Billy Nash:

November 14, 1895: Traded by the Boston Beaneaters to the Philadelphia Phillies for Billy Hamilton

Is that right? Straight up Sliding Billy? Nash was only a year older than Hamilton, but it turned out he was almost done and Hamilton had a hand in the Beaneaters championships in 97-98 (moving Duffy from CF to LF).

Anyhow, the Phillies 'GM' must have thought highly of Nash.

Is that right? Straight up Sliding Billy?

While Nash had just had a fine season for the Beaneaters, there's no way that I would have made that trade straight up (even accounting for the differing fielding responsibilities back then for third base and centerfield).

Obviously the Phils wanted Nash as their captain. After all, he had captained Boston to 3 straight pennants. But he didn't have the horses in Philly.

Obviously the Phils wanted Nash as their captain. After all, he had captained Boston to 3 straight pennants. But he didn't have the horses in Philly.

So did he lead the Phils on the field at all, Cliff? I have never seen it mentioned about him, but they didn't mention him at the helm for 1893-1895 in Boston, either. ;-)

Holy crap! This thread is 2 years old and no posts. Well, ok, one.

Denny Lyons would be an obvious addition to this group BTW.

These are raw (Bill James) unadjusted WS for the '90s guys (assuming that NOT normalizing to 154 or 162 games will hurt them all about the same, whichis not quite an accurate assumption. The guys [like Nash] who played in the '80s are hurt a little more.)

Lave Cross 278/26-22-21-19-17-17-17-16-16-16-16-15-14-11 (14 years ≥ 10 WS, 1891-1906)
Billy Nash 222/25-23-22-20-20-20-19-17-17-15 (10 years, 1886-95)
John McGraw 207/34-31-24-21-20-20-20-16 (8 years, 1893-1901)
Denny Lyons 190/27-27-25-23-21-20-16-14-10 (9 years, 1887-96)
Billy Shindle 167/27-26-16-15-15-15-14-13-12 (9 years, 1888-1897)
Bill Joyce 157/29-25-18-18-18-17-16-16 (8 years, 1890-98)

These numbers are of limited use, actually, because of differing season lengths but the best seasons, other than McGraw's, are probably Lyons', because some of his best are in the '80s in short seasons. At his best he was better than Nash at his best, I think.

OPS+

Lyons 138/193-62-57-47-37-31-30-23-12 (in his 9 years of ≥ 10 WS, about 5000 PAs)
Nash 103/134-23-12-9-7-6 (otherwise < 100 OPS+, about 6650 PAs)

This would certainly support the idea that Lyons was better than Nash at his best, but obviously Nash had the equivalent of about 3 additional seasons. And WS has Nash as an A glove, McGraw B+ and Lyons C-. Cross was of course an A+, Shindle an A and Joyce an F. (Yes, that's an F. It's the only one given to a 3B, I don't remember ever seeing one before.)

Mainly I wanted to compare Nash and Lyons. Here are the other guys, including the '80s guys since the short seasons don't skew the OPS+ in the way they skew a cumulative number like WS.

Joyce 144/179-79-62-50-44-35-27-3 (iin his 8 years of ≥ 10 WS, about 4000 PAs)
McGraw 135/165-57-48-31-29-26-10-7 (in his 8 years of ≥ 10 WS, about 4700 PAs [starting 7 years after Nash])
Williamson 112/149-48*-30-26-16-11-7-1-0 (*148 is adj. downward due to short porch in 1884, about 5000 PAs)
Cross 100/135-31-21-20-13-12-9-0 (8 years ≥ 100 OPS+ in 14 years of ≥ 10 WS, about 9600 PAs)
Shindle 88/124-16 (only 2 years > 100 OPS+, in about 6200 PAs)

McGraw's OBA and/or OPS+ is not that spectacular when compared to his contemporary Bill Joyce, though of course Joyce was one of the worst fielders in ML history. But get this:

Lyons beats McGraw head-to-head in OPS+ in all 8 years that McGraw earned ≥ 10 WS including 191-165 at the absolute peak. That's a big advantage to Lyons. Nash of course was a better fielder than Lyons and played those 3 extra years (equivalent). Still, as a peak voter, it is hard to prefer Nash. And as any kind of voter, this certainly puts the McGraw mystique in perspective.

Holy crap! Between the time I started my post and finished it, 4 posts beat me to the punch!

Marc, are you taling into account league strength with Lyons?

BTW, I have set up a thread for him.

I also have have Lyons as the best ML third baseman only once (1887), while I have Nash as the best numerous times (1888, 1890, 1892, and 1893).

Lyons also gets an unadjusted league strength C- from Win Shares, while Nash gets a sterling A. Lyons most likely would have had (at least) a D rating if he had played in the NL and PL like Nash did.

Good point, John. That 193 OPS+ came in the 1890 AA. My bad.

The 162 came in 1887 however, when the AA wasn't so bad. Unfortunately he didn't play when the AA was at its peak in '85-'86.

And he was at 130-131 in the NL in '92-'93, still only age 26-27. (He came up at age 19 and became a regular at age 21. His last decent year was age 30 and his last year period was age 31.)

The progression was:

AA (getting progressively weaker): 162-47-57-93-37
NL: 130-31-12-(99*)-23-(90*)

* Not as a regular.

