in this year's bouquet. But he has his thorns too. The total lack of black ink really dings him in my evaluation; it marks him, to my eyes, as a compiler, and I've never been keen on compilers (e.g. Tony Perez - I'm not implying a direct comparison, just saying that he also has no black ink and, for that specific reason, also wouldn't make my ballot).

Keep talking though; my ballot is, after all, explicitly a prelim. I posted it early specifically for the debate, and I appreciate your comments.

"Black ink is (a) a mediocre-at-best measure, heavily favoring specialists over all-around talents (like Beltran), and (b) inherently biased against modern players who have bigger leagues to compete with. I would strongly encourage you to use something more comprehensive.

I can't agree that black ink is a mediocre measure. You can't possibly mean that leading a league in a major offensive category is uniformly meaningless, or that players who do so consistently don't tend to be great players.

I do agree, though, that it favors specialists over all-around talents, and I also agree that it's considerably harder for modern players to compile black ink.

I further acknowledge that there's such a thing as irrelevant black ink. I mean, who really cares how many times Ichiro led the AL in at bats? I have the highest regard for Ichiro, he'll probably be #1 on my ballot when he becomes eligible (I'm not sure who his competition will be, so I can't state this with certainty), but that particular black-ink stain is meaningless to me.

Still, for a player of Beltran's career length to have compiled none is, for me, a mark against.

"I would strongly encourage you to use something more comprehensive."

As I stated, I don't rely on any single measure. Super-adherents to single measures (OPS+, or WAR, or what-have-you) have always baffled me.

You have to be damn good to be allowed to compile 11000+ PA's in the Majors - only 39 men have done it, all All-Star caliber players at minimum. But to not earn any black ink in 11000+ plate appearances - I just don't see how that can be avoided as a mark against a player.

"If you're considering postseason for Ortiz, how about Beltran's 1.021 postseason OPS in 250 PA (plus 11 SB/0 CS) as a quality CF?"

Great point, Eric. Beltran inches closer to my ballot with each new comment.

I can't agree that black ink is a mediocre measure. You can't possibly mean that leading a league in a major offensive category is uniformly meaningless, or that players who do so consistently don't tend to be great players.

Of course leading the league isn't meaningless; the issue with black ink is that it considers everything EXCEPT leading the league meaningless.

For what it's worth, among stats listed by Baseball Reference, Beltran led the NL in oWAR in 2006 and championship probability added in 2008; he also led his various leagues in stolen base percentage three times. And if you want to credit him for second-place finishes to account for the fact that modern leagues are almost double the size of the originals, he has a few of those in more conventional stats: runs scored (2006/08), doubles (2011), triples (2001), and stolen bases (2002).

Slightly less volume? 3465 hits from short, which has exactly one historical precedent (Honus), is orders of magnitude more impressive than 2725 hits from center, which has numerous historical analogues.

11K plate appearances vs 12K plate appearances is slightly less volume (both huge). Hitting value is not very well captured by hits, just as baserunning value is only halfway captured by stolen bases (unless you're Rickey or Rock). Beltran had 1078 XBH to Jeter's 870, for example. Ultimately, Jeter's offense was worth 332 runs, Beltran 313, which is why I called them essentially equals offensively (one better by bulk, the other by rate, with close career totals). Jeter had far more hits than almost every other player in baseball history, but of course no one uses that to argue that Jeter is one of the best offensive players to ever play. Which is why I don't think it makes for a particularly good argument about the relative offensive quality of Jeter and Beltran.

You don't buy the Jeter comparison. I was just stating someone who is similar by the value of offensive production, not necessarily the shape of it. Maybe you'd be more receptive to a Dawson comp, which would be highly favorable to Beltran in almost all facets. How do you feel about Dawson as an inductee?

Another comp for Beltran worth considering - Dave Winfield. Both guys had long careers, all around skillsets (at least by repuation, Winfield's fielding numbers are pretty so-so at best). Winfield has 4 black ink, but Beltran is a single Joey Votto double in 2011 from having 3 Black Ink, so I think the comparison is fairly strong. Well actually, I think Beltran blows Winfield out of the water on defense, and every thing else is a bit of a wash. But if you buy into Winfield's defensive repuation over his stats, they are pretty similar players.

Ok, dumb question time - I'm seeing lots of references to Baseball Gauge ratings and Humphrey's DRA when their website got closed long before this discussion thread started. Where are people getting the data from? I keep hearing that B Ref may eventually incorporate some Gauge stuff into their website but have seen no evidence of this happening. I'm apparently one of the few people that preferred the Gauge to B Ref so I've been mourning its demise for almost 2 years now. So if anyone knows where Gauge data can be recovered I'm very interested. Thank you.

