Regular HHS contributor “no statistician but” (or nsb) continues his series examining where the Hall of Fame cutoff line really lies with his look at marginal Hall of Famers. In Part 3, nsb takes a look at the middle and left side of the infield, and 13 selected Hall of Fame inductees at those positions who are outside of the Hall of Stats. More after the jump
Where do you draw the line? I mean, YOU.
The next players for consideration among those forty-eight cited in earlier posts—who belong to no outlying group and appear in the Hall of Fame but not the Hall of Stats— are in a certain respect central to the issue of evaluation, and to my mind they present a much greater challenge for the evaluator than do catchers and first basemen.
These players man the other three other infield positions, in other words, and many of them have light bats and heavy gloves. As in the last post, I’m providing for comparison each player’s Hall of Stats rating followed by his career plate appearances, OPS+, WAR, and dWAR. In addition, I’ve appended as a “Small Hall” reference the name of and figures for the player at each position ranked sixth by JAWS.
Second Basemen
- Johnny Evers 88—7220 PA; 106 OPS+; 47.7 WAR; 15.4 dWAR
- Tony Lazzeri 88—7315 PA; 121 OPS+; 50.0 WAR; 5.2 dWAR
- Billy Herman 99—8636 PA; 112 OPS+; 54.8 WAR; 12.4 dWAR (WW II: 2 years)
- Bobby Doerr 98—8028 PA; 115 OPS+; 51.2 WAR; 13.4 dWAR (WWII: 1 Year)
- Red Schoendienst 66—8479 PA; 94 OPS+; 42.3 WAR 15.2 dWAR
- Nellie Fox 90—10351 PA; 93 OPS+; 49.0 WAR; 21.0 dWAR
- Bill Mazeroski 53—8379 PA; 84 OPS+; 36.5 WAR; 24.0 dWA
JAWS 6th: Rod Carew 158—10550 PA; 131 OPS+; 81.3 WAR; -1.7 dWAR
Third Baseman
- Pie Traynor 60—8297 PA; 107 OPS+; 36.3 WAR; 2.1 dWAR
- George Kell 66—7359 PA; 112OPS+; 37.4 WAR 1.6 dWAR
JAWS 6th: Chipper Jones170—10614 PA;141 OPS+; 85.2 WAR; -0.9 dWAR
Shortstops
- Rabbit Maranville 73—11255 PA; 82 OPS+; 42.9 WAR; 30.8 dWAR
- Joe Sewell 92—8333 PA; 108 OPS+; 53.7 WAR; 9.1 dWAR
- Phil Rizzuto 77—6719 PA; 93 OPS+; 40.8 WAR; 22.9 dWAR (WWII: 3 years)
- Luis Aparicio 94—11230 PA; 82 OPS+; 55.8 WAR; 31.8 dWAR
JAWS 6th: Arky Vaughan 151—7722 PA; 136 OPS+; 72.9 WAR; 12.0 dWAR
The challenge, as before, is to argue for (or against) the presence in the Hall of Fame of any or all of the thirteen players listed here with a HOS rating of below 100. One argument is disqualified: Saying that Richard Roe doesn’t belong because John Doe, who was better, has been passed over for inclusion fails on every count to meet the terms of the challenge. The merits or demerits of the listed players, comparisons to other HOFers, the use of more detailed statistics, historical and biographical information—these and similar bases for argumentation are all welcome. Later on I’ll weigh in with some observations about what I think might be at issue regarding some of the players.
There’s a huge amount to discuss here, not only because of the number of players at issue, but because we’re dealing with three different positions.
I’ve been reading Mike H’s book (and recent update) on players who meet the Hall criteria using his CAWS system, which calculates each candidate’s worth by weighting his ten best seasons (by Win Shares) at 80% and the remainder of his career at 20% (I hope I have that right!), one of many reasonable ways to approach the problem.
Mike’s lists provide a lot of context for where the Hall threshold lies. For position players, he calculates CAWS scores for the best 25 at each position (excluding pure 19th century players), and defines a CAWS Hall threshold for each position (the differences serve as defensive adjustments). He adds scores for all Hall members who fall short of the threshold, whether they are among the top 25 or not. This gives us a larger context of what norms for the Hall seem to be and where cut-off points lie, even if we choose not to make our decisions according to the CAWS system.
I thought it would be useful to supplement Mike’s information by applying some basic WAR-based figures within the rank framework of CAWS. That way we can both see where the players nsb has highlighted sit on (or off) Mike’s list, and look at each of them through a very different angle at the same time. The figures I’m going to use are basically the ones nsb has already noted, but, for me, the best starting point is to look at the rate at which candidates compiled WAR, which I take to be a quality indicator, so I’m going to stress that.
I don’t want to post too much material at the start, so I’m just going to start with second basemen for now — enough for starters since there are seven on nsb’s list (more than we considered in thinking about catchers and first basemen together). Here are the WAR based figures — in CAWS order. The WAR rate is per 500PA, and I’ve included a peak figure which indicates each player’s 7 most productive consecutive seasons (JAWS uses 7 best; CAWS 10 best, with out regard for continguity). The ‘Career’ category counts 5000 PA as 1.0. Since I like to start from WAR rate, I’ve added in parentheses the rank of each player on the list, taking that stats as the index.
WAR…WAR/500PA…Peak7…OPS+…..dWAR…Career
124.0………5.1……….62.9……141………..8.1……..2.4………Eddie Collins (4)
127.0………6.7……….64.4……175………13.9……..1.9………Rogers Hornsby (1)
100.6………4.4……….59.2……132………..3.8……..2.3………Joe Morgan (5)
107.4………5.1……….53.5……150………10.1……..2.1………Nap Lajoie (3)
..65.5………2.6……….41.7……112……….-2.9……..2.5………Craig Biggio (24)
..80.7………3.9……….48.2……124………10.7……..2.0………Charlie Gehringer (9)
..67.1………3.2……….35.5……116………..3.3……..2.1………Roberto Alomar (18)
..69.3………3.9……….47.2……126………..9.3……..1.8………Robinson Cano (10)x
..68.0………3.7……….38.6……114………13.5……..1.9………Ryne Sandberg (13)
..81.3………3.9……….49.2……131……….-1.7……..2.1………Rod Carew (11)
..70.4………3.5……….43.7……110………21.6……..2.0………Frankie Frisch (14)
..55.4………2.9……….34.3……123……….-0.1……..1.9………Jeff Kent (23)x
..71.1………4.3……….39.9……125………16.8……..1.6………Bobby Grich (7)x
..75.1………3.8……….32.5……117………16.3……..2.0………Lou Whitaker (12)x
..65.4………4.2……….49.3……117………18.3……..1.6………Chase Utley (8)x
..61.4………5.3……….51.1……132………10.1……..1.2………Jackie Robinson (2)
——————————————————————————————————
..49.0………2.4……….33.9……..93………21.0……..2.1………Nellie Fox (25)*
..54.8………3.2……….33.2……112………12.4……..1.7………Billy Herman (20)*
..45.1………3.1……….28.3……125……….-2.2……..1.5………Larry Doyle (21)x
..65.9………3.5……….30.9……104………20.2……..1.9………Willie Randolph (15)x
..51.2………3.2……….36.4……115………13.5……..1.6………Bobby Doerr (19)*
..47.4………3.3……….29.2……106………15.4……..1.4………Johnny Evers (17)*
..57.2………4.4……….41.6……120………22.4……..1.3………Joe Gordon (6)
..50.0………3.4……….34.8……121………..5.2……..1.5………Tony Lazzeri (16)*
..44.8………3.0……….37.9……106………..6.0……..1.5………Chuck Knoblauch (22)x
..42.3………2.3……….27.9……..94………15.2……..1.8………Red Schoendienst (26)*
..36.5………2.2……….22.9……..84………24.0……..1.7………Bill Mazeroski (27<)*
The line above Nellie Fox indicates where Mike draws the HOF line according to CAWS. Cano is in italics because he's still active (figures are through yesterday, 4/2/19). An X after the parenthetic rank for WAR rate indicates the player is not in the Hall, and * notes the seven Hall members whom nsb has asked us to consider. Mazeroski is a special case as he is not even among the 25 best second basemen, according to CAWS — I believe Schoendienst has also been bumped down below #25 with Mike's latest update, but he may still be next in line; Maz is separated by a distance (hence his WAR rate rank mark). Jackie Robinson's CAWS figure is calculated differently because of his short career, but he is still almost equal to Fox, whom he outranks after adjustment (if I'm reading Mike's book correctly). Joe Gordon lost his age 29-30 seasons to the War.
There is a link to the CAWS document under Statistics at the top of the page.
There are some differences between Mike’s lists in the update posted under Statistics and those in his book, apart from the fact that the update carries the 2011 figures through to 2018. In the book, Mike considers players whose careers predated the lively ball era, while the update lists are restricted to those playing from 1920 on. So second basemen like Nap Lajoie and Eddie Collins don’t appear on the update. I’ve picked the more inclusive list and supplemented from the update.
Since the update lists are shorter, there may be additions that should be inserted in the lower halves of the list, but I think the stats here should provide the necessary context for considering our candidates for Hall expulsion in terms of current thresholds. (And I should add that in his discussions, although Mike draws hard cutoff lines on the lists, he doesn’t actually dismiss borderline cases like, say, Fox, from Hall consideration — if you look at the update, the lists stops at the cutoff CAWS cutoff number of 260 for second basemen, with Robinson a rare exceptional case at 257; Fox’s number is 258 . . . sort of within margin of error.)
Just want to throw something out for discussion that we talked about a while back, which I think is highly relevant to this particular group—the reliability of fielding stats, particularly those of earlier-in-the-20th Century players. Context was different–equipment was far worse, field conditions spottier, hitters put the ball in play more. I wonder even if scorers were seeing the game the way modern scorers would.
Mike (L — we’re in danger of confusion on the Mike Front): I think this is a really interesting question, and it has three components: the stats collected; the conditions of play; the judgment of scorers — all considered over time.
As I understand it, the basic fielding stats collected remained pretty stable from early baseball through the 20th century: total chances (assists + putouts) and errors. I believe the quality of collection of these stats was pretty high from the start. Fielding performance was initially seen as equal in importance to hitting, with pitching less so early box score player lines included both hitting and fielding stats. Those counting stats are the basis for the Range (/G and /9 inn.) rate stats, which now serve as quality stats when compared to league averages. However, I believe Retrosheet doesn’t have complete fielding records until 1957, so although there’s lot of fielding data, it’s not truly complete (like base running data).
As for equipment changes, improved fielding conditions, and scorer judgment, I think that over the long arc of the sportswriter-as-official-scorer era, which was about a century, the standards for guiding judgments was fairly loose, and the judgments themselves tended to evolve as expectations for fielders rose with the continual development of gloves and more uniform requirements for field conditions. So although there may have been wider ranges of scorer judgments concerning hit/error decisions early on, and the nature of those judgments tended to change over time, those changes probably reflected reasonable expectations about what field should and and should not be held accountable for as fielding conditions improved. (I believe that over the past 20-40 years, scorer training and review have become far more rigorous.) To the degree that’s so, changes in equipment would have relatively little impact on the quality of fielding-percentage figures (the figures themselves rise continually because of improved fielding conditions, performance innovations, and training, but not because of changes in judgments about hits and errors). Since WAR-based figures are all normalized against replacement value each year, the rise in replacement fielding expectations would have kept pace with rises in performance, and the normed dWAR figures would retain their basic function. (Bill James, by the way, has written extensively about how conceptually awful the category of “error” is, but acknowledges that the awfulness decreases the farther back you go: it was an appropriate concept in the era of barehanded, high-bobble ball, for which it was designed.)
However, it seems to me that in this century advanced fielding stats are creating a real rupture with past figures. Rdrs measures much more than outcomes and errors: it measures every play against fielding quality norms for specific types of challenges. Fielders are increasingly assessed on the basis of how they perform against degree of difficulty, not against outcomes. We have the tools to do this for hitting too — Statcast measures exit velocities and angles that are much better diagnostics for hit quality than whether or not a hit, out, or error results, but although I’ve read reports that teams are now assessing players primarily according to those stats rather than the ones B-R, etc., report (a lot of this data is available on MLB.com and other sites, but I haven’t actually explored using it). Moreover, those stats can be related to the pitch speed/movement data Statcast collects to create degrees of difficult on each “hitting play,” with the pitcher’s execution being measurable against norms (which may or may not be calculated by teams) of optimal pitch approaches related to batter/situation types. So at this point, when looking at B-R as the master database for assessments, I suspect that contemporary fielding stats are, in fact, more objective and probing than hitting and pitching stats, which simply measure pitch/contact outcomes.