So the 162-47-57 in 1887-88-89 are probably fairly good indications of his real level of offensive play (fairly meaning within 5-8 percent). But clearly the 193 is 25 percent too high (193 X .75 = 145 as a very gross estimate). The 137 is too high, too, though that must be seen as an off year because he came back at 130-31 in the NL. So adjusted the following seems fair for his years as a regular:

Denny Lyons 154-40-44-45-16-30-31-12-23
Ken Boyer 139-32-29-27-23-21-21-13-1- (+3 years < 100, obviously a better fielder)
Bob Elliott 112-5-23-19*-19*-3*-(99)-48-45-38-43-28-0 (WWII years discounted 10 percent, clearly on steroids after the war)

Not sure he shouldn't be a candidate. I am quite sure he is better than McGraw (see the McGraw thread), not sure he is better than Nash. People at the time seemed to like Nash. And the AA competitive discounts may not be deep enough, I don't know. Too late for him in any event, but a man worthy of being remembered.

Oops, thought I was on the Lyons thread there.

Yes, Grandma, Nash led the Phils in 1896 and the first half or so of 1897.

Here BTW are the consensus 3Bs.

5. Joe Sewell ;-)

14. Ken Boyer--23 ballots, no top 3
32. Tommy Leach--12 and 0
37. Bob Elliott--7 and 0
39. John McGraw--4 and 2
40. Pie Traynor--4 and 1
52T. Ed Williamson--4 and 0
63. Al Rosen--3 and 0
85. Bus Clarkson--1 and 0

My top 3Bs vs. the consensus

16a. Stan Hack
22. Ed Williamson (+30)
23. Ken Boyer (-9)
42. Bus Clarkson (+43)
50. Pie Traynor (-10)
57. Al Rosen (+6)
64. Bob Elliott (-32)
77. John McGraw (-38)
77a. Ezra Sutton

Swap Williamson and Clarkson for Elliott and McGraw and doing OK.

Lyons beats McGraw head-to-head in OPS+ in all 8 years that McGraw earned ≥ 10 WS including 191-165 at the absolute peak. That's a big advantage to Lyons.

Whoa, now, let's re-analyze here. I have McGraw ahead of Lyons in offense EVERY year, even before taking league strength into consideration:

On the Lyons thread, John mentioned that Lyons had played the first half of his career in the AA, which I had overlooked. Note that I have McGraw in my top 100 and Lyons not, though he is in my Honorable Mention group just behind that (and Nash is not). Still Lyons' first 2 NL years were OPS+ 130-131. If we discount his AA years they probably come in in the 130-140 range on average, so the guy could hit. McGraw is a lot like an Al Rosen or Ralph Kiner--stratospheric offense for 3-4 years but very much down to earth otherwise (e.g. OPS+ 110-120-130), except that his peak is not all that stratospheric. If this thread says anything, it is that McGraw, while probably the best, is not as far ahead of some of his contemporaries as we might suppose

And I might add that Ed Williamson, who is in my PHoM, is not so far ahead of them either. What is it about 3B that makes it so hard for a guy to excel, I mean really excel? And doesn't all of this make HR Baker look so much better?

Meyerle's better than any of them. 241 OPS+ in 1871 makes Norm Cash look like a piker!

Anyone know if he had a MiL career after retiring from the majors?

I'm pretty sure that in an earlier discussion of the International Association, it was mentioned that Meyerle played in it after leaving the NL (part of the argument that the IA was possibly the stronger organization 1877-79). If I recall correctly, Ned Williamson and Fred Dunlap both played in the IA also before moving to the NL.

Does anyone have data for the IA years of Meyerle, Williamson, and Dunlap? This thread would be a good place for it to go.

Yes, Grandma, Nash led the Phils in 1896 and the first half or so of 1897.

Thanks, Cliff!

If this thread says anything, it is that McGraw, while probably the best, is not as far ahead of some of his contemporaries as we might suppose

If McGraw had stayed healthy, he would be the king of 19th century third basemen. But...

Meyerle's better than any of them. 241 OPS+ in 1871 makes Norm Cash look like a piker!

There's no way that he would have had a 241 OPS in a 154 or 162-game schedule. The man who could hit, but he wasn't Babe Ruth. :-) Some adjustment needs to be done.

With that said, I used to have him on my ballot many moons ago. In fact, did anyone else have him on their ballots? I seem to remember you having him, karlmagnus, but I'm not 100% sure about that.

I had him at one stage quite high on my ballot; he was downgraded after the pre-ML research showed that he was nothing special in 1869-70 -- that made him a pretty short career, shorter than Pike (though Meyerle was clearly better.) If he did a lot in the late 70s IL, given that we've been more open to MiL credit than we used to be, he needs a medium sized revaluation which would put him back on the ballot, as he's currently about #40 and the ground is pretty flat after #10 or so.

Of course, if Williamson had the same (productive IL years) then he would stay ahead of Meyerle. Ditto Dunlap. If somebody knows where to get those numbers let us know. These are three interesting players, to say the least.

Long ago, I described Meyerle's injury (broken leg?) in reply to someone's (charlemagne's?) ballot comments.

--
1897-04-19, Monday
Opening game, Philadelphia at Boston (not part of the opening series beginning Thursday for all eight teams)

"The Boston-Philadelphia game of April 19 was played on that date in order to take advantage of "Concord Day" [now Patriot's Day], a holiday special to Massachusetts, having taken the place of Fast Day several years ago by special act of the Legislature."

Before the game, there were presentations at home plate to Captain Duffy and ex-Captain Nash.

