Here is my prelim 1901 ballot, with comments on guys who have moved.

1. Jack Glasscock (-) - No brainer #1 choice for me. Career value.
2. Hardy Richardson (2*)
3. Pud Galvin (4*) - I flipped Galvin and Bennett. A little more pitcher re-evaluation.
4. Charlie Bennett (3)
5. Tim Keefe (10) - Keefe and Radbourn moved strongly up as part of my "annual" pitcher changes.
6. Harry Stovey (6)
7. George Wright (7)
8. Old Hoss Radbourn (11)
9. Ezra Sutton (8)
10. Fred Dunlap (9)
11. Pete Browning (13) - Starting here is my "dont think they will make it in" section. I reorganized this part. It may be completely different when I finally vote in '01.
12. Joe Start (14)
13. Ned Williamson (!) - Back in the ballot after a year absense.
14. Tom York (12)
15. Bob Caruthers (15)

Spalding, McCormick and Whitney also were considered for the last 3 spots.

Rob, I don't understand what you are doing with your pitchers. Are you upgrading the pitchers from "last year" because you feel your analysis of their accomplishments were incorrect or are you moving them up because you need to have pitchers up in the top ten? Obviously, I have no problem with the former, but the latter is a problem for me.

We are supposed to pick the best 15 players and put them in the proper order, If you had Keefe at #10 for the last ballot, unless you felt you were misinterpreting his statistics, he should be around the same ranking as he was last year.

If I'm reading what you wrote in your post incorrectly, I apologize.

1) Al Spalding

2) Ezra Sutton

3) George Wright

4) Jack Glasscock I have him basically tied with Wright, except he's more career value than Wright (and George is more peak). If elected, he will have the most unfortunate surname in the HoM for at least 120 years. :-)

5) Cal McVey

6) Dickey Pearce

7) Hardy Richardson

8) Joe Start

9) Tim Keefe

10) Charlie Bennett

11) Levi Meyerle

12) Ed Williamson

13) Fred Dunlap

14) Lip Pike

15) Arlie Latham Longtime thirdbaseman. Value more on career than peak.

John,

Legitimate question. No, Im not boosting them to get them in the top 10. In fact, I dont like the top loaded distribution of them, and was considering moving them down to spread them out, but that would be just as bad (actually, if McCormick or Whitney or Spalding had been included, the balance would have been better). Im constantly refiddling with my rankings. I am making a slow adjustment over time based on how pitchers were used (1 man rotations to modern 5 man rotations) but that doesnt really affect Pud and Hoss does it?

One thing that happens is that I have a file with a bunch of different numbers in it (career value, peaks, era+, etc.) and I use it to rerank from scratch each time. Then I compare with past year, tweak a little, and then wait on someone to point out something horribly wrong and adjust from there.

Also, this "year" I considered Spalding again. I look at the top vote getter who isnt in my list, to see if I maybe have him wrong. And, he was much higher this time than the last time I considered him. He still didnt make my top 15 though.

Next "year" I may decide Im putting too much emphasis on peak value, and return to something more like my 1898 ballot, which was almost a pure career value vote. Or maybe not, It will be 2 weeks before I have to worry about that.

Anyway, when I did my initial rankings this time, Pud finished 2nd above Richardson, but I flipped them, because I didnt agree with the result. The same may happen somewhat with Keefe and Radborne, but I left them where my rankings came out.

However, I do use some positional adjustments in the middle of my ballot. The differences within the values of the guys from about 5-10 are so small to be within any margin of error of any system (same for the guys from 11-18), so I do pay a little bit of attention to positions, but it really has been unnecessary so far. There hasnt been an overload at any position.

At the top, I pay no attention to position. Anson and Connor will be 1-2 when they hit the ballot, and if Brouthers somehow isnt in, they will be 1-2-3.

Rob:

That sounds cool. I've done the same thing in the past. I just wanted to be sure that I was clear on your attentions.

No problem, John. We are supposed to defend our picks. Im too schizophrenic to rank players the same way every year. There are too many legimate methodologies. Now, if I could get the voters who left Richardson off their ballots to explain that, I would be much happier. I think we need a 2nd base thread like we had the pitcher thread. Both Richardson and Dunlap are being undervalued by some.