That’s very good, Bob. I want to think about how that would impact comparisons between players essentially from different eras. I suppose if the backbone of the comparison were the counting stats (maybe the old fashioned one) you’d get a reasonably good metric. I happened to look at Maranville’s 1914 Boston Braves–of the 548 runs they gave up, 105 were unearned. Other NL teams: Giants 576/121, Dodgers 618/190, Cardinals, 540/163, Pirates 541/119, Cubs 648/230!, Phillies 676/206, Reds 653/200.
It would take a while to get a fix on how those unearned run rates relate to defensive quality. UER is one indicator, but pitchers play a role in that, whereas fielders are presumably fully responsible for their errors, and in team terms, errors as a percentage of chances. More to the point, the number of chances relate to high/low range factor, which should correlate in some way with Defensive Efficiency, or percent of BiP translating to outs.
The 1914 Cubs gave up a lot of UER (though ten fewer than you’ve noted here), and also committed a lot of errors. However, they were second in DefEff, ahead of Boston, though it’s very hard for me to figure out why. The pitching record was unusual: lots of Ks, lots and lots of BBs, very few hits, lots of errors.
Thanks to Bob for posting the CAWS scores for second basemen – I appreciate that.
Jackie Robinson was mentioned as falling just below the CAWS HOF benchmark for second basemen of CAWS = 260. This is true but Jackie meets a different CAWS HOF benchmark (“a short but great career”). In my research, I have discovered that since 1900 there appear to have been only 12 position players who have played in fewer than 1800 games but achieved a CAWS score of 250 – and ALL ARE IN THE HOF (except Dick Allen). (Thanks to Paul for pointing out that Allen belongs in this group.)
……………………………..Games….CWS…..…CV…….CAWS
Joe DiMaggio………….1736…….387……..325……..341
Dick Allen…………… …1749…….342……..304……..314
Elmer Flick……………..1483…….291……..280……..283
Earl Averill………………1668…….280……..268……..271
Hank Greenberg….…1394…….267……..262……..263
Lou Boudreau…………1646…….277……..255……..261
Bill Terry…………….…..1721…….278……..255……..261
Larry Doby……………..1533…….268……..257……..260
Jackie Robinson………1382…….257……..257……..257
Mickey Cochrane….…1482…….275…….250……..256
Kirby Puckett…………..1783…….281…….247……..256
Bill Dickey…………….…1789…….314…….235……..255
Joe DiMaggio and Dick Allen stand out among the players in this group as the ones who achieved the most in a relatively short career. But note that Jackie Robinson played the fewest games among this elite group – and yet he was still able to achieve the CAWS benchmark.
As an interesting aside, note how close the numbers place Jackie Robinson and Larry Doby – the two players credited with integrating the National League and the American League, respectively. Each of these players had a core value (CV) of 257 meaning that each averaged almost 26 win shares over his ten best seasons – an outstanding accomplishment. So, aside from being the integration pioneers, both Jackie Robinson and Larry Doby were terrific ballplayers.
The short-career benchmark is an interesting feature that gives the CAWS system valuable flexibility. It’s interesting to compare one of the second basemen on our list, Joe Gordon, with an outfielder on the “short career” list: Earl Averill.
Averill played 1668 games, while Gordon, who lost two years to the War, played 1566, so both qualify for “short career” consideration. Averill scores 271 in the CAWS system, as Mike indicates. Gordon, by my calculation (using Baseball Gauge stats), scores only 235, based on a 241 CWS and 232 CV. So Win Shares-based CAWS puts Averill over the CAWS 270 threshold for center fielders, while Gordon is 25 under the CAWS 260 threshold for second basemen.
WAR-based figures give a different picture:
WAR…WAR/500PA…Peak7…OPS+…..dWAR…Career
..57.2………4.4…………41.6……120………22.4……..1.3………Joe Gordon (CAWS 241; Best10 232)
..48.0………3.3………..35.4……133……….-5.3……..1.4………Earl Averill (CAWS 271; Best10 268)
Basically, it seems to me that key difference between Win Shares and WAR here concerns defense. Gordon was an outstanding defensive second baseman, according to WAR, No second baseman exceeds him in terms of his rate of dWAR accumulation (and on our second baseman list, only Maz edges him out for total dWAR, despite Maz’s career being about a third longer). Averill is assessed as a poor fielder in a less demanding position: the positional adjustments for second base and center field range from 0 to +1.0 for 2B in Gordon’s time (1938-50), and from -1.5 to -3.0 during Averill’s career (1929-41).
Win Shares assigns to Averill a 162-game average of 5.4 defensive Win Shares while Gordon earns 7.2. So Gordon is viewed by Win Shares as only 33% more productive than Averill on defense. Win Shares doesn’t use negative numbers, so that sort of calculation is possible, where it is not in dWAR (in terms of Rfield and Pos, B-R assigns Averill -32/-27, Gordon 150/66).
In terms of offense, as OPS+ figures suggest (with Averill having about a 10% advantage), Averill exceeds Gordon in oWAR, 51.1 to 41.5; Win Shares makes the difference in Offense 215.1 (21.6/162G) to 172.1 (17.8). Both give Averill about a 25% advantage, which makes sense since OPS+ is relative to average, not replacement. So I think Win Shares and WAR align pretty well on offensive measures. The defensive measures, however, seem to me world’s apart.
This may by why CAWS sees Averill as Hallworthy, while nsb, looking at B-R, puts him on the questionable list, and why CAWS sees Gordon as lacking Hall credentials, while nsb didn’t put him on the list of questionable second basemen.
For the record, Joe Gordon is not profiled here because he makes the grade in the Hall of Stats, and by a comfortable margin with a 118 rating. That is likely due to Gordon’s impressive 37.2 WAA, far ahead of Averill at 22.8 and only an 88 Hall of Stats rating.
Right. I completely forgot for a moment how nsb generated his lists. I should have named Adam Darowski.
This describes one of several reasons why win-shares any anything based on it should be tossed.
mosc, I wouldn’t toss any reasonable system, including traditional stats and, of course, Win Shares. They provide data from different perspectives and, as long as we keep alert to systematic biases, can help stats improve.
I do feel Win Shares dramatically under-represents the impact of fielding, but Mike, who is based in the WS system, has several times pointed out the from his perspective WAR dramatically overstates fielding. As long as we’re grounded in one or the other we can’t really assess what the dispute is about and the degree to which each system may capture something the other misses. (In the case of fielding, as I’ve said too often, I think that since adopting Rdrs B-R is very much on the right track, but I can’t say the same for earlier eras, most of baseball history.)
Win Shares and WAR are asking very different questions: 1) how great a share of how many real-world wins did a player earn on the teams he played for; 2) how many statistically-based wins did the player earn by comparison with others on an annually set scale. I think both those questions are good to ask and try to answer. I think the second relates more closely to the issues the HHS community cares about, while the first probably relates more closely to what Hall of Fame votes are about.
I would have expected you to be a little more sympathetic towards WS on other grounds: the fact that it resembles WAA+ in essentially writing off a player’s negative impact on his team’s record and assessing his quality on the basis of his demonstrated positive talents.
Well I agree with you Bob. I’d love to write down some of the statistical strength and weaknesses with the WAR system. It’s just too much time to put it all down and make sure it’s coherent.
In summary, WAA is good at normalizing the sample size except for individual factors. It’s bad when the individual components are correlated. That can be extremely hard to remove. Good pitchers don’t want to go to offensive parks and good offensive players don’t want to go to pitching parks. A pitcher’s pitches are not equally easy or hard to field given the same trajectory off the bat nor are park factors uniform on their impact to all balls in play, etc etc.
Another fundamental problem with WAA is WAR. People do not like WAA because a damn good player gets a 0.0. That guy is average. He might have played 1 game or 1000 games to get that total. Due to accumulation of positive against negative (good players play longer), the overwhelming majority of major leagues have a negative WAA but the negative gets construed as “hurting the team” which is not the statistical design. Replacement level is a construct that works somewhat well to show the accumulation of mediocre value (not intended nearly as derogatory as it sounds) which sets a somewhat looser bar to target contributions above a AAA level of play. The uniform addition based on durability reflects better the value to the team created but only to team who’s only other option is AAA. In reality, bad teams may view AAA as nearly status quo and good teams could see a positive WAR player as a tremendous subtraction relative to the rest of the team.
But I think knowing the problems with the current system does not mean ignoring the statistical data set that allows for several independent factors to be isolated and measured in relative importance.
I really want to write more on this but I’ll have to stop there.
It’s been pretty quiet on this thread. Here are some thoughts on second basemen; if they’re outrageous enough, perhaps it will liven things up.
Of the seven second basemen we’re considering, two stand out for fielding superiority: Maz and Fox. There’s a fairly clear cut among Hall-eligible members of this list between five dWAR leaders at 20-24 dWAR, and everyone else, at 16.3 and below (Utley’s not yet eligible). Like SS, second is a position of enough defensive importance that teams make trade-offs of offense for defense, and among the five defensive leaders here, Maz and Fox are the two that best exemplify that.
Maz was a truly weak hitter, compensating with truly great defense. I don’t think that translates to a Hall case: it translates better to a solid career as a regular despite poor hitting (but I was jumping up and down with his famous home run). Fox is closer to the borderline in terms of both the Hall of Stats and CAWS, and he has an added consideration: as the second-slot hitter behind the leading base-stealer of his time on a low-scoring team, Fox often had to prioritize advancing the runner over getting on base, which is reflected in his high Sacrifice Hit totals, but would have included also many grounders to the right side that might have been more promising hit chances if Fox had simply been swinging away. Nevertheless, I’ve never felt either Maz or Fox belonged in the Hall, and I still feel they fall short. Fox may be borderline for CAWS, but I don’t actually see his defense as of a scale to bring him up that high.
The remaining five we’re considering really do not seem to me to have very strong cases either. Obviously, Herman and Doerr are borderline for both CAWS and HoS, but I don’t really see it. We have a list of 27 players here: when it comes to the category I prefer to start with, WAR/500PA, Doerr and Herman rank 19th and 20th. If you use the WAA+ figure that mosc prefers, they rank 19th and 20th. On defense they do better: 13th and 14th (same order), but not to a degree that particularly compensates for moderate WAR/WAA rankings. Both did lose some time to the War, and that adds strength to their cases — pushes them back towards the borderline for me. The rap on Doerr, as mentioned in an earlier thread, is that he was a one-park wonder — in Fenway he was a fine hitter; everywhere else he was mediocre or worse, relative to his own career standard (he actually hit at his own career average rate in Sportsmans Park, but nowhere else). I don’t see any problem with that in terms of productivity: it evens out, and it’s a plus if a player learns to make the best of the park where he plays fully half of his games. But if part of the qualification for the Hall concerns athletic excellence — assessment against the five-tool ideal — then Doerr suffers a little: could he have translated his skills into productivity as well had he wound up with a team other than the Red Sox in quirky Fenway Park? Herman is just behind Doerr in a number of key categories (though not by CAWS), but lost more time to the War. They seem even to me — perhaps it’s my small Hall bias, but I don’t feel the Hall is stronger by including them and I would not.
That leaves Schoendienst, Lazzeri, and Evers. I see no good case for Red, much as I liked him as a player in his later years. He was solid on defense without giving too much away in hitting, and he lost a late-career year to an illness that had probably handicapped him over the prior season, but the gap is too far to bridge. He places just above Maz in almost every key category, but without Maz’s exceptional defensive record. Schoendienst was a good baseball man, very likable, and a successful, World Series-winning manager, and I expect that accounts for how he crossed the Hall threshold in ’89. Lazzeri’s case probably relied on his association with the great Yankee teams of the Ruth era, and he drew attention as a player who was successful despite dealing with epilepsy. He was a good hitter without giving too much away on defense, and his WAR rate stat is not near the bottom of the second baseman list above (though he’s well below the median), but he ranks 24th of 27 in terms of WAA+, just above Larry Doyle. That’s not my favorite stat, but it does suggest that Lazzeri’s case isn’t very strong.