Source: Sporting Life 1897-04-24

Of course, if Williamson had the same (productive IL years) then he would stay ahead of Meyerle. Ditto Dunlap. If somebody knows where to get those numbers let us know. These are three interesting players, to say the least.

Decent seasons from the IL for Williamson could make him a phoenix on my ballot, Marc.

Speaking of the International Association, does anyone know where to find reasonably complete stats for the 1877-1879 years? I have Nemec's 19th C. Encyclopedia, but the IA stats are not in there.
Thanks

mulder, Are you still interested in IA playing statistics? If so, what is your mailing address?

--
DL from MN, 16 May 2007
"Carlos Moran" #17
If I go with the MLEs [Moran] ends up in the 40s. That put him near Lave Cross and prompted me to doublecheck [Cross's] numbers. I hadn't been giving Cross any credit as a catcher. If I give Cross that credit he jumps up above Dunlap, Leach, Ben Taylor and the CF glut to show up as the 2nd most deserving pre-1930s player (behind Cravath). Anyone want to talk me down off that ledge?

Lave Cross may be the second (or first or fifteenth) most deserving player who remains from before 1930. His career is long enough (longest?), and there is enough uncertainty in our understanding of 3B and C defense, to underwrite that wide range.

For seven seasons age 21-27, Cross was his team's second catcher, clearly second, 37-55 games at the position. In the last three of those seasons he was a more or less regular player, moving to more demanding positions in the field when not catching.
OF : 3B : SS
43 : 24 : 1 (1891)
25 : 65 : 5 (1892)
10 : 30 : 10 (1893)
Then in 1894 only 16 games as catcher and
0 : 100 : 7 (1894)

1895-1897, mainly 3B with significant SS and 2B play, 1400 plate appearances at OPS+ 74, thirty points below his otherwise career average 104. That following OPS+ 129 in 1894, the highest-scoring season.

1898-1906, age 32-40, he was everyday 3Bman, average 141 games except 1901.

Tough job, part of 1899, leading the Cleveland Spiders.

Around 1900, the Cleveland Press (adopted hometown newspaper) considered him a great fielder, maybe the greatest.

Win Shares gives Cross only two gold gloves, 1895 and 1902, compared with Billy Nash six to 1894 and Jimmy Collins five to 1903. At 6.03 Win Shares per 1000 estimated innings, Cross with about 15000 innings beats everyone down to Lee Tannehill at about 6000 innings, then Bobby Wallace at about 3600 innings.

Among 171 players with 5000 estimated innings at 3B, the leaders in Win Shares per 1000 all played during the 1880s-1910s. Among the twelve rate leaders down to 3000 innings, listed here, only one played in another era (Mark Christman, one of the anonymous 1944 St Louis Browns).

Third Base fielding Win Shares (rate per 1000 estimated innings)
fWS/1000 inn ; inn (1000s, nearest 0.5)
6.58 Lee Tannehill (6)
6.03 Lave Cross (15)
5.93 Jimmy Collins (15)
5.76 Tommy Leach (8.5)
5.67 Art Devlin (10.5)
5.51 Billy Nash (13)
5.49 Ned Williamson (6.5)

Down to 3000 innings (290 players down to Pedro Guerrero), the other rates in this range are 5.69 Charlie Reilly, 5.86 Buck Herzog, 5.64 Mark Christman, 6.73 Bobby Wallace, 6.59 Billy Clingman.

Among the twelve named here, Wallace moved from 3B to SS at age 25 (1899); Williamson, Clingman, Tannehill, and Herzog at age 28 (1886, 1898, 1909, 1914); Christman at age 32 (1946).

Cross moved from C to 3B, described above. Leach moved to CF. Collins, Devlin, Nash, and Reilly played 3B essentially throughout their major league careers.

Innings played at SS
16046 Wallace, #22 in SS innings, estimated
12058 George Davis (another contemporary ex-3B, for reference)
_4013 Herzog
_3925 Williamson
_3261 Clingman
_3223 Tannehill
_3048 Christman

That is among 290 shortstops with 3000 innings played at position.

Honus Wagner moved to SS where he played 16971 innings est., from 3B (fewer than 3000 innings) and other positions. By Win Shares, Tannehill and Wagner were great fielding shortstops; Wallace, Davis, Herzog, and Clingman were good ones, Herzog maybe a notch better.

This thread is about John McGraw too, another 3B contemporary to many of those listed here.
5.02 John McGraw (6.5)
That is still a good rating although there are several others around there, of course.

What about Meyerle? I took a renewed look at him and he looks good, even ballotable. I want to know more before taking the plunge.

If I recall correctly Meyerle gave up almost as many runs with the glove in 1871 as he contributed with the stick.

One warning about Meyrele's times: BA/OBP/SLG alone do not come close to fully explaining run production, due to the very high prevalence of reaching on error. Clearly, many more of Meyerle's balls in play were scored as hits in comparison to his teammates, but he scored only about the same number of runs as did various teammates. I don't honestly know what to make of that - but that, and the small samples, and the extreme team quality differences, tend to make me cautious.

OCF may be onto something but it one should check every season.

I think I recall seeing that he batted eighth. (Should I use an exclamation mark here?)
--in one game or regularly? His numbers of AB and BB support batting eighth regularly.

player.. _G PA
Malone.. 27 143
Fisler.... 28 150
Reach... 26 138
Meyerle. 26 132
Radcliffe 28 151
Cuthbert 28 160
Sens'er. 25 127
(Heubel. 17 .77)
McBride 25 139

The batting order was not quite fixed but exceptions were few. Among the eight everyday players, Meyerle and Sensenderfer alone "batted" (AB+BB)/G only about 5 times per game.