The pitcher thread helped me a lot, and also helped others see the value of Pud Galvin, so I dont feel so alone ranking him as high as I do. Im beginning to believe others may be right about Bennett, but I havent lowered him (much) as of yet.

Both Richardson and Dunlap are being undervalued by some.

I can understand leaving Dunlap off your ballot (though he's on mine), but Richardson? You got me.

I think what's needed is a good "Who's better? Richardson or Sutton" argument and a good "Who's better? Radbourn or Spalding" argument, with advocates for each separating the twins. Actually, a "Stovey or Browning" battle might work as well, although Stovey seems to win that one simply by his not being a butcher in the field (anyone want to counter the prevailing 'wisdom' there?

I think what's needed is a good "Who's better? Richardson or Sutton" argument and a good "Who's better? Radbourn or Spalding" argument, with advocates for each separating the twins. Actually, a "Stovey or Browning" battle might work as well, although Stovey seems to win that one simply by his not being a butcher in the field (anyone want to counter the prevailing 'wisdom' there?

I think what's needed is a good "Who's better? Richardson or Sutton" argument and a good "Who's better? Radbourn or Spalding" argument, with advocates for each separating the twins. Actually, a "Stovey or Browning" battle might work as well, although Stovey seems to win that one simply by his not being a butcher in the field (anyone want to counter the prevailing 'wisdom' there?

Howie - I agree. In triplicate. Ive started putting together the
Richardson-Sutton comparison. Give me a few days...

Actually, a "Stovey or Browning" battle might work as well, although Stovey seems to win that one simply by his not being a butcher in the field (anyone want to counter the prevailing 'wisdom' there?

I have Browning about equal with Stovey until I give the Gladiator a positional bonus for the higher attrition rate for the positions he played over Harry.

"Who's better? Radbourn or Spalding" argument, with advocates for each separating the twins.

I don't have Radbourn on my ballot, though he still might be the best pitcher of the 19th century (if we include his pre-NL work). I don't know.

"Who's better? Richardson or Sutton" argument

I take Sutton over Richardson for the longer career at a more demanding position.

(I'm in Pittsburgh for a training class this week)

Hope the Pirates find room for you on the roster, Joe! :-)

1. Spalding (3-2-1)--still the giant on the board
2. Keefe (--5-4)--THAT close to Clarkson, should follow soon
3. Radbourn (10-10-6)--pardon my profusion AND confusion re. pitchers
4. Wright (8-4-4)--a truly great player, but follows Spalding
5. Glasscock (new)--best 19th century career at SS but not as valuable in other respects as Wright (peak) or Ward (pitching)
6. Richardson (9-9-8)--like Muldoon, he's a solid man
7. McVey (5-6-8)--a poor man's Deacon White but with a better peak
8. Caruthers (--11-9)--a unique career, as hard to peg as the other pitchers, but moving up

In/out line somewhere between 7 and 10, maybe here:

9. Browning (--8-10)--mixed bag: career OPS+ 164, short career, AA discount
10. Pike (6-7-13)--great peak, much much higher than Start or Sutton

11. Bennett (--13-11)--best catcher til Ewing
12. Start (15--12)--like Pike and others, hard to peg pre-'71
13. Stovey (--12--)--follows Browning
14. Sutton (13-14-14)--no peak
15. Dunlap (14----)--back after two-week absence

Dunlap replaces Williamson, head-to-head: better hitter, good fielder though not as demanding a position as Big Ed's. Plenty of pitchers on the ballot, no more room but if there was Mullane, Bond, Welch, McCormick would all be ahead of Galvin. Pud's record more than anything else suggests a guy who never threw the ball hard enough to hurt himself.

Prelim. ballot
Pretty much just moved people up after losing two players to HOM. I had about 5 people going for my 15th place spot, but this is what I got.