Evers is a tough call for me. He was such a pronounced personality in early 20th century baseball, and was not only a member of utterly exceptional teams (the 1906-8 Cubs and the Miracle Braves) but their spark plug (in several senses, not all of them good — Giant fans were certainly not enamored of him). He was strong defensively, though not near the top of this list, and didn’t give too much away in hitting, which is impressive, given his very small size. His WAR rate is just below Lazzeri’s (good among all players, but not strong for a Hall candidate), and he’s a little higher in WAA+, but still near the bottom of this list (tied with Fox). He’s got such a terrific narrative that I’m actually happy he’s in the Hall, but I would not have voted for him if I’d been on the Tinker-to-Evers-to-Chance committee.
So in this case I’m 0 for 7 in my vote. Disappointing. I went into this hoping to defend some player as I did Chance last round: it’s fun to do. Perhaps someone else can take on that role in disputing my assessments here.
Tangentially, I think the most surprising element of the stats I posted is how poorly Biggio does — really poorly: his WAR rate is well behind Chuck Knoblauch’s. It doesn’t get better with WWA+ either (even though the ‘+’ adds back a lot of WAA Biggio gave up in his late career pursuit of 3000 hits). The Hall of Stats likes him (127), and so do CAWS and JAWS, mostly because of his best season totals — the ones that induced Bill James to herald Biggio as the great successor to sliced bread (a judgment he later toned down, if I recall). (I assume the HoS figure is heavily influenced by eliminating negative WAA and adding a partial catcher bonus for Biggio, who was really not a very good catcher.)
But by my reckoning here, Biggio is really not Hallworthy. A lot of the problem is defense, but his OPS+ rating is not very impressive either, especially if it’s compensating for poor defense. His great strength, though, is his peak, not just “best seasons,” but the terrific consecutive run from 1995 through 1998, with fine seasons surrounding those great ones — those seasons were why James rated Biggio over Lajoie in his “New Abstract.” And perhaps Biggio is a case where we really need to pay less attention to the career record than to individual accomplishments (like all those doubles in the Astrodome).
Bob,
Here’s age 27-32, 1893-2018, OPS+, Middle Infielders, 2000+ PA’s
1 Rogers Hornsby 187 1923 1928
2 Nap Lajoie…… 169 1902 1907
3 Honus Wagner 168 1901 1906
4 Joe Morgan…. 155 1971 1976
5 Eddie Collins.. 144 1914 1919
6 Jackie Robinson 141 1948 1951
7 Robinson Cano 138 2010 2015
8 Craig Biggio… 135 1993 1998
Biggs is in pretty fair company here stretching it 2 years back to 1993. I believe James indicated he was better than Griffey. But, even WAR gives him credit in 1997 (9.4 – second only to Walker among all position players in MLB). His 35.3 WS are 3rd in the NL to Bonds and Piazza. Walker is 2nd among NL RFers (Gwynn). Biggio went 15/32 on stolen bases in 1993 and followed it up the next season with 39 steals out of 43 attempts.
He took a walk, stole a base, got hit by pitches at a ridiculous rate (contraption on left arm) and played on pretty good teams. That certainly helped sway voters along with 3,000 hits.
When we ask the question, “what were the voters thinking”, the answer may lie in the fact that they didn’t have a chance to compare Billy Herman to Ryne Sandberg, Robinson Cano,……or Chase Utley or whomever has been great since Herman’s 1975 induction.
On another note, I would think that Doerr scores above 100 if he doesn’t hurt his back at a relatively young age (33)
RE: Knoblauch and Biggio
They were only three years apart in age (actually 2.5, but 3 in “baseball ages”). Both players tooled around at more important defensive positions in their youth (Knoblauch at short, Biggio as a full-time catcher) before settling in at second, then moving to the outfield as older players (Knoblauch due to the yips); both had a fair amount of positional flexibility, let’s say. Both are small players – under 6′, not too brawny. So the comparison is good.
But the apples-to-apples comparison quickly becomes apples-to-crabapples when you include a VERY long decline phase for one player, while the other played himself out of baseball at an earlier age. So let’s see whether Biggio actually accumulated WAR at a lower rate. Knoblauch played from age 22-33. Fortunately, Biggio was also in the league all of those ages, though Biggio played less at 22. Anyway, they have a different number of PA by less than 100, which is essentially equal. Both players have, via WAR, one absolutely outstanding season (9.4 for Biggio in 1997, 8.7 WAR for Knoblauch in 1996).
From age 22-33, Knoblauch accumulated 44.8 WAR; Biggio, 56.2. From ages 34-41, Biggio accumulated 9.3 WAR; Knoblauch 0.0. This is the problem with looking at WAR as a “rate” stat; it doesn’t scale well, because, in comparing players to replacement, you’ve already factored in the rate! That’s why we do it that way. And it’s why Win Shares doesn’t work; it gives positive credit for negative performance. WAR is giving positive credit for positive performance, and zero credit for zero performance. You can’t claim that Knoblauch’s teams won more games faster because of Knoblauch not having an extended decline phase. They were closest in WAR right after age-31; at that point, Biggio leads the comparison 44.6-44.5, in spite of a 500 PA disadvantage. From that point on, Biggio was a vastly better player. Knoblauch was the better 20-something player. But as a whole, his career did NOT accumulate WAR faster.
I understand the objection you’ll make, Bob: you’ll say, as you have before, “You can’t have different WAR rates when you compare to different players.” What I’m saying is that dividing WAR by PAs that way is absolutely misusing the statistic, because you’re going to get anomalous results. It’s always going to be better to have a massively bad year at, say, age-33, and then be out of the league, than it is to have a long, slow decline. And in fact, while the player with the long, slow decline tends to actually be the better player, he will register worse by the measure. It’s just not a good way to compare players’ WAR to one another. Now, if you’re going to use it to look at making up for strike time, or shorter schedules in the early days of baseball, or something like that, I have no problem with that (though, really, in those cases, it would have to be regressed a bit). But just dividing WAR by PA and calling it “good” simply does not work. It doesn’t give a “different kind of answer;” in a case like Biggio vs. Knoblauch, it gives an incorrect answer.
To work my final point into an analogy I just thought of, let’s say you and I run a race; we decide to run until we get tired. We start at the same time. You come out guns blazing and you toast me. You finish in four minutes. I finish in five minutes (these are not remotely realistic, at least for me; Bob, for all I know, you were a world-class miler). After that five-minute mile, something I couldn’t even do in high school, I quit. However, you’ve started your second mile, so you decide to keep going. You finish your second mile. But, winded from your Olympic-like first mile, it takes you twice as long – eight minutes – to finish the second mile.
By your logic, the faster runner in this example was me; I ran 12 MPH, you ran 10 MPH. Except the problem is, at no point in the race was I running faster than you. While I was running 12 MPH, you were running 15; while you were running 7.5 MPH, I was running 0. You were faster at every point, so you were faster overall, even if your overall rate of speed was slower than mine. Really, the way to do it would be to compare our speeds to a hypothetical, non-zero baseline of a below-average runner, to account for both of us running FAST at first, and me running slower-than-slow the second mile, while your second mile was fast… which is EXACTLY how WAR already handles the problem! No need to bother looking at it as a rate stat, because it pretty much already is. When you look at it like a rate stat, you end up drawing the conclusion that I wrote about here, that the slower runner is actually faster.
Doom,
Re “Really, the way to do it would be to compare our speeds to a hypothetical, non-zero baseline of a below-average runner, to account for both of us running FAST at first, and me running slower-than-slow the second mile, while your second mile was fast”
For that matter, why not do the comparison by using an average “runner” – you know 81-win players versus 50-win players? Since most ML GM’s will hopefully dump below average players (kind of like you sub-par casual joggers having no business competing in a track event at the highest level) , by comparing to the “average player” we get an idea who really wasn’t much help at all and who, actually, was a detriment to winning?
But, I’m all for peak comparisons and don’t really care much about comparing players who hung on too long or weren’t contributing. It’s their peak that counts – kind of like Joe Louis, shot and shopworn, losing badly to Marciano. I don’t think Marciano was as good as Louis. When we have that barbershop argument, you know, it’s Ray Robinson in his prime versus Hagler in his prime or, Duran (not when he’s 40) versus Pernell Whitaker, etc…
Paul, your prime=100% approach is certainly one of many valid ways to look at the Hall. CAWS takes a prime=80% approach. Other people may take other percents. It’s not as though there were an actual right answer. Yours has the virtue of being extremely simple to apply.
Well, we’ve had this general argument many times, Doom, and I do appreciate your point. For starters, I’d want to note that I’m not at all saying that WAR/500PA gives you an answer as to who had the better career. I like it as a starting point; obviously, longevity counts. If you want to use WAR or WS as the first cut that’s fine too, but if you do a full assessment you’re going to have to deal with what the rate stat shows at some point, because everything a player does on the field matters, no matter what age he may be, and someone who uses a portion of his career to be a compiler, as Biggio did and as Pujols is doing now, has to be identified as such.
Biggio’s last seasons were not simply a matter of decline: he was a below average player who finished up well below AAA level. In our race, where I embarrassed myself with my awful 8-minute mile, I hurt no one but myself, and, of course, the feelings of my many fans. But to the degree that he was a weak link in Houston’s line-up over that final period, Biggio was causing damage to the purpose of the game: to win. To make your analogy serve its purpose our mile runs should be part of a team effort — a relay race — where I boosted my team by running my customary 4-minute mile, leaving you in the dust, but you (like Knoblauch, once the PEDs failed him) handed the baton off to another good runner while I insisted on running a second leg and wound up costing my team the race.
All this doesn’t make Biggio any less great when he was great, but it counts: it really happened and it does diminish his career. He was a 1.2 WAR guy over a period of eight seasons, full time, 5000+ PA, with a cumulative WAA of -5.3. That is unfortunately just as much a part of his career as the nine seasons when he was great. They both really happened. WAR doesn’t just give positive credit for positive performance, it also gives positive credit for barely-MLB performance. Although you view Biggio as still adding value with his play in that long decline, I do not see it that way: he was not adding value to the team goal of winning. The only teams that benefit from a 1.2 WAR full-time second baseman (or center fielder) are teams that aren’t actually in the running. To win a division or pennant, that sort of performance by a regular is something that needs to be overcome by his teammates.
On a very different tack, I think you’re ignoring my final point. We know that for a sustained period Biggio was a great player. When the stats tell us his defense was poor (which it was) and that his homogenized career quality was not at all what we’d expect, our job, as I see it, is to reconcile those two different views. His peak was really terrific and when you look at his traditional stats you can see his virtues in terms that are very clear: the WAR rate tells us he was a very ordinary player, but the peak stat tells us to look again and in detail. My general argument is that no one number can substitute for a detailed historical assessment: we need an array of stats and often the narrative as well.
If the question were whether Biggio should be de-Halled I’d be arguing in defense of his enshrinement in spite of his long period of lesser or greater mediocrity. No way Knoblauch is as good as Biggio. But what you won’t find me doing is ranking Biggio ahead of players like, say, Gehringer, Grich, or Robinson.
Anyway, glad to create some argument here at last. And, just for the record, my recent miles really haven’t quite met my 4-minute standard, so I’ve given up relays.
Going way out on a limb, Bob wrote: “But by my reckoning here, Biggio is really not Hallworthy.”
Forget all about our “modern” fancy stats – any player who collects 3000 hits at the major league level is Hallworthy – and deserves induction on the first ballot.
Mike, I don’t have any automatic qualifying numbers and you’re not going to be able to convince me that Palmeiro belongs in the Hall (and I suspect you’ll have a tough time convincing Doom that Rose belongs), but Hall criteria are up for grabs, as Bill James illustrated so entertainingly years ago. Our disagreements are what makes discussion interesting.
And, as I pointed out to Doom, after saying that by my reckoning Biggio isn’t Hallworthy, I was hoping I’d made clear that this meant my “reckoning here” couldn’t be the last word.
Bob,
I assume we are talking here only about whether players have “HOF numbers” – achieved on the field of play – and not about bans or PED issues which are clearly another story. And it is interesting to debate whether “borderline players” (according to the numbers) deserve to be in the Hall (as we are doing in this thread). And using “creative stats” and other considerations to do this is fine.
But players like Pete Rose and Craig Biggio and Rafael Palmeiro are not borderline players according to the numbers – all clearly have HOF numbers (as do a number of other suspected PED abusers).
Your point’s well taken, Mike, and I’ll withdraw mine concerning Palmeiro and Rose. But I would want to consider larger issue that Biggio’s case points to, but does not exemplify (because I believe Biggio does belong in the Hall).