Just cross-posting my timeless case for McGraw from another thread...

My contention (as always) is that Win Shares and BP WARP just get him terribly, terribly wrong. The reasons are:

1. Win Shares and BP WARP both use replacement levels that are far below the empirically demonstrated value of freely available talent. This leads them to grossly over-reward "just showing up" at the expense of actual excellence on the field. If you correct this flaw (by subtracting around .015 WS or .004 BP WARP per PA) from everyone, McGraw's peak looks much more impressive.

2. The run estimators in Win Shares and BP WARP are too inflexible to capture McGraw's value in context. In higher run-scoring environments, the relative value of OBP to SLG goes up, while in lower run-scoring environments it goes down. (The easiest way to understand this is to think of a hypothetical league in which no outs are ever made--in which case a single is just as good as a HR, since all runners will score eventually--and then compare it to a hypothetical league in which no runners ever reach base, in which case the only hit with any value is a HR). BP whiffs on this because it appears to first do the UEQR-EQR translation, moving from the actual run environment to a 4.5 RPG environment, and *then* calculating BRAA, which changes the win value of out avoidance. I don't have the Win Shares book so I don't know what RC formula he's using for the 1890s, but eyeballing it he systematically understates the value of OBP in that era. If a WS expert can walk me through the process of calculating BWS in that era, I should be able to spot the error. Anyways, the point is that McGraw's greatest skill--out avoidance--was worth more in his time than in any other, and neither system gives him full credit for it.

3. Neither system (as far as I can tell) properly credits him for his speed. BP appears to use a flat 75% assumed success rate for seasons where CS was not available, and again I don't have WS's 1890s run estimator handy but I'd be stunned if it did this correctly. The point is that there is an extremely strong (logarithmic) relationship between a player's SB attempt rate and his success rate. If you apply any sort of linear value to SB when you don't have CS data (either a flat assumed success rate or a flat run value), then you are overrating guys with middling SB and underrating guys with very high SB, since in general guys with middling SB tend to have break-even or worse success rates whereas guys with very high SB tend to have terrific success rates. McGraw's extremely high SB totals suggest he should also have had one of the best SB percentages of his era, which means they should be credited more than someone who had only 20-30 SB per year.

4. Neither system recognizes the changing depth of 3B over time. BP uses the same ratio of 3B to SS FRAR in 2007 as in 1897 (about 55%), and Win Shares only increases it by 8% (from 67% to 75%), representing a difference of just 0.6 Win Shares/0.2 wins. In fact, the gap between replacement third basemen and replacement shortstops grew from 0.6 wins in the 1890s to 1.7 wins today. This leads these systems to further understate McGraw's value (and that of all 1890s 3B).

>Anyways, the point is that McGraw's greatest skill--out avoidance--was worth more in his time than in any other, and neither system gives him full credit for it.

Unfortunately, some of those outs he avoided by sitting on the bench. I could avoid those outs.

>This leads these systems to further understate McGraw's value (and that of all 1890s 3B).

Yes,then there's the issue of all those 1890s 3B. McGraw is not that freakishly unique in an environment where Bill Joyce is excelling along with Lyons and Nash and Latham and others.

Sunnyday, you know very well what I'm talking about. He is underrated by OPS+ or EqA or whatever you want because they don't properly weight the contextual value of his OBP, regardless of what replacement level you set to determine the relative importance of rate to playing time.

I don't know off the top of my head who "all those 1890s 3B" were. Joyce was an atrocious fielder; none of Lyons, Nash, or Latham hit much in the 1890s. The only one we've elected is Jimmy Collins. Regardless, the point is that a lot of teams were playing really terrible guys at 3B back then--just take a look at, say, Bill Gray or Billy Shindle in 1898, when McGraw was tearing up the league.

OCF, if you're reading this, could I trouble you to post one of your RCAA charts for the following players?

John McGraw
Bob Johnson
Atanasio Pérez
Kirby Puckett
Hugh Duffy
George Van Haltren
Ken Singleton

Thanks very much.

McGraw's extremely high SB totals suggest he should also have had one of the best SB percentages of his era, which means they should be credited more than someone who had only 20-30 SB per year.

Not necessarily. It could simply be (probably is) a reflection of Ned Hanlon's style, where *everyone* stole bases. The Orioles were typically well ahead of every other team in the league in SB during McGraw's years as a regular, and up-and-down in other speed-based indicators. I don't know that you can extrapolate CS in that era as well as you can in other eras.

-- MWE

I just saw Dan's post - a pretty motley list, that. Different positions, different times, different strengths and weaknesses, different peak or career tendencies. The only unifying factors are that they're high backlog and have offense as a large fraction of their value.

All of the usual caveats apply. This is based on the numbers from a Stats Handbook. Different RC formulas are used for different time periods. You can say that that cast doubts on comparisons between 1890's players and 1980's players - but how can you cast new doubt on something that was already that doubtful anyway?

These are not adjusted for season length in any way, or for league strength, or for position (RCAA, not RCAP) - with the one exception that I did slightly discount Johnson's leagues for WWII depletion. (Else his 59 would have been a 69). The measure has proved relatively insensitive to playing time, so I have personally felt the need to separately deal with and account for playing time. That's why I've never supported McGraw (or Frank Chance, whose offensive peak is even higher than McGraw's). At the same time, I do understand what someone like KJOK is doing when he does support McGraw and Chance.