1. Tim Keefe(2)
2. Harry Stovey(3)
3. Pete Browning(5)
4. Hardy Richardson(6)
5. Joe Start(7)
6. Ezra Sutton(8)
7. Jack Glasscock(n/a)
8. George Wright(9)
9. Al Spalding(10)
10. Hoss Radbourn(11)
11. Pud Galvin(12)
12. Charlie Bennett(13)
13. Charley Jones(14)
14. Mickey Welch(15)
15. Cal McVey -Was off my ballot last time. Realized I was seriously underestimating him.

Just missed Williamson, Mullane, Caruthers, Pike

'Parisian Bob's still the best, y'all need to give my boy a little more love.'
The problem I have with Caruthers is that he packed all 5 peak seasons into his AA seasons. When he went to the NL, it all went downhill and its not like he was old when he got there(26). His 3 seasons in the NL went like this: Very good, Above Average, Freakin Terrible. Then he retired after his age 28 season. My point is, he was great in the AA, but there's not a whole lot of evidence that he could stay 'great' when playing against the best. This is why I lowered him in my rankings 'last year'.

1901 Preliminaries. I am taking a new interest in "positional peak", i.e., how often was he one of the top two players at his position for the year? I've always been a big fan of peak, so this isn't changing my ballot too much.

1. George Wright (2) -- People talk about Wright like he's the "peak" pick over career value, but that peak (if you include 1869 and 1870) made him one of the top two shortstops in baseball for 10 out of 11 years between 1869 and 1879. That peak alone should be good enough for career value people.

2. Joe Start (5) -- I'm moving him up here. He has 9 Top 2 finishes at first base, and that doesn't include any of the 1860s. He'd have had more if he weren't crowded out by Brouthers and Connor, who were often the best in the league, let alone the best at first base.

3. Ezra Sutton (3) -- Seven top 2 finishes.

4. Bob Caruthers (4)

5. Jack Glasscock (--) -- Ten top 2 finishes in his league, but several were due more to league dilution. He's better than I thought at first, and may move up by the final ballot.

6. Tim Keefe (6)

7. Hardy Richardson (7)

8. Pud Galvin (8)

9. Al Spalding (10) -- I had Ward 9th, so everyone else gets an extra bump.

10. Harry Stovey (11)

11. Charley Radbourn (12)

12. Pete Browning (13)

13. Cal McVey (15) -- Dropping Ed Williamson this time, which leaves room for first appearances by . . .

14. Lip Pike (off ballot), and

15. Charlie Bennett (off ballot)

Meyerle, Dunlap, and Mullane are bubbling under the list, fighting amongst each other for a premiere spot when a new place opens.

The McVey spot could easily go to Levi Meyerle, who was also great in the NA, but I'm just now noticing had his greatness spread over second base, third base, and the outfield, so didn't rank among the best at any single position. Also just under the radar are Charlie Bennett and Tony Mullane

Off course, ignore that last paragraph that I accidentally cut and pasted from my last ballot.

Richardson-Sutton is a good comp to study. I tend to like peak value more than career, especially in the 19th century when so many players did not have long careers, so I like Richardson over Sutton. But I realize this is not in-depth.

I like Browning over Stovey for the same reason, and a 164 OPS+ versus Stovey's (what?) 140.

Finally, I don't see a Spalding-Radbourn pairing. The real question is how to rank the short '70s careers, some of which take additional value from the '60s and some that don't (i.e. I'm not including the unusually long careers of that time like Start and Sutton and White in this group). And I see the short '60s-'70s careers as being:

1) Spalding, 2) Barnes, 3) Wright, 4) McVey, 5) Pike, 6) Meyerle.

Of course, you've got to slot them in amongst "the English" eventually but first you gotta get 'em right among themselves. I see Spalding and Clarkson as a better pairing in that each was clearly dominant in his "class."

Clarkson representing the 1-man rotations from the late '70s and '80s, and those clearly are divided into 2 classes. The 300 game winners are (in order) Clarkson, Keefe, Radbourn, Welch and Galvin; and the high peak guys/short career guys like Bond, McCormick, Caruthers, Foutz, Whitney, Corcoran, etc. etc., who have me completely and totally baffled (well, except Caruthers who is clearly the class of the group). And then you've got 'tweeners like Mullane and maybe Mathews.