Biggio accumulated 3000 hits, but it was at cost: (1) throughout his career his offensive contributions were usually balanced by indifferent to poor defense; (2) over about 30-40% of his career his contributions were below average, costing his team wins. Now, if we place on one side of the Hall scale his outstanding decade of offense and on the other his persistent defensive problems and his below average period, how does that affect the way we assess his benchmark of 3000 hits? This is especially pertinent if we see his pursuit of 3000 hits as the driving force that generated a number of those poor seasons. I’d argue that the negatives significantly detract from his hits, especially if you look at every batted ball Biggio failed to reach that an average fielder would have as the equivalent of one lost hit.
I think the balance is still in Biggio’s favor, but I’d argue that at some point, that balance could tip without regard to the hit total. Add a few more errors each year, a few more caught stealing, a few more GIDP (which Biggio avoided well), and take away many of those HBP . . . After all, Biggio was not a high-average hitter; he only reached 200 hits once. He got to 3000 by adding seasons as a compiler. His last six seasons, he produced 30% of his career hits while contributing a 92 OPS+ and significantly sub-replacement-level dWAR at positions that carry positive positional adjustments, averaging double-digit negative Rfield. Why should he be celebrated for that? He could never have started an MLB career like that because he would have been back in Triple-A.
I think Biggio belongs in the Hall despite the cost of reaching 3000 hits, not because he crossed that arbitrary line.
I really don’t like throwing around a phrase like “creative stats”. It’s not appropriate. A hit is itself relatively arbitrary. Did the fielder make an “error” or was it by your own credit? If there’s a guy on first and you’d be safe at first yourself but the guy on first doesn’t make it to second before the throw, that may or may not be a “hit” depending on a lot of subjective criteria that doesn’t really relate to the outcome of the play.
3000 “hits” is still a historical complexity too because the number of games has varied, seasons have been suspended, military service, segregation, let alone just good old blackballing. That also ignores all manor of off the field injury or personal tragedy. Does the guy hit on Yom Kippur? Did he hit in a segregated league? How much of one player’s opportunities are solely a reflection of their own talent? Where did they hit in the lineup? I could go on and on.
All stats are “creative”. The purpose is to find patterns that are predictive. Baseball is a complex game and poorly reflected in a single stat line for value but if you do, it should be WAR. If the hall has any arbitrary lines based on career accomplishment in a single stat, WAR would be the least offensive. Hits is pretty far from the top of my list.
On the other hand, you can get more hits if the other guy doesn’t pitch on Yom Kippur.
mosc,
I do not see anything in your note that would change my view that any player who produces 3000 hits is a first-vote Hall of Famer (bans and PEDs aside). Can you suggest any player who has 3000 hits who would not satisfy almost any other HOF criteria?
Mike, I think that currently all 32 players who have reached 3,000 hits were very good offensive players, and they probably all belong in the Hall. You can quibble about a few – Lou Brock has certainly been talked about quite a bit over the lifetime of this website – but for the most part they all fit the bill. But I think 3,000 hits isn’t a great criteria because it’s very possible for a player to reach that bar and not be anywhere near HoF quality. Slap a few more seasons onto Doc Cramer’s career – have him come up at Age 20 or so – and he might reach 3,000 hits. And he still wouldn’t be within spitting distance of a HoF worthy career.
Obviously every number only makes sense in context, but something like, I don’t know, 700 HR’s – it’s tough to imagine a bad offensive player ever getting to that. Maybe the game changes so much that players are hitting 100 HR’s a year and suddenly journeymen are getting to the 700 HR mark. It’s possible, but about as probable as a team winning the World Series because the opposing team’s ace is abducted by aliens. But a mediocre offensive player getting to 3,000? It’s quite possible to imagine it because it’s almost happened.
Now you are really reaching. It is not possible for a player to reach 3000 hits and be “mediocre.” History has proven that to be true. Any player who reaches 3000 hits is a Hall of Famer. I cannot imagine it otherwise. “what-ifs” do not count – only reality counts.
Mike, why is it implausible for another player like Cramer to pop up except getting just 10 or so more hits a year? Or an Omar Vizquel type player who gets just 5 more hits a year? Vizquel was not mediocre, but his hitting was, and he ended up very close. This isn’t some incredibly out there scenario.
Mike, I would agree that it’s not possible for a player to reach 3000 hits and be mediocre. Certainly none has. But “better than mediocre” is not the standard for the Hall.
In your book, you calculate Lou Brock as barely crossing the Hall threshold, with a CAWS score of 285, only 5 points above the minimum for his position. That is the Win Shares-based view. Consider Brock’s WAR-based numbers — let’s compare him with an outfielder CAWS and the BBWAA sees as nowhere near Hall status, Dale Murphy:
WAR (fWAR)…WAR/500PA…Peak7…Best10…OPS+…..dWAR…Career
45.3 (43.2)…………2.0…………..29.8……….39.7……109……..-16.3……2.2………Lou Brock
46.5 (44.3)…………2.6…………..37.2……….47.1……121……….-6.8……1.8………Dale Murphy
Note that I’ve included WAR’s version of best ten seasons as a matching alternative to the CAWS WS-based formula, and that on these two players, bWAR and fWAR are closely aligned. Murphy is ahead — in some cases far ahead — in every category other than career length. Brock, in his career-closing pursuit of 3000 hits, piled up a three-season total of -2.0 WAR (including sub-replacement level offense and defense) in collecting the final 300 hits necessary,
Murphy collected 2111 hits; Brock 3023. But I see no way in a WAR-based framework even to begin to suggest that Brock was the superior player, and Murphy is no Hall of Famer. Look at Brock through 1974:
WAR…WAR/500PA…Peak7…..Best10…OPS+…..dWAR…Career
44.1……..2.5……………29.8………39.7……114……..-12.3……1.8…….Lou Brock, 1961-74
In a meaningful sense, this is Lou Brock’s cumulative peak, right after his 118-SB season, and he’s within shouting distance of Murphy — not quite there, but almost on a par with that certain non-Famer. He has at this point 2388 hits. From there on he becomes like late-Biggio without early-Biggio:
WAR…WAR/500PA…OPS+…..dWAR…Career
..1.2……..0.3……………..92……….-4.5……0.4…….Lou Brock, 1975-79
Do you really want to argue that the 635 hits Brock accumulated during this period add to his Hall qualifications? Win Shares, which knows no negatives, assigns Brock about 15% of his career value for these five seasons (essentially the margin it assigns him above Murphy), but are you so committed to the absolute inerrancy of Win Shares and CAWS that Brock’s skin-of-his-teeth Hall qualifications under that system show that every 3000-hit player’s credentials are beyond question? The WAR stats indicate that Brock doesn’t deserve to be in Hall; JAWS ranks him the 36th best left fielder ever; the Hall of Stats assigns him a 72 score.
Brock was a talented player, as was Murphy — neither was “mediocre” — but I would not put either close to the Hall. Brock attracted press coverage for his stolen bases (the writers largely ignored the caught stealing counts that virtually negated Brock’s net value on the base paths); his postseason play in ’67-’68 was superb (although when he was caught stealing in ’68 it abruptly turned the tide of the Series). Brock’s reputation as a speedster and association with winning teams in the ’60s gave him the leverage to play on and on after his value had fallen to replacement level, far below average. That is how he gained time to pile up 3000 hits. If you deem 3000 hits a sufficient criterion for Hall membership, then, by definition he’s Hall member. But if you mean to say that 3000 hits is invariably an index of a career whose value meets Hall criteria independent of that single stat, I think Brock is strong counter-evidence.
Bob,
Well said. Your argument is that it is possible to accumulate 3000 hits and still fall below what would be considered “accepted HOF numbers” by many. And I would agree with your point. I am thinking of Ichiro as another possible candidate here.
Having said that, I would say that I would definitely vote for any player who has 3000 hits for the Hall of Fame – just as I would vote for any position player with 400 career win shares or a core value of 300 or a pitcher with 300 career win shares or a core value of 200.
I guess what I do believe is that in some cases longevity is, in fact, a sign of true greatness.
Mike, Bill James’s Hall insight was to capture the fact and consequences of there being no single set of Hall standards. The Hall has been around 80+ years, and we’re not going to change that — at least, I hope not: I’m sure the standards the “authorities” designated would not be mine. (If they let me make the decisions, I’m ok with it.)
So I see the point of the discussions here as crystallizing the varieties of positions that baseball obsessives like us adopt with regard to what Hallworthiness means and how that should determine its membership. You and I don’t agree, but agreement isn’t the point: it’s understanding. I see the sorts of benchmarks you name — a traditional and contemporary mix — as arbitrary: why should 3000 be different in any significant sense from 2999? But I’m not a completely soulless nerd: I resonate to the magic of these exceptional, historic round numbers, and I can fully understand how it would be possible to feel the Hall was diminished by leaving an eligible 3000-hit player on the outside.
Now that discussion is beginning on second basemen (although not really on nsb’s question so much as the way my comment about Biggio has annoyed people), I’m adding the comparable list of third basemen. These are listed according to CAWS assessment, but the numbers are WAR-based, as before (OPS+ being a control stat). Again, the numbers in parentheses are ranks according to WAR/500PA, which, in fairness, I should not that Doom objects to as a distortion of WAR. Players not in the Hall are denoted by X (and there are many, 3B being a famously underrepresented position), and the two asterisks are the players nsb has asked us to comment on, since they are in the Hall of Fame, but not the Hall of Stats. (The first line is Mike H’s CAWS cut-off for Hallworthiness; the second indicates the end of the list of the Top 25 CAWS candidates; Kell and Lindstrom are an unspecified number of ranks further down.)
WAR…WAR/500PA…Peak7..OPS+…dWAR…Career
106.8………5.3……….57.2……147………18.4……..2.0………Mike Schmidt (1)
..96.6………4.8……….51.4……143………..5.6……..2.0………Eddie Mathews (2)
..88.7………3.8……….47.8……135………..2.2……..2.3………George Brett (7)
..91.4………4.3……….56.3……131………13.9……..2.1………Wade Boggs (3)
..85.2………4.0……….41.4……141……….-0.9……..2.1………Chipper Jones (5)
..70.5………3.8……….53.8……125………..8.7……..1.9………Ron Santo (9)
..95.6………3.9……….45.3……116………29.3……..2.4………Adrian Beltre (6)
..58.8………2.7……….28.6……119………..0.2……..2.1………Darrell Evans (21)x
..78.4………3.3……….42.2……104………39.1……..2.4………Brooks Robinson (14)
..52.6………3.1……….29.4……119………..1.4……..1.7………Stan Hack (19)x
———————————————————————-
..61.5………3.7……….41.5……119………..8.5……..1.7………Sal Bando (10)x
..70.2………4.1……….41.8……122………21.2……..1.7………Scott Rolen (4)x
..53.3………3.6……….35.6……113………16.8……..1.5………Jimmy Collins (11)
..48.2………3.4……….35.5……118………..8.6……..1.4………Heinie Groh (12)x
..50.4………3.1……….31.6……124……….-3.0……..1.6………Bob Elliott (20)x
..68.0………3.3……….40.9……110………21.4……..2.0………Graig Nettles (15)x
..62.8………3.8……….45.1……116………10.7……..1.7………Ken Boyer (8)x
..36.2………2.2……….24.0……107………..2.0……..1.7………Pie Traynor (25)*
..53.8………3.2……….36.6……121………..6.6……..1.7………Ron Cey (17)x
..66.3………3.3……….40.5……109………23.8……..2.0………Buddy Bell (16)x
..56.1………3.4……….32.4……114………17.9……..1.7………Robin Ventura (13)x
..34.1………1.9……….24.5……109……..-10.7……..1.8………Eddie Yost (27)x
..48.3………3.1……….27.2……109………10.8……..1.5………Larry Gardner (18)x
..33.5………2.4……….25.4……116………..1.6……..1.4………Ken Caminiti (23)x
..38.5………2.2……….27.9……102………10.5……..1.8………Tim Wallach (26)x
———————————————————————
..37.4………2.5……….25.6……112………..1.6……..1.5………George Kell (22)*
..28.3………2.3……….22.0……110………..2.6……..1.2………Freddie Lindstrom (24)
I need to add a name to the third base list. In consulting Mike’s updated list for 2018, I failed to notice that Frank “Home Run” Baker’s name had been dropped, as were some other “Dead Ball Era” players. He belongs in the 7th rank, just below Ron Santo, and his WAR-based stats are these:
WAR..WAR/500PA…Peak7..OPS+…dWAR…Career
..62.8………4.7…………45.5…..135…….9.6……..1.3…………….Frank Baker (3, with others moving down)
Alex Gordon set the KC record for being hit by pitch. And manager Ned Yost quipped that he never got hit in his career, because he was too easy an out.