Anyway, with all that said, here's your table:



Van Haltren's -13 was in the first year of his career, when he was a pitcher. Singleton's -25 was in the last year of his career: he suddenly got old, and wasn't removed from the lineup until the damage had been done.

OCF, thanks very much for posting that. I requested it for precisely the reason you described: If we're going to elect one more Professional Hitter™, which one should it be?

Is there any chance you could do a season length-adjusted version? Or just tell me which seasons are which (email cooberp@gmail.com), and I'll repost with that info? It's kind of hard to assess when you've got 130-game seasons from McGraw, Duffy, and Van Haltren mixed in with 162's.

I really appreciate your taking the time to help.

Dan

Duffy's peak shows up pretty nicely in that table. I guess many don't like his defensive prowess as much as I and his other best friends do.

One warning, rawagman: the 58 in the second spot on Duffy's line is the 1891 AA and I didn't take a strength of league discount for that season.

I think it would also be interesting to see the RCAA for Heinie Reitz (1895; OPS+ 90), Joe Quinn (1897; 67), Art Ball (1898; 33), Charlie Harris (1899; 73), Pat Dillard (1900; 56), and Jack Dunn (1901; 63).

These are some of the third basemen who had to take over for McGraw when he was out of the lineup.

Brent, that's an argument *for* McGraw, not against him! If 1890's 3B replacement level was low--which, as you point out correctly, it was--that *increases* the value of everything McGraw did. If he hadn't been there, his teams would have been stuck playing Reitz/Quinn/Ball/Harris/Dillard/Dunn all year long!

Now, it also means that a 3B of that era who accumulated the same RCAA by playing more games at a lower rate would be more valuable than McGraw, which is what you're getting at. But, of course, there weren't any besides Jimmy Collins (including defensive value of course), and Collins was rightly inducted a long time ago.

The shallowness of the position in the 1890s is essential to McGraw's case. Whatever he provided in the games he did play was just light years beyond what his teams would have been stuck with otherwise (and what they indeed were stuck with when he wasn't there).

Anyways. The reason I asked OCF to provide the chart, which he so kindly did, was basically as a spot-check of my own work. My system sees McGraw as having an otherworldly peak, not just for contextual factors like 3B rep level but also because it sees his offense as WAAY more valuable than what his OPS+ would indicate. It's a bit hard to tell without the season length, but I imagine the 72 and 66 are in the 154-game 1898 and 99 seasons and the rest are in approx. 135 game seasons, so he would be 76 69 55 43 41 41 37 32 12, which makes him the second-best backlog hitter in year 1 to Duffy's 1894, the best in year two, tied for the best in year 3 with Singleton, and then lagging Singleton after that. Of course, McGraw was playing 3B when it was like 2B and playing it well while Singleton was playing corner OF mediocre-ly and DH, but this was just to get a sense of whether my system's evaluation of McGraw's offensive value lines up with others. Appreciate it, OCF.

Dan, did you get the email I sent you? The 66 and 72 are, indeed, in 1898-99. But then, I have Frank Chance's 1903 through 1906 as 66-52-66-78. What argument can you make for McGraw that doesn't also work for Chance? Yes, 3B has much more defensive value than 1B (I never entirely bought what karlmagnus was selling about 1B defense when he was stumping for Beckley). But Chance (sharing the same can-he-stay-in-the-lineup flaw) played more games than McGraw, his offensive peak was even higher, and we're not exactly overrun with 1B of Chance's generation, either.

I don't vote for either McGraw or Chance; I can't get past the games played issue. But I wouldn't draw a giant separation between them, either, and in my off-ballot reaches I have Chance ahead of McGraw.

The fact remains that McGraw's teams weren't just getting an other-worldly third baseman--they were generally getting about 2/3 of a season of a great third baseman and 1/3 a season of a bad bench player. When there are players available with slightly lower peak credentials but substantially better within-season and across-season durability (Carlos Morán with an OBP-heavy 120 MLE OPS+ in 1750 estimated games; Bill Monroe with a 16-season career and probably an even better hitter; Sal Bando with a 128 OPS+ and an average of 157 games over 10 peak seasons), I will go with them.

OCF, I just got it now, thanks very much for the forward.

Let's see, McGraw vs. Chance:

Glossary

The following numbers are all standard deviation-adjusted. SFrac is the percentage of the season played (compared to a player with league average PA/G in 162 games). BWAA is batting wins above average, BRWA is baserunning wins above average, FWAA is fielding wins above average, Replc is wins above average a replacement player at the same position would have accumulated in the same playing time, and WARP is the first three minus the fourth (wins above replacement). All seasons are straight-line adjusted to 162 games. Sorry for the goofy formatting, but the PRE tag doesn't seem to be able to handle consecutive whitespaces anymore. I'm just translating BP numbers (including the AA discount) for 1891 and 92.

John McGraw



3-year peak: 24.3
7-year prime: 44.9
Career: 54.1

Frank Chance



3-year peak: 19.2
7-year prime: 34.8
Career: 43.7

So the two aren't close for me. Well, first, my system doesn't agree with yours that Chance's offensive peak was higher. Summing BWAA and BRWA, I have McGraw's best as 6.6 in 1899, 5.7 in 1898, 5.2 in 1900, 3.7 in 1893, and 3.4 in 1897, compared to 5.8 for Chance in 1906, 5.2 in 1903, 4.8 in 1905, 4.0 in 1904, and 3.0 in 1907. Not sure what the source of the disagreement is--a season length issue? Run estimation? Floating out values? (An OBP-heavy OPS+ is worth far more in high-offense 1899 than in low-offense 1906; do your RCAA numbers factor that in?) It's not a huuge difference, but I have McGraw's career OPS+, adjusted for baserunning and OBP-heaviness, at 143; Chance's at 137.