So I see Spalding as "the dominant man from the '70s" rather than as a "pitcher," because he does not seem to comfortably fit in that "class."

1901:
DICK BUCKLEY,1895
OYSTER BURNS,1895
CON DAILY,1895,1896 -> 9
DAVE FOUTZ,1895,1896 -> 2
JACK GLASSCOCK,1895
ARLIE LATHAM,1895,1896 -> 8,1899 -> 6,1909 -> 4
GEORGE TEBEAU,1895
WALT WILMOT,1895,1897 -> 11,1898 -> 35
BILL HUTCHISON,1895,1897 -> 6
PHIL KNELL,1895
HARRY STALEY,1895
SCOTT STRATTON,1895

There is a question on Wilmot's eligibility. We will wait for the Committee to decide.

To me Glasscock is the only top-shelf player on this list. There will be some borderline favorites here: Hutchison was the top pitcher for a couple of years, not arriving until 30, just a few years before 60'6"; those who love Caruthers will like Foutz; Stratton is another hard-hitting pitcher that might have made some noise if he had adapted to the mound change; Walter Arlington Latham is a character with a long career while being "The Freshest Man on Earth", leading the league in heckling and insults for many seasons; George "White-Wings" Tebeau is not to be confused with "Patsy" (his younger brother? Oliver Wendell Tebeau).

Prelim:

1. Keefe (1)
2. Radbourne (4)
3. Galvin (8)
4. Richardson (9)
5. Caruthers (7)
6. Foutz (new)
7. Stovey (6)
8. Mullane (3)
9. M. Welch (12)
10. Glasscock (new)
11. Browning (11)
12. McCormick (13)
13. Sutton (10)
14. Latham (new)
15. Hutchison (new)

Off: Williamson, Wright

I have done some shuffling, with Galvin and Richardson the big beneficiaries and Mullane slipping the most.

This is far from final.

jimd wrote:

"those who love Caruthers will like Foutz"

As a Caruthers-lover, I want to say emphatically that I do not like Foutz, and I think looking at why they are not at the same level will demostrate why Caruthers is a clear "in" and Foutz is a clear "out."

First, they should be easily directly comparable, as they played in the same leagues at about the same times (often for the same teams).

Both pitched, essentially, from 1884-1892. Foutz continued for 2 more years as a hitter. Let's compare each area separately.

Pitching:

Caruthers -- 302 PRARP (218-99, 2828 IP, 123 ERA+)
Foutz -- 183 PRARP (147-66, 1997 IP, 124 ERA+)

The difference is clear. Caruthers and Foutz had similar pitching careers, but Caruthers pitched about 50% more than Foutz did. It's really not even close. Their peak seasons (1885-1887) were comparable, but Caruthers had three more years (1888-1890) that were just one notch below his peak(92-37 record), while Foutz fell off a cliff (17-8 record in those years).

Caruthers versus Foutz as pitchers is about as close to a no-brainer as there is.

Now, let's look at hitting. Foutz has the advantage in plate appearances, 4847 to 2906. Essentially, even though Caruthers pitched 50% more innings, Foutz had almost 50% more plate appearances. Does that narrow the gap?

Not at all. To see why, you can look at the pattern of their OPS+. Caruthers' OPS+ was fairly consistent, or, at least, the second half of his career is not obviously worse than the first.

Caruthers pitched 340 games, and hit with an OPS+ of about 135. That is just incredible, and probably makes him the best hitting pitcher since Babe Ruth (or, would have been, if Ruth had come first.) Foutz pitched in 240 games, and even though his OPS+ for his career has 104, for the first half of his career he was a much better hitter. Call his OPS+ as a pitcher 115 (the straight average of his OPS+ from 1885 to 1890).

That leaves the non-pitching part of their careers. And here's where the even bigger difference is:

Caruthers played another 380+ games, mostly in the outfield, and with an OPS+ of about 135 (they were all scattered among his pitching performances, and there's no reason to think he hit any better when he wasn't pitching.)

Foutz, on the other hand, played in another 900+ games when he wasn't pitching (more than twice as many as Caruthers), but the vast majority of them was 1889 and beyond. He played 595 games at first base and 320 games in the outfield. If you adjust his OPS+ to take into account when these games were played (primarily later in his career), his OPS+ as a 1B/OF player was in the low 90s.