He had 640 PA.
Wondered what the record might be.
A lot more than 640:
3664 … Mark Lemke
3229 … Bill Bergen
2324 … Mickey Witek
2069 … Herm Winningham
1920 … Tom Hutton
1646 … Scott Livingstone
1643 … Rob Andrews
1494 … Jim Norris
1435 … Coaker Triplett
1393 … Bobby Brown
1344 … Wes Ferrell
1324 … Bullet Joe Bush
Wes Ferrell, a pitcher, the only good hitter in the top 11.
Lemke had another 257 PA in the post-season… and didnt get hit then, either.
I suppose we can draw the inference that none of these folks were sufficiently feared that pitchers had to go inside to keep them from extending their arms?
Certainly nobody was afraid of Lamke. But still, Lemke probably saw 15,000 pitches. And not one of them missed badly enough to hit him. There’s some luck involved there.
Voomo
Good to see that Gordon has retained some skill or another. Here’s OPS+ LF/RF, 60% G , age 32-34…the bottom 8
RK… Player……………….OPS+………PA
10 Dan Gladden .. 85… 1541
11 Gino Cimoli 85…. 1258
12 Juan Pierre 82…. 1884
13 Jack Tobin 82…. 1097
14 Vernon Wells 81…. 1249
15 Alex Gordon 80…. 1615
16 Lance Richbourg 80…. 1027
17 Vince Coleman 71…. 1071
While he was hit by pitch 7 times early in his career, Ruben Sierra went 5652 PAs without being hit (his 1991-2006 seasons or 1991-end of his career). He had a 100 OPS+ during those years.
Game logs in 1990 show he was hit on September 4th, so we can add another 112 PAs from September 5th through the end of the 1990 season. He was hit in his first plate appearance on 9/4/90, so we can add another 3 PAs. I’m not sure if 5767 PAs is the longest streak without being hit by a pitch, but there are only (as of today) 748 players in history with that many career PAs.
Artie, You’ve found the winner. Eddie Murray seems to be second with 4038 PA, from his final PA on July 8, 1990 through the end of his career in 1997. During that period, Murray had an OPS+ close to 110 (estimating the rest of 1990, when he got red hot in the second half — .361 BA — and had the best season OPS+ figure of his career, 159, before slipping to 103 for his remaining seasons).
To all and sundry (with a special greeting to Mike H, as a new and welcome fellow contributor to the nonsense):
I’ve been in the process of moving this week, and at any age, much less mine, its an exhausting endeavor, so I haven’t contributed the way I probably would have in normal circumstances, especially since this post is my project.
Here’s what I expected to happen in my absence that didn’t, or hasn’t so far re the infielders: Arguments for or against this or that individual player. You can only go so far, it seems to me, with beating a dead horse, which in this instance is that none of these guys makes the top echelon or even comes close by modern statistical reckoning. That in fact, is the given.
But what, say, about Luis Aparicio? I hated him when I was growing up, mainly because he was on the White Sox, But I never believed him to be a mediocre player, quite the contrary. His fielding was outstanding, and even before Maury Wills and Lou Brock, he rejuvenated the stolen base as a tactical weapon. In 1959 he swiped 56. Mantle was 2nd with 21. In 1960, he led with 51, Landis had 23. He led the AL his first 9 seasons with an unusually high SB% year after year. Does this or doesn’t it constitute the type of “Fame” that the Hall of Fame was designated to celebrate?
Got to go now, but I welcome your responses.
nsb, Mike H has already started responses to your comment below, but here’s another.
I agree that Aparicio played an historic role in the revival of the stolen base, though, as Mike’s comment pointed out, not quite as dramatically as it may have seemed at the time. A lot of Aparicio’s profile in theft was the product of the press’s presentation of the ’59 Go-Go Sox, who were, indeed, a great story (anyone displacing the Yankees would have been at that point). It was only in ’59 that Aparicio moved from the league leader in a sluggish SB period to a truly great base stealer, and the following year Wills, a virtual rookie, joined him at the 50 SB-level. Aparicio had a good career as a slick fielding lead-off man with unusual speed, but I think the Hall’s “fame component,” which has to be judged by voting results, is too uneven and uncertain to use as a standard. Generations recited “Tinker to Evers to Chance,” but how many recalled Aparicio’s role decades after his career ended, and is that a standard to use in any event (although when it comes to Adams’ jingle, as I see it two out of three ain’t bad)? Luis’s revival of the base-stealing specialist role was, in my view as in yours, historic, and he was a fine shortstop; however, I think his career stats fall too far short to sustain his claim to fame at Hall level. I think the basic reason is simple: although he stole lots of bases when he got on base, he simply didn’t get on base enough. Compare the OBP of this list of past and future “Stolen Base specialists”:
Max Carey .361
George Case .341
Luis Aparicio .311
Maury Wills .330
Lou Brock .343
Rickey Henderson .401
Even “non-historic” base stealers like Ron LeFlore (.342) and Tim Rains (.385) got on base far more than Luis.
If I had to pick a Hall of Fame member among the four candidates for de-Hallification that you’ve listed, I’d probably plunk for Sewell. Sewell’s offensive WAR stats are far better than Aparicio’s (and CAWS likes him better too, despite his shorter career with fewer WS), and Sewell’s historic forte, avoiding strikeouts, has stood up over time: he’s the man when it comes to that circus trick. He wasn’t the fielder Luis was, but he was an average shortstop, not a weak one, just below zero in Rfield. I’m not advocating for Sewell for the Hall, and I don’t argue with the claim that the press made Luis more famous for stolen bases than Sewell is for making contact, but I’d think Sewell would be in line before Luis on overall quality/fame grounds. Of course, they’re both in anyway.
I’d like to respond to nsb’s query regarding third basemen, his challenge being to argue for or against the Hall members who fail to meet Hall of Stats criteria: Pie Traynor and George Kell.
I’m having a hard time imagining good arguments for either. Both were good hitters and minimally capable fielders, but “good hitter” has to be defined in terms of the standard of the time: batting average. Traynor has an exceptionally good lifetime batting average, .320. However, Traynor played in a high-average era: his top average (1930) was .366 — great right? — which was good for 9th place in the NL (he had only one Top 5 finish: fifth, in 1927, with .342). Moreover, his BA was pretty empty: little power and few walks (but also few strikeouts and a lot of good bunts). He did hit triples, but in those days Forbes Field was a triples factory — the Pirates almost always led the league in triples, usually by a lot (Owen Wilson played in Forbes when he set the record for triples). Traynor was elected to the Hall in 1948, when batting average was still king, and it’s not surprising, but great a hitter as he was thought to be then, it’s more valid to see his offense in terms of his 107 OPS+: underwhelming. As we’ve often noted, third base was the field position least likely to attract quality players before the middle of the 20th century, and Traynor, with his high average, was widely viewed as the greatest of all time there. But I see no case for him from contemporary perspectives.
Kell was another high average hitter: .306 in the ’40s and ’50s. Those decades weren’t the hitting fest that the previous two had been, but they were still a relatively high-average era. Like Traynor, Kell’s average was empty of much power: where Traynor hit triples, Kell hit doubles, and like Traynor he avoided strikeouts and was a league-leading bunter. (During Kell’s career, the premier AL third baseman was probably Al Rosen, who amassed almost as much WAR in a career only about 60% as long.) Neither Traynor nor Kell was a distinguished fielder; their defenses were just above replacement value. (It’s worth noting that between Jimmy Collins (1895-1908) and Brooks Robinson, there really aren’t any particularly distinguished fielders among leading third basemen.)
Both Traynor and Kell had an extra advantage in terms of Hall voting: both went on to careers as baseball broadcasters, which kept them in the eye of the public and the minds of the BBWAA and Veterans Committee voters. Traynor was, by most accounts, a terrible radio broadcaster but much beloved. Kell was pretty good — he was one voice of the Tigers during the ’60s, when I lived near Detroit, and he was skilled enough not to be an embarrassment when paired with his partner, the great Ernie Harwell. Nevertheless, I was very surprised when he was named to the Hall.
So in the case of third base, I’m not tempted even a little to advocate for the two Hall of Stats de-Hallification designees.
In speaking of Luis Aparicio, nsb stated “he rejuvenated the stolen base as a tactical weapon.” I beg to differ. The player who rejuvenated the stolen base as a weapon was Willie Mays (the greatest all-around player in history) – just one of his many accomplishments.
In 1956, Mays stole 40 bases. When was the last time prior to that that anyone stole 40 bases? He followed that with 38 in 1957, 31 in ’58 and 27 in ’59 (more than Mantle’s 21). Aparicio was just one of a number of players who “followed Willie’s lead.”
This is an interesting issue with an outcome that seems to me unclear. Mays stole 40 bases in 1956 and led the NL, while Aparicio, a rookie, led the AL with 21, which shows how low SB totals had fallen. But they hadn’t fallen for long. The answer to the question of the last time prior to Mays when soeone had 40 SB is 1944, when Snuffy Stirnweiss had 55; George Case had stolen 61 the year before. But, more to the point, a number of players had totals close to Mays’s more recently: e.g., Jackie Robinson, 37, in 1949, Sam Jethroe, 35, in 1950 and 1951, and Bill Bruton, 34, just two years before Willie.
So why is Aparicio’s name associated with the “revival” of the stolen base? Because while 40 was Willie’s peak in 1956, Aparicio made the stolen base his trademark skill (like Bob Bescher many years before, then Max Carey and George Case). Aparicio led the AL for 9 straight years, during which he averaged over 40 SB per season, four times breaking 50. By the time his streak was done, another trademark thief, Maury Wills, had stolen a march on Luis by averaging over 60 in the course of taking six consecutive NL titles, his top total of 104 almost doubling Luis’s best (57).
So I think it’s fairest to say that the eclipse of the stolen base was somewhat overstated (league-leading totals were particularly low during the pre-War years in the NL, and during the post-War years in the AL), that Mays was one of several NL players who maintained the importance of the SB during the post-War period, that Aparicio brought back the role of the “stolen base specialist” that had lapsed after George Case, and that Wills demonstrated that dead ball era levels of the stolen base could be duplicated in the modern era.
Stolen base leaders tend to be concentrated among a limited number of players, certainly much more so than other offensive categories. These are the players since 1901 to lead their league in SB for three or more seasons consecutively.
NL
Bob Bescher, 1909-12, 4 leading seasons overall
Max Carey, 1915-18, 1922-25, 10 seasons
Kiki Cuyler, 1928-30, 4 seasons
Bill Bruton, 1953-55, 3 seasons
Willie Mays, 1956-59, 4 seasons
Maury Wills, 1960-65, 6 seasons
Lou Brock, 1966-69, 1971-74, 8 seasons
Tim Raines, 1981-84, 4 seasons
Vince Coleman, 1985-90, 6 seasons
Tony Womack, 1997-99, 3 seasons
Jose Reyes, 2005-07, 3 seasons
Michael Bourn, 2009-11, 3 seasons
AL
Ty Cobb, 1915-17, 6 leading seasons overall
Ben Chapman, 1931-33, 4 seasons
George Case, 1939-43, 6 seasons
Bob Dillinger, 1947-49, 3 seasons
Minnie Minoso, 1951-53, 3 seasons
Luis Aparicio, 1956-64, 9 seasons
Bert Campaneris, 1965-68, 6 seasons
Rickey Henderson, 1980-86, 1988-91, 12 seasons
Kenny Lofton, 1992-96, 5 seasons
The current 23 year gap since the last AL three-peat (Whitt Merrifield could do it this year) is the longest for the junior circuit and a close match to the longest NL gap of 25 years between Cuyler and Bruton.
Doug, I think your list suggests that players with streaks of stolen base titles fall into three main categories (with borderline cases). There are some terrific all around players who included speed on the basepaths in a more extensive tool kit of outstanding skills: e.g., Mays, Cobb, Henderson. There are players who were a notch lower in general quality, but whose skills as stolen base threats magnified their strong abilities to boost them into Hall territory: e.g., Minoso, Lofton, Raines. And there are the true stolen-base specialists, whose strings of SB titles are what identified them as players, some even parlaying those into Hall careers despite limited qualifications otherwise: e.g., Bescher, Carey, Case, Aparicio, Brock, Campaneris (there’s a reason Carey, Aparicio, and Brock are all on nsb’s list).