Second, although they were both approximately average fielders at their positions over the course of their careers, McGraw's defense was much more up-and-down, and his good fielding seasons lined up with big offensive years, giving him a much higher peak, while Chance was consistently below average until he settled down at first base, and then consistently somewhat above average during his prime.

But those are minor discrepancies. The big difference is positional. Even if you call hitting and defense a wash, McGraw was the equivalent of a middle infielder, while Chance was playing first base--yes, 1B was more important then than it is today, but it's still 1B. If value above average and career length are anywhere *close*, the fact that one played 1890s 3B and the other played 1B just opens up a yawning gap. I imagine if you did RCAP rather than RCAA it would show this.

Again, thanks for the chart.

Brent, the Negro Leaguers are in the eye of the beholder, but I definitely don't see the case for Bando. He wasn't nearly the hitter McGraw was, and he was a poor fielder. The gap in value above positional average (summing offense and defense) is so great between McGraw and Bando that you'd need to use a Single-A replacement level or something to have the playing time gap count for enough to compensate.

What about Ed Williamson? Well, I know you can't run pre-'93 players. But seriously, the guy was good on both sides of the ball and he stayed in the lineup. A Ryne Sandberg type of player, he even played in Chi-town.

I mean, seriously. Willliamson never played more than 132 games in a season because that's all there was. He led the league in GP at 63, 98, 113 and 127. In 13 years he piled up 1200 games and about 5100 PA.

McGraw once played 143 games in a season, because he could (the games were on the schedule). Big Ed never had that opportunity. Yet in 16 years Big John played in 1099 games with <4800 PA. Don't tell me Big Ed had a short career. He retired in 1990 and McGraw started in '91 yet Ed played ≥ 100 games 5 times, same as McGraw.

Ed's games were all in large increments except for 120 games in his final 2, declining years. He had 11 seasons and almost 1100 games of ≥ 75 percent of team games. About 475 of McGraw's games were played in increments of about ≤67 percent of the team's games for the year. That means he played about 600 games in 5 seasons in increments that significantly drive toward a pennant.

McGraw's OPS+ 135 vs. Ed's 112. Ed was an A glove, McGraw B+. Ed led the league in FA 5 times, John never did. Ed led the league in BB once and stole > 80 bases in 3 seasons as a regular in which SB were accounted for, with a high of 45. Obviously McGraw did more of those 2 things but Ed was not a slug. And Ed had almost 2X XBH, about 380 to 200. That is not insignificant, about 450 more XB. McGraw stole 436 bases to get most of these back, unless you count Ed's SB too, but of course we only have 3 of his 11 years as a regular to go by. If he stole bases at the same rate as a younger man as he did as an older man, then he probably had about 350, or maybe 85-100 fewer than McGraw. So his advantage is TB + SB is still 375 bases, give or take. Ed made about 3400 outs, McGraw 2600, just in the interest of full disclosure.

OK, McGraw was a better player, with more ability. Ed was more valuable, period, and substantially more valuable in years of ≥75 percent of his team's games. IOW Ed had 11 such years, McGraw 5, and Ed had more than twice as many such seasons. Too bad you can't run pre-'93. I think you'd find Ed Williamson to be Ken Boyer with a little bit more glove when the glove was a little more valuable.

A curious coincidence in Dan's numbers. Chance and McGraw have almost exactly identical profiles (career, 3-year peak, 7-year prime, even season-by-season comparison) if you take away McGraw's extraordinary 1898 season. In other words, if you buy Dan's methodology John McGraw is Frank Chance plus one historically great season.

I imagine you mean his '99...his '98 was an outstanding, MVP-type season, but I certainly wouldn't call it "historically great."

I think the best way to look at the Chance/McGraw comparison is that they are extremely similar in all regards except for one critical one: one played the equivalent of a middle infield position, and the other played first. That factor is simply gigantic and puts them in completely different classes.

Yes, I meant 1899. His 1898 comps to Chance's 1906 and then the comparisons keep coming. As I said, just a curious coincidence.

I agree that analyticaly your description of the difference as "Same offensive value, same level of defensive facility, huge difference in defensive responsibility" is more salient.

But those are minor discrepancies. The big difference is positional. Even if you call hitting and defense a wash, McGraw was the equivalent of a middle infielder, while Chance was playing first base--yes, 1B was more important then than it is today, but it's still 1B. If value above average and career length are anywhere *close*, the fact that one played 1890s 3B and the other played 1B just opens up a yawning gap. I imagine if you did RCAP rather than RCAA it would show this.

Just wanted to note that McGraw's RCAP is 459, which is BY FAR the highest of anyone we haven't elected. Chance's RCAP is 309, which is very very good, but still a long way from McGraw.