Ok, dumb question time - I'm seeing lots of references to Baseball Gauge ratings and Humphrey's DRA when their website got closed long before this discussion thread started. Where are people getting the data from? I keep hearing that B Ref may eventually incorporate some Gauge stuff into their website but have seen no evidence of this happening. I'm apparently one of the few people that preferred the Gauge to B Ref so I've been mourning its demise for almost 2 years now. So if anyone knows where Gauge data can be recovered I'm very interested. Thank you.

Regarding tables, using the "pre" and "/pre" tags work if you have pre-formatted tables that you want to post (keeping the formatting). I have found it looks better if you also bold the table entries using the "b" and "/b" tags (all tags going in between < > symbols).

For example, below I will precede the text lines with < b > < pre > and follow them with < / pre > < / b > where there are no spaces inside the < >.

text line one goes here        755
text line two would go here    600
text line three would go here   75
etc.

@Rob_Wood: Many thanks for the tables information. :)

@Jaack: Having seen him play more times than I can count, I've never agreed with Winfield's defensive metrics. The man's range, the sureness of his glove, and the power and accuracy of his arm were wonderful to watch. He's not the best I've seen, because he shares the planet with Ichiro, but he was very, very good.

That's subjective of course, and some might prefer to rely strictly on the metrics - but allow me to point out another historical player, George Sisler, whose metrics show him as being average, at best, at first base. All contemporary sources agree that he was either equal to or second only to Hal Chase as a fielder - and of course, Sisler was honest and Chase was not, so clearly the trophy goes to Sisler. In his case I trust contemporary opinion over the metrics. Winfield's too.

Having said this, your central point -

"Another comp for Beltran worth considering - Dave Winfield. Both guys had long careers, all around skillsets... They are pretty similar players."

- is spot on. I can't but agree 100%.

Guys, I think I may have exactly reversed two deserved ballot places. But before I commit to that decision, may I ask the group:

What's the group's consensus on Ortiz? This is a case where posting early is to my disadvantage, because I can't read other posters' current evaluations of him.

It's perhaps telling that with 541 homers, two postseason MVPs, and perhaps the greatest final season in baseball history, Ortiz didn't pass muster on last year's ballot.

Edgar M., who's in, isn't an adequate comp in my opinion, because

a) he was a toally different type of hitter, and

b) Edgar played more than twice as many games in the field (592 to Ortiz's 278), played a more important defensive position, and (I think we'd all agree) played better defense. Edgar had top-5 seasons in assists, double plays, RF/G, RF/9 and TZR. Ortiz never posted anything like that; as a first baseman, he was an excellent DH.

In your collective view, is my placement of Big Papi appropriate? He's the one guy I hemmed and hawed over the most. Last year I followed my instinct and left him off-ballot; this year I bowed to the big counting and postseason numbers, though not to the extent of giving him an elect-me place.

Here's my current thinking: The more I look at the arguments in this thread, the more I think that Beltran, not Ortiz, should be in my top ten, and that Ortiz, not Beltran, should be left off-ballot.

This is why I posted early. Posting here has sharpened my logic, my writing, and my knowledge of baseball history - and for that, I bow to all of you.

"Hitting value is not very well captured by hits, just as baserunning value is only halfway captured by stolen bases (unless you're Rickey or Rock). Beltran had 1078 XBH to Jeter's 870, for example. Ultimately, Jeter's offense was worth 332 runs, Beltran 313, which is why I called them essentially equals offensively (one better by bulk, the other by rate, with close career totals). Jeter had far more hits than almost every other player in baseball history, but of course no one uses that to argue that Jeter is one of the best offensive players to ever play. Which is why I don't think it makes for a particularly good argument about the relative offensive quality of Jeter and Beltran."

Perhaps I phrased myself poorly... What I mean to say is not that Jeter is one of the "best offensive players to ever play," but rather that he is (and I think this is indisputable) one of the best offensive shortstops to ever play. Beltran, by contrast, is not one of the best offensive center fielders to ever play, and again I think that's indisputable.

This, I think, makes them distinctly unequal as offensive forces - or, to put it another way, it would be much more difficult, historically, to replace Jeter's function in a given offense than it would be to replace Beltran's.

But of course, Beltran, with his well-rounded credentials, doesn't need that distinction to be elected to the HoM - and Jeter, who wasn't an all-around, five-tools player like Beltran, absolutely did.