Another “ancillary skill” specialization that shows some of the same patterns is HBP, where you also find many league-leader streaks and a similar division among players, although without any at the Mays-Cobb-Henderson level (unless you see Frank Robinson that way, or A-Rod, who didn’t ever lead his league, but who was a skilled plunkee). Some HBP specialists who enhanced an already strong profile would include Hughie Jennings, Craig Biggio, and Chase Utley. The lesser specialists, who basically used the HBP to help sustain their roles as regulars, would include Frank Crosetti, Ron Hunt, and, at a somewhat higher level, Don Baylor. Crosetti and Hunt seem a lot like, say, Bescher and Case. The one guy who appears prominently on both lists is Minoso, with those three SB titles (with moderate numbers) and ten HBP titles (in 11 seasons — his only career qualifying seasons).
There’s a kind of half-parallel with walks. Players like Roy Thomas, Topsy Hartsel, Donie Bush, Eddie Stankey, and Eddie Yost survived and stood out for their skills in working the walk — they aren’t black-ink guys, but they have plenty of black ink in the BB column. But the league leader category there is also filled with the best hitters (Williams, Bonds, etc.) for obvious reasons. Still, I think there is real charm in looking at the lifetime BB leaders and seeing the Top 10 list of great sluggers, followed at #11 by Eddie Yost, a .250ish hitter without power whose career was shorter than any of the greats listed above him. (And it’s really spooky that Yost’s skill in drawing a walk was almost matched by the identically named Eddie Joost [yoost]. Yost batted .254 and walked .176; Joost batted .239 and walked .154.)
“Eddie” is definitely the name with the highest walk rate. Beyond Yost, Joost and Stanky, you’ve got Collins and Lake. Mathews and Murray weren’t OBP specialists, but both have good walk rates. Heck, even Eddie Robinson drew his fair share of walks. And the greatest OBP man ever, a certain Mr. Gaedel, also an Eddie.
But I think you make a good point about the tops of these lists tending to split between truly great players whose skills portfolio included other abilities and guys whose careers derive almost entirely from that one skill. Ferris Fain, Max Bishop, Roy Cullenbine – these are good players whose walks provided a lot of value, but I don’t think anybody, even the most OBP obsessed, considers them Hall of Fame candidates.
Those guys tend to have a symbiotic relationship with players lower in the lineup, the lower OBP power guys that make big bucks driving them in. That’s the conventional lineup wisdom, of course. How good would an offense comprised of only the OBP guys be? The best offenses in history all have good OBP’s, of course, but they have good slugging too. If your team is Roy Thomas, Elmer Valo and Roy Cullenbine in the OF, infield of Yost, Joost, Bishop and Fain, Wally Schang catching (obviously Cochrane is the best OBP catcher but he’s got too much power), are they going to outscore the 1950 Red Sox?
Good points, CC. But per your final observation, let me bring this back around to Aparicio. The 1959 ChiSox — the only AL team to best the Yankees over a full decade span — featured two Top 10 OBP members, two Top Ten SB members, three Top 10 in HBP, and four members of the Top 10 in sacrifice bunts. The team was 7th of eight teams in slugging, led by catcher Sherm Lollar’s 22 HR and 84 RBI (neither a Top 10 figure). Could they have beaten out the ’50 BoSox? Maybe not, but they won their pennant by five games, with no one else close.
Not quite your all-OBP team, and not a true counter-argument to your correct general point. But there are always these oddball exceptions. (The 100-victory ’69 Mets easily topped the other eleven clubs in the NL despite being 11th in slugging and 10th in OBP. If the pitching all came together just right, the All-Eddie All-Stars might make it after all.)
I strayed a little in that last reply . . . “Outscored?” No.
It’s hard to isolate offensive factors. Nobody ONLY takes walks, let alone a team full of non-hitters. Certainly can line up 4+ events in a row of a decent IBO rate every few innings to score but the odds of that are staggeringly small. The bases open followed by 4-walk inning is not impossible, just unlikely. Certainly the solo HR somewhere surrounded by 3 outs is more common so a direct comparison is difficult.
The best we can do is do correlation of W/L records or run scoring totals for teams based on their relative slugging or OBP rank. That’s pretty clear though, leading the league in slugging is less correlated with team success than leading the league in OBP. Both are good and it’s not a skill do do one in spite of the other, but OBP is a better single stat that slugging.
Since nsb has started the discussion of shortstops by flagging Aparicio as a player with substantial qualifications for his place in the Hall of Fame, Hall of Stats notwithstanding, I’m going to post figures for the shortstops either in the Hall or highly rated by CAWS. The rank order below is according to CAWS, but the figures are, for the most part, WAR-based, with the rank in terms of WAR-rate in parentheses after the names. Asterisks are for shortstops nsb has flagged for discussion, and X marks players not currently in the Hall. The first line represents the CAWS cut-off for Hallworthiness, the second is the end of the list of highest CAWS-ranked shortstops; the four names below are Hall members ranked various points lower. Note that Peak7 figures for Davis and Maranville are in italics — Davis played only 4 games in 1902 because of the league wars, and Maranville played only 11 games before enlisting in the Navy in 1918: I have skipped those years in calculating their “consecutive” Peak7 totals.
WAR..WAR/500PA…Peak7..OPS+…dWAR…Career
130.8………5.6……….65.0……151………21.3……..2.3………Honus Wagner (1)
117.8………4.8……….57.6……140………10.6……..2.4………Alex Rodriguez (2)x
..72.9………4.7……….49.3……136………12.0……..1.5………Arky Vaughan (3)
..95.9………3.7……….46.9……112………35.7……..2.6………Cal Ripkin (9)
..77.3………3.2……….43.8……115………..6.8……..2.4………Robin Yount (21)
..72.4………2.9……….37.5……115……….-8.3……..2.5………Derek Jeter (25)
..74.5………3.6……….36.2……113………19.0……..2.1………Luke Apxpling (13)
..84.3………4.1……….40.1……121………24.0……..2.0………George Davis (5)
..66.4………3.8……….39.5……119………14.3……..1.8………Joe Cronin (8)
..75.4………3.6……….32.0……110………28.5……..2.1………Bill Dahlen (14)x
..70.4………3.9……….32.6……116………14.4……..1.8………Barry Larkin (6)
..67.5………3.2……….47.3……122………..5.1……..2.1………Ernie Banks (19)
..66.3………3.5……….33.7……..99………25.4……..1.9………Pee Wee Reese (16)
..63.0………4.5……….46.3……120………23.4……..1.4………Lou Boudreau (4)
..70.7………3.8……….40.7……110………22.7……..1.9………Alan Trammell (7)
..70.3………3.7……….39.4……105………28.7……..1.9………Bobby Wallace (12)
..47.3………2.6……….36.6……108………..6.9……..1.8………Miguel Tejada (28)x
..76.9………3.6……….41.1……..87………44.2……..2.2………Ozzie Smith (15)
———————————————————————-
..45.5………3.1……….32.7……119………..9.1……..1.4………Vern Stephens (22)x
..53.7………3.2……….35.9……108………..9.1……..1.7………Joe Sewell (20)*
..48.7………3.3……….40.6……113………..7.9……..1.5………Jim Fregosi (18)x
..48.5………2.9……….34.3……..98………23.5……..1.6………Dave Bancroft (24)
..45.3………2.6……….29.3……101………14.9……..1.8………Tony Fernandez (27)x
..42.8………1.9……….25.8……..82………30.8……..2.3………Rabbit Maranville (30)*
..53.1………2.8……….34.8……..89………21.1……..1.9………Bert Campaneris (26)x
———————————————————————
..53.1………3.7……….31.6……..96………34.3……..1.4………Joe Tinker (10)
..40.8………3.0……….28.2……..93………22.9……..1.3………Phil Rizzuto (23)*
..55.8………2.5……….26.9……..82………31.8……..2.2………Luis Aparicio (29)*
..44.0………3.3……….32.1……102………22.9……..1.3………Travis Jackson (17)
In using Mike’s CAWS lists to organize WAR-based data, I’ve become aware of how disparate the WS-based and WAR-based approaches seem to be in many cases. While you may agree with Doom that I should not be choosing WAR-rate as my primary rank contrast with CAWS on the three lists posted here, using total WAR, career prime (Peak7), or a combination of OPS+ and dWAR all generate rankings significantly different from CAWS. There haven’t been all that many cases where there was enough contrast to argue interestingly between Hall of Stats and either CAWS or WAR, but there are lots of cases where I think it would be very interesting to argue between CAWS and WAR figures (e.g., the different stories the figures tell about, say, Stan Hack at third or Miguel Tejada at short).
I’ve resisted raising these issues here because the focus is on nsb’s project and because Doug has let us know that Mike will be offering some full posts on CAWS in the future — I think that would be the best time to explore these issues. But what has kept me from writing about this more has been the fact that to understand the issues, we really need to have a firm grasp of the differences between Win Shares and WAR approaches. I don’t have that grasp now, and I hope when CAWS does become a focus for us it will be an occasion to understand the contrast better.
Just a quick note at this point re the disparity between SOME (I would not say “many”) WS-based and WAR-based results.
As I am sure many of you suspect, it seems to have to do with how bWAR (and not necessarily fWAR) SOMETIMES treats fielding results. That is, bWAR tends to EXAGGERATE the importance of fielding when a player is perceived to a very good or very poor fielder. The problem is NOT necessarily with the defensive score itself but rather with the manner in which bWAR COMBINES the offensive and defensive scores to produce the total bWAR score.
Perhaps the best recent example of this “problem” is how bWAR ranks Matt Chapman and Charlie Blackmon in 2018. Chapman is considered a great fielder and is ranked #3 (much too high) while Blackmon is considered a poor fielder and is ranked #231 (when WS has him ranked #23 and fWAR has him ranked #66). If you go to seamheads.com, you can read recent articles I have written about this.
It has been suggested that WS undervalues fielding to some extent and that may also be true – particularly where the player is a truly outstanding fielder.
I should have added that one good example of this disparity is the career ranking of Derek Jeter and Ozzie Smith by JAWS and CAWS. WAR considers Ozzie to be a great fielder (which he was) and Jeter to be a poor fielder (which he was) – and so tends to allow that fact to distort their career value.
Jaws ranks Ozzie at #6 and Jeter at #9 for their careers – while CAWS has Jeter at #5 and Ozzie at #13.
I think the disparity in defensive assessment is a major factor in the way WS and WAR see players, but I’m not convinced that bWAR is distorting the effect through whatever oWAR + dWAR mechanism it uses (or is distorting the effect at all). I noted some reasons in a response to a previous reference you made on HHS to your Blackmon article. We also had a discussion here about your article on Chapman (in my contribution there, I really messed up a hyperlink).
And it can’t all be about the weight given to defense either. Compare, for example, Stan Hack and Scott Rolen.
CAWS…WAR…WAR/500PA…Peak7..OPS+…oWAR…WS-Off…dWAR….WS-Fld…Career
271……..52.6………3.1…………..29.4…..119……52.8……321.1…….1.4………66.0……….1.7………Stan Hack
262……..70.2………4.1…………..41.8…..122……52.8……307.2…..21.2………70.6……….1.7………Scott Rolen
Here are two guys who are equivalent in oWAR and career length (they’re within 8 PA of one another), but WS gives Hack a significant offensive edge. Moreover, independent of how oWAR and dWAR are combined, WAR rates Rolen as worlds ahead of Hack on defense, while WS shows them as only slightly different (on a per 162G rate, Rolen is 5.6 and Hack 5.5).
So the issue is not in the formula that adds oWAR to dWAR to produce bWAR — I haven’t even cited the bWAR difference. We need to drill down further to understand how such different results can emerge. Until we understand how the systems differ, we can’t assess how best to use them, whether to rely just on WS or bWAR, or to decide that each has its virtues and defects and we need to consider both angles when asssessing players (not to mention fWAR . . .).
I think this is not the string to do this work. I just wanted to flag it as an issue that the hybrid lists I posted on this string raise. (Besides, before I start spouting off I need time to go back and re-read “Win Shares”: I once actually understood that system, but I had many more neurons back then.)
Sorry — the link to Mike’s Blackmon article was dropped in the post above.