A tidbit that might budge the needle on some evaluations of John McGraw's candidacy: he has a strong case for holdout credit in his superb 1900 season. McGraw was transferred from Baltimore to Brooklyn, which had the same owners, after the Orioles were contracted, and then sold to St. Louis. McGraw refused to report until May 8, when the Cardinals offered him a contract with no reserve clause. He signed that deal, absolutely obliterated the extremely tough contracted 1900 NL (.505 OBP in a .339 OBP league, over 50 points higher than the runner-up), and then returned to Baltimore when the AL went major the following year. If Edd Roush got credit for holding out, so should McGraw, and this was over 10% of a season right in the middle of his peak. Crediting him for it improves his WARP2 in my system for 1900 from 6.2 to 7.2, a meaningful improvement for a peak year.

I also found out why he missed 37 games in his monster 1899: his wife died of a ruptured appendix and he was apparently too aggrieved to continue playing. Strange for someone who had a reputation as a hard-a$$, and who remarried three years later. But I'm afraid I can't quite make a case for bereavement credit--nor for his being credited for the 5 games in 1902 when he was suspended for "rowdyism."

Another fun tidbit: McGraw tried to sign a black 2B by passing him off as a Cherokee Indian named "Chief Tokohama" for the 1901 Orioles, but Charlie Comiskey nixed the idea. Apparently his name was Charlie Grant--no chance he was confused with HoM'er Frank Grant, was there?

The '99 Orioles are an interesting team. The syndicate Brooklyn/Baltimore ownership did move Keeler, Kelley, and Jennings to Brooklyn, along with Ned Hanlon, so the old Orioles no longer existed as a unit. But the owners didn't utterly gut Baltimore the way the St. Louis owners notoriously gutted Cleveland. With Hanlon, Kelley, and Keeler, and with the acquisitions from other sources of Bill Dahlen and Duke Farrell, the Brooklyn team made the instant jump from the second division to the pennant. But they left McGraw behind in Baltimore, and sent a 20-year-old Jimmy Sheckard the other direction, and they let 28-year-old rookie Joe McGinnity have a breakout year. (In the Cleveland scenario, anyone who showed a spark of life got shipped to St. Louis.) However, even though they were allowed to keep some talent, their 24 games over .500 finish has to be a major marker in the evidence that McGraw was a great manager.

Well, McGraw accounted for about 10 of those 24 wins over .500 all by himself just as a player. Moreover, the team went just 20-17 after McGraw's wife died, versus 66-45 beforehand, so that gives a pretty good sense of a what a .547 OBP with 73 steals and Gold Glove defense at a scarce position will do for a team.

Does anyone know why McGraw, in particular, stayed behind in Baltimore when Kelley, Keeler, and Jennings moved to Brooklyn? It wasn't that Brooklyn had an established 3B. Billy Shindle was awful in 1898; Doc Casey was better in 1899, but still made 3B look like the weak link position on that combined superteam. Was it because McGraw himself insisted (and was listened to)? Was it McGraw leaping at a chance to manage? Did he have a falling-out with Hanlon?

McGraw insisted on staying in Baltimore because he had a stake in a profitable saloon business there. He was sold to St. Louis in 1900 because he announced he would not report to Brooklyn (for the second straight year), and held out until he got offered a contract without a reserve clause (virtually unprecedented in those days). Then he went postal on the super-difficult contracted 1900 NL for the remainder of the season.

Charlie Grant and Frank Grant are indeed 2 different players, if that's the question.

Credit for a player holding out in order to maintain his profitable saloon is one of the most ridiculous arguments I've ever heard in the HoM. Unlike players which receive deserved black-list.collusion credit (Jones, Raines), St. Louis was not acting unlawfully.

What if he had held out because he had a lucrative teaching job?

What if he had held out because he had a lucrative teaching job?

Its like kosher laws. Its all not kosher, but eating a whole piglet, killed in a slow, painful way, cooked in its actual mother's milk, is somehow more offensive.

Another fun tidbit: McGraw tried to sign a black 2B by passing him off as a Cherokee Indian named "Chief Tokohama" for the 1901 Orioles, but Charlie Comiskey nixed the idea. Apparently his name was Charlie Grant--no chance he was confused with HoM'er Frank Grant, was there?

Charlie Grant was already a top player and he is on the short list of best turn-of-century black players. He played for the Columbia Giants 1899-1901, maybe the strongest team in 1900, and he replaced Frank Grant for the Philadelphia Giants in 1904-06 (Sol White's team, probably the strongest team).

--
McGraw was one leading man in the American Association drama of 1899-1900, organizing a rival major league. In 1900-1901 there were two rivals in making. McGraw ended up with the winner but I think he had options with both and probably the NL would have welcomed him, too.

For the success of the 1899 Orioles, McGraw and Robinson were recognized as great leaders, including or emphasizing leadership in the sense instilling a style of play, tactics and competitive drive in this case.

Robinson did get some of the credit. His career didn't begin as McGraw's assistant in New York. St Louis paid a lot for Robinson as well as a king's ransom for McGraw.

I meant to say, they moved to St Louis in 1900 because the AA collapsed. And the flaming holdouts --refusing to play in Brooklyn (or anywhere in the NL?), far in advance of any practical deadline, must have been partly a tactic in building the AA: open commitment hoping to elicit the same from money men in some other crucial cities.

Muggsy is making a mad dash to the finish line! Ever more members of the electorate are impressed by his otherwordly OBP, mayhem on the basepaths, and slick glove at a middle infield-equivalent position in a single-league era! Don't get left behind--his train is leaving the station! Vote McGraw in '08! :)

Ever more members of the electorate are impressed by his otherwordly OBP, mayhem on the basepaths, and slick glove at a middle infield-equivalent position in a single-league era!