I have Beltran nearly equal in value to Jim Edmonds and Edmonds had no trouble being elected. Duke Snider is another comp.

David Ortiz slots just below other elected sluggers I have placed lower than the consensus (Vlad Guerrero, Will Clark, Ducky Medwick) with my placement mainly because I have more pitchers represented in my PHoM than the actual HoM. I like him better than Helton, Kiner or Terry. He's not likely to make my ballot but he meets our established standards.

I had Andy Pettitte near the bottom of my ballot, so I consider his election somewhat questionable. But Pettitte has a good case and does pretty well in my Win Value stat where each start of a pitcher's career is evaluated compared to how a league average pitcher would likely have fared given the team's run support in the game. Pettitte is in the same tier of pitchers along with Tommy John in my pitcher stats (who also has appeared near the bottom of my previous ballots).

However, Jim Kaat is a big step down compared to Pettitte and John. In my evaluations Kaat is in a group along with Mickey Lolich, Frank Tanana, Kenny Rogers, and Jamie Moyer. There must be dozens of pitchers significantly better than this group who are still eligible for a HOM ballot. These better pitchers include Tim Hudson, Roy Oswalt, Kevin Appier, Tommy Bridges, Ron Guidry, Bucky Walters, Jimmy Key, Tommy John, Chuck Finley, Dwight Gooden, Vida Blue, Orel Hershiser, Frank Viola, Jerry Koosman, Steve Rogers, Jack Morris, John Candelaria, and Mark Langston.

And I don't think Kaat deserves any extra credit for arguably being the best AL pitcher in 1966. It is a close call and in any event he led the league with a fairly low "score". Again, many other pitchers can claim to have been the best pitcher in the league in a single season during their careers, and many others had far greater seasons that Kaat's 1966 season.

Pettitte is (IMO) a pretty standard inductee for relatively down-ish years of candidates. I also think Ortiz would not look out of place in the HOM, and though I don't have him close to my ballot (66th), it's not hard to see what is attractive about his candidacy to other voters, and even being 66th is really closer to the ballot than it sounds with how tightly packed players are after the top 5 or so. I agree pretty closely with Rob Wood about the cadre of pitchers that are a level or two above Kaat in my rankings (37th), but again I don't think Kaat would be a poor inductee in isolation, just that at this point there are numerous pitchers I perceive as better. Both Kaat and Ortiz are perfectly defensible choices for a ballot spot, IMO.

Beltran vs Ortiz? Yes, I feel safe in saying Beltran was a more valuable player in his career, and if I was choosing to put one or the other on my ballot it would be an easy choice for me. I have Ortiz behind other sluggers like Giambi, Olerud (is Olerud a slugger?), and McGriff. I may be lower on Ortiz than most, I don't know, but his support for HOM doesn't seem resounding so I don't guess that I'm that much of an outlier. I don't see anyone on your ballot who looks really indefensible - Gilliam is maybe kind of pushing it unless you have minor league credit to fill in for 1949-1952. I think your ballot skews far towards old-timers, but I know you have your defenses for that choice.

And I don't think Kaat deserves any extra credit for arguably being the best AL pitcher in 1966. It is a close call and in any event he led the league with a fairly low "score". Again, many other pitchers can claim to have been the best pitcher in the league in a single season during their careers, and many others had far greater seasons that Kaat's 1966 season.

The mid-'60s in the AL were a particularly weak period for great starting pitchers. I have been intermittently working on extending a modified version of Bill James's starting pitcher rankings (based on a rolling weighted average of game score with penalties for missing time) throughout modern baseball history (as far back as we have box scores); I may throw some results from this project into various threads in these parts once it's semi-complete. Anyway, one of the more striking results so far is that the NL holds the #1 spot for nearly nine consecutive years in this period, between a stint from Camilo Pascual in '63 to Gaylord Perry's first time at #1 in '72. Koufax, Drysdale, Marichal, Bunning, Gibson, Seaver, and Jenkins is a remarkable concentration of high-level peaks in the same league within a short timespan, and the AL basically has nothing to compete.

Gilliam has some similarities to Tony Phillips with the super-utility skills. I think the case that Gilliam is better than Phillips is pretty hard (Phillips was a substantially better hitter, and I don't see Gilliam's minor/negro league career making up for that) but if you love Phillips, Gilliam starts to look pretty good too. There's a lot of room for interpretation for Gilliam as a candidate as well, so I'm not a hard no. It's just a tough case to make, when Tony Phillips is sitting right there, and Ben Zobrist is coming up soon as well.

Flip page