It’s pretty clear now that Traynor doesn’t belong – not only because we have the benefit of hindsight and distance, and not only because we can now view his numbers in the context of the high run scoring era he was a part of, but also because since he retired we’ve had a surfeit of great third basemen. But what about when he was actually voted in?
That was in 1948. At the time, 18 retired or active third basemen (defined as having played 70% or more of career games at 3B) had over 30 career WAR. Traynor, at 36.3, is 11th among those 18. Of the 10 above him:
Home Run Baker (62.8 WAR) – Massively underrated at the time, potentially due to his career ending right as the HR revolution began. Of course, the $100,000 infield was world famous, so it’s not like his achievements weren’t well known among fans and writers. The VC would rectify the mistake of the BBWAA and put him in the Hall in 1955.
Jimmy Collins (53.3 WAR) – Considered by acclamation the best third baseman of baseball’s first 80 or so years, Collins was already in the Hall by the time Pie got the call.
Stan Hack (52.6 WAR) – Probably would be in the Hall if his career started a bit closer to Pie’s. In 1948 he had only been retired for one year, though he never got much BBWAA support. I’m not certain, but there may have been some perception that he was a “wartime superstar” like Bill Nicholson or Snuffy Stirnweiss.
Larry Gardner (48.4 WAR) – A favorite of mine, and I know also a favorite of nsb’s as well. Probably not a Hall of Famer, but almost every winning team has a Larry Gardner among its ranks. Gardner of course was on four WS champs. Never outstanding but routinely very good – you could basically pencil him in for 4 WAR every year, and he was a heck of a gloveman.
Heinie Groh (48.2 WAR) – Undoubtedly hurt by timing – if he were Pie’s contemporary, he’d have likely been close to his .320 average.
Lave Cross (47.2 WAR) – Solid player with most of his value in the 19th century. Without great baseball references books, his HoF case rested on older fans singing his praises, and I guess the songs never got loud enough.
John McGraw (45.7 WAR) – Already in as a manager by the time of Pie’s election. Given his .334 average, probably had an outside chance of making it as a player.
Harlond Clift (39.2 WAR) – Good player, though most of his value comes from an early three year peak. Given the pre-occupations with average at the time, it’s understandable why few would have believed Clift and his .272 average were more vlauable than Pie and his .320.
Bill Bradley (37.2 WAR) – Part of some great Indians infields with Nap Lajoie and Terry Turner. A lot of his value is coming from defense, and the rest comes from a truly great three year peak in the AL’s salad days. Hit .317 with good power from 1902 to 1904, and outside of that was a slightly above average hitter. His career batting line certainly wouldn’t have impressed any voters in the 40’s.
Art Devlin (36.4 WAR) – Contemporary of Bradley’s, except with an even shorter career. Speed threat on the bases, but 20% of his oWAR comes from his 1906 season – his career batting numbers are pretty unspectacular. dWAR has him as a stupendous fielder, which is boosting him a lot.
If you were wondering, the 7 ranked below Traynor are:
Denny Lyons (35.5)
Willie Kamm (35.3)
Billy Nash (33.7)
Ken Keltner (33.1, but still active – would retire with 33.8)
Bill Joyce (31)
Harry Steinfeldt (30.7)
Arlie Latham (30.6)
When he was selected, was he a mistake? Perhaps. But not a huge mistake. The real mistakes were leaving off Stan Hack and not electing Baker much earlier.
This is a good approach, CC. We’ve been looking at the way players appear from the perspective of 2019, but your historical perspective is at least as valid an approach, and one that seemed more and more interesting to me as I read further into your post.
A note: when I wrote above that there were really no high quality fielders at third between Collins and Brooks, I stared at Gardner’s figures for a while. I decided to make the cut between Gardner’s 10.8 dWAR and Collins at 16.8, but Gardner was clearly very good. (Gardner’s career raises what I now think of as the Lou Whitaker issue: how valuable is it to a team to have a consistent and solid player over many years vs. a break-out superstar for a short peak. Doom has validly pointed out that very high WAR players seem to be critical for teams to reach championship status, but there are very high WAR players on teams that don’t win pennants, because those teams don’t have enough dependably solid long term members. When we think of the 2912 Red Sox or the 1920 Indians, it’s people like Wood and Speaker, Bagby and Speaker (again) we tend to think of first (well, I suppose Chapman comes first in ’20, for special reasons), but Gardner is a critical and rather unsung piece of those teams.
I’m interested in following up on how the players you name, many of which were well known even after their day, stack up when looked at more closely. I would be a great (and very large) project to develop a clearer view of the changing profiles of “greatness” at all the positions.
Bob Eno,
“The Glory of Their Times” was a series of interviews with stars from the first couple decades of the past century. Very interesting material that is non-quantative and totally anecdotal. Don’t recall if “Wahoo Sam” or “Smokey Joe” indicated if Speaker or Sisler were the equal of Cobb but, IIRC, it was mostly personal stories of their rise to the big leagues and experiences there
Thanks, Paul. On my shelf it has sat next to the Bible for forty years.
Bob, your point about Gardner/Lou Whitaker is an interesting one. My nominee for “that” guy would be Willie Randolph, 54 BWAR with Yankees over 13 years, added two more 4 WAR seasons, one with the Dodgers, the second with the Brewers, to close out his career. Peak WAR 6.6, lowest WAR in a full season, 2.8. Never considered a star. Named on 5 ballots his only year in HOF voting. Old fashioned player. 1243 lifetime BB, 675 K’s
Bill James was a notable Harlond Clift fan, ranking him 9th among all third baseman in his first (1985) Historical Abstract. Clift seemed to suffer the Cinderella midnight curse when his clock turned to age 30 in 1943 and he turned in easily the worst season of his career. The Brownies didn’t cut their erstwhile star much slack, trading him away in August to the second place Senators, who hardly used him the rest of that season. Clift lost almost the whole ’44 season to illness (mumps) and injury (thrown from a horse), then closed out his career in ’45 with a mediocre campaign much like his season of two years prior.
For his 8 year peak (age 22-29), Clift averaged about 4.5 WAR with 32 doubles, 19 HR, 86 RBI and 106 BB. He seemed to be quite adversely affected by the inferior wartime baseballs, with his doubles rising but his home run output suddenly plummeting to just 7 in 1942, and remaining in single digits the rest of his career. Using Bill James’ Favorite Toy, after his age 29 season Clift was projected to play 6½ more seasons and finish his career with 60.3 WAR; instead, it was 1½ seasons (244 games) and just 39.2 WAR.
From 1934 t0 1943 Clift led the Browns by far in WAR, 37.6 to runner-up George McQuinn’s 14.6. So he was the team’s star player but unfortunately he soured the year before their 1944 pennant. He was the first third-baseman to hit 30 HR in a season.
I love the discussions on this site, but I don’t get involved in the conversations, because I simply don’t understand enough of the advanced metrics to speak intelligently. Most likely, this is not going to sound too intelligent, either, but in the case of Fox, and to an extent, players like Aparicio and Evers, shouldn’t their HoF credentials be considered from an historical context.
I believe Bill James made a similar argument in his Hall of Fame book for Richie Ashburn, since he competed directly with Mays, Mantle and Snider, but was much better than the majority of CF’s from the ’30s and ’40s. Fox was arguably the best all-around 2nd baseman in baseball from 1951-60, but the game simply did not employ many players like Ernie Banks in the middle infield, generally, during that time.
I started following baseball as a kid about the time of Fox’ decline as a player, but in my opinion, the White Sox needed his type of player to be successful. They were quite successful during that decade, one pennant notwithstanding. I’d like to know if you all think someone like Jeff Kent, who is in the Hall of Stats, would have made them better?
Sounds intelligent to me, Gary. I think your point is closely related to Cursed Clevelander’s discussion of third basesmen, above, though not quite the same.
CC’s point, if I can paraphrase my understanding of it, is that we need to look at the Hall as an evolving record of what constituted greatness in the historical perspective of voters contemporary with their times, rather than as a constantly revised standard from the perspective of the present — which is constantly revising itself anyway. I really like that point, though it requires lots more work to figure out how to implement it in our reassessments.
Your point, if I understand it, is that we need to assess players according to multi-dimensional historical comparisons, one with the way players stack up against historical all-time greats and the other with the way they stack up within their own eras. (Thus Ashburn should be freed from the shadow of the New York City sluggers in center field during the ’50s, and Fox should be profiled against the relative dearth of offensive/defensive whizzes in the middle infield a few years later on.) These two angles seem contradictory, but I think it just means that an all-round perspective will, indeed, embrace a multiplicity of very different views. I think nsb’s argument for Billy Herman, below, actually fits very well with your post.
As for your question about Fox vs. Kent, I think the answer is that Kent would not have made the ChiSox better, but that this doesn’t mean he was not the better player. Chicago built its team around a profile unique for the era, almost a dead ball throwback. Fox and Aparicio were the keystone, so to speak. Kent was a different style of player and would have required the Sox to rebuild their approach. The odds of equally good success with a second profile don’t seem good to me.
In that sense, we might think that Fox and Aparicio both made the Hall, despite being borderline quality, much the way that Tinker-to-Evers-to-Chance did (with Fox, Tinker, and Chance right at the border, and Aparicio and Evers well below it). The success that marked them for consideration was really an expression of a multi-player unit, rather than the way any one member carried the team. I’ve never felt that the 1900’s Cubs jingle was such a scandalous basis for the Hall’s decision to admit the trio, and I feel the same way about the Go-Go Sox core duo. If the Hall is about individual greats, then those decisions may be wrong in part or in full (and that’s the model I use myself). But if the Hall is about the history of greatness in baseball, perhaps there ought to be room for some historically memorable couplings. If you look at the Cubs trio, all elected together, the plaques of Tinker and Evers specifically cite the Adams jingle formula, while Chance, their manager, whose accomplishments were broader, is treated as an individual. Luis and Nellie were elected separately and over a decade apart by different processes, but if you look at the actual voting, they were quite close when Luis was elected in ’84, and the following year Fox missed by only 0.3%, indicating a strong sense by voters that the two belonged together. But given Aparicio’s solo status in ’85, neither man’s plaque notes the other.
And I think a related point that we’ve touched on before, though I believe it was in the context of the CoG, is how we balance the idea of absolute benchmarks for induction and relative benchmarks for induction. That’s sort of the Ashburn question – is “well, he was only the 4th best CF during his career” a knock on somebody’s HoF candidacy? It might be – but when the competition is two of the 10-15 best players ever and another strong member, it’s not really a fair thing to hold against him. I believe similar arguments popped up about the 90’s pitchers – when you get to people like Brown and Schilling and Mussina, and even beyond that to somebody like Appier, you can start saying “This guy wasn’t even one of the 10 best pitchers during his career – how could be possible be worthy?” But again, the guys above him are inner circle or close to it – it was a period that had a surfeit of excellent pitchers. (I believe Bill James also noted something I intuitively agree with but do not know if there is a statistical proof of, namely that a high offense era like the 90’s allows the truly great to shine brighter because the average and below avarage pitchers are skewing the numbers in one direction, so those who excel look even more fantastic by league adjusted measures like ERA+)
The opposite case exists, where we might have a period of 10 or 15 years where almost no players or a certain position have been elected, and one can argue that since John Doe was the 2nd best third baseman of this decade-plus stretch, that’s a good argument for induction. But if his absolute benchmarks are low – if his pure numbers, even taken into proper context of the league (maybe it’s a low scoring era), just seem too poor, do we need to represent every period with X amount of players from each position? My view is that of course we don’t, but we should also make sure we’re not falling into context traps. If a 15 year period produced no Hall of Famers at a position, I think it would be worth really poring over the numbers and making sure we’re not missing some nuance that makes someone a more palatable selection. But if we end up concluding that no one fits the bill, I think there’s no problem with that conclusion, as long as it’s arrived at after careful scrutiny.
Great points, CC, and I especially like your comment about how in a high scoring era the greatest pitchers have a chance to stand out even further (and vice versa with hitters?). How would we set up a test for this, and what would positive results imply in terms of assessing players across eras? (This is related to, but not the same as, the knock on earlier baseball greats, which discounts their numbers because the general level of play was low, and the counter-knock on PED players, which argues that the general level of play in their time was artificially high.)
A lot to sort through. And your earlier comment had already lengthened the agenda . . .