Actually, we're only talking about one more vote for him so far (and a 15th place one, at that).

Not that I would be upset if he did make it this year, since I used to be one of his strongest supporters. He may not have been that durable, but he was as great as any third baseman in history when he did play.

Way to rain on the parade, John. We're trying to build some momentum here. :)

Ya know, you could always resubmit your ballot with him on it, thus making it two new votes for him...it's not too late!

Way to rain on the parade, John. We're trying to build some momentum here. :)

Sorry, Dan. Of course, a Raines-R. Smith-Lundy ticket would be more my speed. :-)

23. mulder & scully Posted: September 15, 2006 at 01:46 PM (#2177860)
Speaking of the International Association, does anyone know where to find reasonably complete stats for the 1877-1879 years? I have Nemec's 19th C. Encyclopedia, but the IA stats are not in there.
Thanks
24. Paul Wendt Posted: May 16, 2007 at 10:59 PM (#2366316)
mulder, Are you still interested in IA playing statistics? If so, what is your mailing address?
------

The implied offer still stands. I can provide some reasonably complete playing statistics for 1877-1880 by mail. I believe there are some important open questions, more in number and more open than for NL statistics.

By the way, in 2007 election results Tommy Leach is suddenly the second non-pitcher in the backlog.

In "Ranking ... Group 3" OCF quoted Howie Menckel's list of HOM player movements and replied
143. OCF Posted: February 25, 2008 at 12:30 AM (#2699041)
> 1899-00 - McGinnity and Sheckard go from Baltimore to Brooklyn

That's just the cleanup phase, involving players who weren't yet well-known superstars. The main Baltimore to Brooklyn move was in 1898-99 when Keeler, Kelley, and Jennings made that move.

Many of the listed "moves" are incomplete; the moves were more comprehensive but other parties are not HOM players. In this case, the crucial missing party is manager/minority-owner Ned Hanlon. In a sense, Hanlon and majority-owner vonderHorst chose to break up the Orioles in this fashion. They got ownership shares in the Brooklyn club and vice versa. In effect, Hanlon was the modern "general manager" of both clubs and the field manager in Brooklyn. McGraw was the field manager in Baltimore.

. . . Really, the most surprising move involving Baltimore, Louisville, and Cleveland is the one that didn't happen: McGraw didn't go to Brooklyn.

McGraw refused to go, or bargained aggressively to remain in Baltimore. He was immensely popular in Baltimore and owned a sporting business there with his pal Wilbert Robinson. (The popularity and the business were associated, and jointly associated with his desire to stay in Baltimore.) He was ambitious to manage his own team, too. Finally, he hoped for part ownership (player-manager-owner) in a new Baltimore club, in a rival major league expected to organize post-season.

I don't know whether McGraw's wishes were decisive, even in context of some personal relationship with Hanlon. (Until late in his Brooklyn tenure, I believe, Hanlon had the baseball credibility and power in the organization to make the McGraw assignment himself.)
The prospect of syndicate profit in Baltimore may have been necessary. Certainly they managed for profit in both cities; it wasn't a StLouis-Cleveland situation. A fighting Baltimore Orioles team with McGraw at the helm and Robinson as his right-hand man would be popular. Maybe more popular than Hanlon-led team with better players?

McGraw got (requested? demanded?) and Hanlon gave (offered? conceded?) the promise that he would not move players between Baltimore and Brooklyn in-season. He would make the player allocation decisions pre-season and it would be McGraw's job to do the best possible on the field and at the gate with that team behind him.

[all from memory]

This is a good example of why McGraw is such a no-brainer Hall of FAME candidate. He was a force of nature, a great player with great import even beyond his managing (which we don't count).

He even has a respectable case for HOM, though the missed games ultimately have sunk him so far.

The new lists were a way of revisiting these discussions we had 5 years ago, updating with what our eventual Hall of Merit choices did back then.

And thanks as always to Paul for the valuable insight...

Cross-posting McGraw's comparison to other elected 3B from the ranking thread:

Here's McGraw versus the elected 3B I have in or below his neighborhood:


Hack (including war discounts)



3-year peak: 19.2
7-year prime: 39.4
Career: 57.1


Boyer



3-year peak: 20.5
7-year prime: 41.7
Career: 53.1


Nettles



3-year peak: 18.8
7-year prime: 38.6
Career: 62.2


Brooks



3-year peak: 20.6
7-year prime: 39.9
Career: 63.4


Muggsy

(Based on BP stats and league difficulty adjustments for 1891-92, and including holdout credit for 1900)



3-year peak: 25.3
7-year prime: 45.9
Career: 54.1


And here's Rosen, just for good measure:



3-year peak: 21.2
7-year prime: 32.4
Career: 32.4

Yes, I know Rosen deserves two years of minor league credit, but I couldn't find posted MLE's for him, and anyway, adding on even two MLB All-Star seasons, say a 5 and a 4 WARP2, don't get him into serious candidate territory.

So, by 3-year peak in this group, we have:



By 7-year prime, it's:



By career, the order is:



Finally, by my salary estimator, I rank them (in millions):




So: OK, if you're a strict career voter, McGraw's not there--but neither is Boyer, or Hack after (reasonable) war deductions. But if you care at *all* about peak or prime, it's hard to see how McGraw doesn't make your 2009 ballot. He was just monstrously good. Yes, there were a lot of missed games, but without them he'd be a no-brainer.