Gary, I’m glad you raised this. I’ve never been able to completely understand how we compare players from different eras when they were playing in different environments. Obviously, there’s the difference between the Dead Ball Era, and pre-integration, and the War Years, where we have come to some sort of consensus, albeit occasionally a forced one. But I still don’t think I see how we reconcile differing approaches to game management between, say, the three true outcome modern environment and the put-the-ball-in-play-move-the runner-up. We’ve had 30 seasons of 190+just since 2004. Nellie Fox struck out 216 time in his entire career. This wasn’t just because of incredible bat control or even hard throwers, it was an approach to the game that making contact, even if it resulted in an out, could move runners around the bases and maybe score runs. Another incredible Nellie Fox stat–in 80 career PA’s against Bob Feller, he never struck out.
So, how do we adjust for players whose stats reflected, in part, the way their managers and then-conventional wisdom said they should play?
Someone in this thread, probably Bob Eno, has strongly suggested that Billy Herman doesn’t belong in the HOF. But to me, he is the one player in the group whose credentials are worthy without question. Not an inner circle HOFer, no, but a sizable cut above anyone else, even Bobby Doerr, for whom Doug put in a good word in an earlier post.
The first thing that should be noted about Herman is the fact that he lost two productive years to WW II. How productive? Well, in 1943 he led the Dodger position players with 4.7 WAR and an OPS+ of 135. In his 1946 return, playing 122 games for the Dodgers and Braves, he put up 3.8 WAR with an OPS+ of 128. It’s hard to believe that he wouldn’t have garnered another at least another 6-7 WAR in the seasons he missed. Or put it this way: the premiss of the discussion here is about players in the HOF who fail to make the Hall of Stats 100 rating cutoff. Had Herman played those two seasons he wouldn’t be on the list above, since his HOS score of 99 is already just a point away.
How good was Herman? No power, but he hit for average with lots of two-baggers. He was the best second baseman in the NL without question in the 1930s and arguably the best position player on the near dynasty 1930s Cubs as well, plus a key acquisition for the Dodgers in their 1941pennant run. A ten-time All Star. Finished in the top five in MVP voting three times. Drove in 100 runs in 1943 with only 2 home runs, not an argument for HOF inclusion, but an interesting fact. Is this a live-ball era record—fewest HRs in a 100 RBI season? Maybe someone with time on his or her hands and more statistical chops than I have can find out.
I am indeed the guilty party, nsb, but my suggestion was not “strong,” though the word was in there. I concluded a joint assessment of Herman and Doerr with this summary: “Herman is just behind Doerr in a number of key categories (though not by CAWS), but lost more time to the War. They seem even to me — perhaps it’s my small Hall bias, but I don’t feel the Hall is stronger by including them and I would not.” If your Hall is bigger than mine, Herman would be over the line. The Hall of Stats leaves the line where it is and shuffles players across it both directions, and Herman would probably have made it in with credit for his War years, as you suggest, though we’d need to look at where he’d stand if all others sitting outside got their wartime credit too (and, perhaps, everyone who played in 1943-45 were docked a bit for accomplishments abetted by a lower-quality competitive environment).
I’d pick a few small bones with your points. The 1929-38 Cubs weren’t really a near dynasty: they were a competitive team that won pennants every three years like clockwork (and Herman, who missed 1929, added 1941 on the Dodgers to keep his clock on time — if I recall, Durocher thought the trade for Herman was the key to that ’41 pennant, and that makes Herman a foundation of the Branch Rickey Dodger dynasty). You could stretch it and say that they were in dynastic rotation with the Cards and Giants — from 1928 to 1938 those three teams won all the NL pennants by turn. Herman was indeed perhaps the Cubs’ top position player; others like Frisch and Ott were much better in that rotating line-up of teams; Herman was not in their class, but he was in a second, very good class, and may have led it.
On missed wartime years I think we’ve got to be a bit careful in making estimates, in line with my comment in the first paragraph. Baseball was mildly depleted in 1942 at about the scale it was in the first postwar year, 1946. Wartime baseball was essentially 1943-45, when players were increasingly drained away by the draft and on-field quality diminished. Herman and Doerr both lost time to the War and deserve credit, but in assessing the level of credit by examining the pre-/post- years, there does need to be some discount if the pre- year falls during the period 1943-44, when normally strong players had an advantage over their emergency-replacement competitors. Herman’s 1943 WAR needs to be discounted to some degree, probably guided by his performance the previous two years, where he was showing normal career decline. He does, however, redeem some of that with his strong comeback in ’46 as a semi-regular, before retiring a few games into 1947.
Herman’s 100 RBI on 2 HR in 1943 tied a lively-ball era record, of sorts: Pie Traynor’s 103 RBI on 2 HR in 1931. Traynor managed 124 RBI three years earlier, but needed to slug no fewer than 3 HR to get there. (My statistical chops reflect my past willingness to fork up a few bucks to use the B-R Play Index.)
Bottom line: I don’t think Herman’s inclusion strengthens the Hall, but I don’t think it weakens it — he’s a borderline candidate, and therefore a very good player, whose qualifications will look different depending on your image of what the Hall should be. I do see Herman and Doerr as essentially of equal value, and think it would be splitting hairs beyond what attempted objectivity can claim to accomplish to divide the pair.
With the PI, I can indeed confirm that Herman and Traynor are the only live ball players to satisfy RBI > or equal to 100 and HR < or equal to 2.
Ed Delahanty had 2 HR and 109 RBI in 1900, and Lave Cross had 108 RBI and 0 HR in 1902. Everybody else who fits these criteria was in the 19th century.
On a not-quite unrelated note, Scott Sanderson just passed away at the age of 62.
How do you evaluate a player like Sanderson? He was in the starting rotations of four division winners, was injured a lot, put up wins despite a mediocre ERA+ in most seasons, went 16-10 for the dismal ’91 Yankees (71-91) with 3.9 WAR. Can all this just be ascribed to “luck”?
Statistics aside, I think that some pitchers win more than expected in spite of their mediocrity, and Sanderson is a good example. Do their teammates rise to the challenge? Who knows. Other pitchers lose more than one would expect, considering their capabilities, a good example being Sanderson’s contemporary, Dave Stieb. You’ll have to ancient to get the reference, but when I think of Stieb, I think of Joe Btfsplk.
N SB
” I think of Joe Btfsplk.”
Nicknamed “Eyechart”?
Sorry to hear about Sanderson, nsb.
I’m not sure there’s too much of a discrepancy between Sanderson’s performance and his game outcomes. Sanderson’s career ERA+ was 102 and his W-L Pct. was .533. They don’t seem too out of line to me, though obviously the W-L is a bit better.
In ’91, a lot of the Yankees’ dismal quality was lodged in the remainder of the pitching staff, which had a 4.52 ERA vs. a league norm of 4.09 (with Sanderson at 3.81). The team didn’t have anyone else close to a qualifier apart from Sanderson (it was a really terrible staff). 16-10 is probably within normal range for Sanderson’s 109 ERA+, and it’s a little low, I’d think, for a 3.9 WAR season. It’s true, though, that Sanderson had pretty good run support that year. I think his real year of good fortune was 1990, when he went 17-11 — virtually the same as the next season with the Yankees — for a very strong Oakland team, while compiling an ERA+ of 95. For this B-R awards him 0.3 WAR. Things average out: Sanderson earns 2.1 WAR/season for his two-year 33-21 stretch.
I didn’t know Joe Btfsplk, but then, our home newspaper didn’t carry comics; I knew Capp’s strip mostly from a recording of the B’way musical. We see eye to eye when it comes to Stieb.
nsb
I am ancient and I think I got the reference. Wasn’t Joe the guy perpetually under the rain cloud? What cartoon series was that?
L’il Abner.
FWIW,
Matt Chapman, 3B Oakland, was on pace to surpass the all-time dWAR single season record for the hot corner before yesterday’s game. He was at 0.5 after 16 games and now, after 17 games, he’s at 0.4….He must have more moves than El Cordobes
I still think Chapman’s special sauce is the Coliseum’s spacious foul grounds, largest in the Majors.
What stands out most in Chapman’s fielding stats is the high proportion of putouts among his chances (37% — compare, e.g., Nettles’ career 25%; Brooks’s 29%).
Chapman in the Coliseum has recorded 45% of his chances on putouts; 24% away. His 17 home putouts include 8 foul flies caught — 7 in only two games (3/28 & 4/3): that’s 21% of his total chances at the Coliseum just on foul outs.
Bob
Yes. I can not disagree with you but shouldn’t dWAR emphasize assists by a 3B over putouts? Sure, it’s great that “Chapman’s ranging under it” but, how many times did he call off the SS or even the LF to catch that “can of corn”? In this day and age of MLB, it seems unfathomable to me that Chapman’s athleticism exceeds all the other guys at 3B by that much. I’d certainly be much more impressed if his assists count was that much more higher than his peers- no kidding.
Paul, I wasn’t defending Chapman’s high rating, just trying to explain it. I don’t think he should be receiving credit for the size of Oakland’s foul grounds. On the other hand, if he’s exploiting it better than previous home team third basemen, that’s to his credit.
As for putouts and assists, I don’t know the answer for which should be seen as more valuable. Foul outs are unassisted; a force out at third assisted by a throw is not that much different from a first baseman’s putout; a tag out on an outfield assist may deserve greater credit. And assists may likewise vary from routine to circus-quality. BIS and Rdrs should be tracking all these distinctions (including whether a foul out is a can of corn or a leap into acrobatics) and assessing them in terms of both difficulty and RE24 factors.
If you’d be impressed were Chapman leading his peers by a lot in assists, then be impressed: as of this morning Chapman has 39; runner up Machado has 34; no one else in MLB is above 30.
Bob
Haven’t the As played more game than the rest of MLB due to the Japan series with Seattle?
Hadn’t thought of that. Machado’s Padres have played three fewer games. Others have played fewer too, but Chapman’s lead in assists is far greater.
But Paul, I’m trying to figure out where this is going. You noted Chapman’s figures were very good and I suggested a reason why he might not be quite as good as he looks in the figures (the Coliseum). You raised the issue of assists vs. putouts (not sure why), and it turns out, Machado possibly excepted, Chapman is well ahead of others. I’m not boosting Chapman, just looking at the stats you brought up and finding their basis. Is there more to say?
Bob,
Sorry, my remarks concerning his high dWAR totals were in disbelief and sarcasm. That being said, I believe the emphasis for evaluating a 3B’s fielding prowess should be on assists.That’s it – plain and simple…. it’s much more difficult to make the rather routine 5-3 than the routine pop-up foul ball putout. Chapman’s putout totals are a product of the foul territory at Oakland Coliseum – as you’ve pointed out. At the risk of offending anyone, I’d also like to see a few 150 OPS+ seasons before anointing him the second coming of Schmidt……or even Santo. I believe even Sal Bando had a few 150+ OPS seasons at Oakland.
Maybe that’s the dWAR distrust in me-I dunno. I believe Carlos Gomez recently had a high overall WAR total with the lowest oWAR total of any CFer with that much WAR in a single season
Well, Paul, I don’t agree with you on assists — routine plays are routine; my interest would be in above average abilities to make plays of either type that are exceptional — but since I have no ax to grind with regard to Chapman either way I’m happy to give you the point. I read your initial remark as disbelief but heard no sarcasm.
Speaking of plays at third, have you checked out Machado’s routine-burger on Friday?
Bob
Yes. And half of MLB 3b make that play but certainly not in as effortless fashion
Since we’re in a sort of holding pattern, I wonder whether I can toss out a general question related to dWAR.
Neither B-R nor WS on Baseball Gauge assign fielding credit to pitchers (in terms of dWAR or Fld). B-R assigns pitchers a positional adjustment, which is clearly used in oWAR calculations for them. But there are, of course, good and bad fielding pitchers: since B-R assigns 0 Rfield and dWAR to all pitchers, I’m having trouble finding where the value of their defensive contributions is reflected. In it’s discussion of positional adjustments, B-R states, under “Pitcher Positional Adjustment,” “Since pitcher fielding is included in Pitcher WAR, we do not need to consider it here.” I can’t however, figure out where we see pitcher fielding included in Pitcher WAR.
Does anyone know how this works?
It must be counted separately in B-R’s model. It would make no sense to separate out fielding by positional opportunity and not deal with the pitcher.
Well, B-R’s system builds positional adjustment into offense as well as defense (so a pitcher who bats .240 with 2 HRs counts as a good hitter), so the rationale for calculating the adjustment for pitchers exists independent of fielding. Like you, I assumed the adjustment’s defensive component would be counted separately somewhere, but I sure can’t find it. If you can, please let me know.