Essentially, Caruthers and Foutz were comparable pitchers, but Caruthers pitched 50% longer. Caruthers and Foutz were both good hitters when they were pitching (although Caruthers was significantly better).

But when they were not pitching, Caruthers threw in four extra seasons of Fred McGriff/Rafael Palmiero type output, while Foutz threw in almost a decade of Travis Lee.

Now, counting stats are all good and nice, but even if Travis Lee continues at his established level of play until he is 77 years old, smashing all established records in the process, he's still not a good player and would never get my vote.

So, here's are my mathematical formulas:

Bob Caruthers = (2/3 Tim Keefe -- career pitching) + (1/3 Fred McGriff -- career hitting)

Dave Foutz = (2/5 Tim Keefe -- career pitching) + (1/5 Tino Martinez -- hitting early in career) + (2*Travis Lee -- hitting late in career)

If Clay Bellinger plays until he's 77, I'll put him in just for making it that long!
Seriously though, you've definitely convinced me that Caruthers was much better than Foutz. I still don't believe Caruthers belongs in so that makes Foutz a clear 'out'. We'll call Caruthers a 'partly cloudy out'.

Matt, this is concerning your comment:

>Not at all. To see why, you can look at the pattern of their OPS+. Caruthers' OPS+ was fairly
consistent, or, at least, the second half of his career is not obviously worse than the first.

I don't have Caruthers' offensive breakdown when pitching vs. when in the field, other than after he was no longer pitching. But the idea that he was consistent seems at odds with his annual OPS+ figures:

113 (23 games)
83
196
164
113
121
115
120
131 (not pitching much)
119 (not pitching)

I mean, yes, he was consistent during the final six years of his career, but the idea that he was a "great" hitter, not just a great hitting pitcher but a great hitter, obviously derives from the 196-164, at which level he was clearly not consistent. Maybe I'm being too hard on the guy, and he has been moving up my ballot. But I don't see that his "established level" which is more like 125 than 175 is "great" other than for a pitcher. I mean Bob Lemon had two seasons at 179-135. He was a great hitting pitcher, not a great hitter.

Jim Gilliam is a comparable modern player, though Latham had more defensive value. Nice player, not worthy of mention on a ballot.

While I agree he probably doesn't belong in the HoM, you're a little too harsh on him, Joe. Do you see a big difference between him and Ed Williamson (who makes your list)? I don't.

As a Caruthers-lover, I want to say emphatically that I do not like Foutz, and I think looking at why they are not at the same level will demostrate why Caruthers is a clear "in" and Foutz is a clear "out."

Was Caruthers ever the best at his position? No. Was Foutz? Arguably yes in 1890 (for all of baseball) at first.

I have both of them about equal. They both don't belong.

I'm going to be honest here. I'll be very disturbed if Caruthers makes it into the HoM (even more than Stovey). I feel he's being compared to a pitcher today instead of a pitcher for his time. Obviously, he wouldn't hit (or play the outfield between mound visits) like that today. He would be something (I'm not say exactly) like Mike Hampton is today. If you compare him to his time, he was very good, but not great.

Bob Caruthers = (2/3 Tim Keefe -- career pitching) + (1/3 Fred McGriff -- career hitting)

Dave Foutz = (2/5 Tim Keefe -- career pitching) + (1/5 Tino Martinez -- hitting early in career) + (2*Travis Lee -- hitting late in career)

This is 100% wrong. Caruthers and Foutz are nowhere near McGriff and Martinez. You're not taking into account the overall competition of the time. This goes beyond the AA. A 135+ is not the same as 135+ today. Check out the standard deviation for their era and ours to confirm this.

Again, you have to compare them to their time.

Andrew:

We need to compare the players to their time. That's about the best thing I can tell you.

There are some here that discount pre-1880 baseball, but the whole 19th century is suspect, IMO. There are many more fine players today than then.

Nevertheless, I think the best players at each position should be honored (regardless of how they stack up to today's players).

Do we have any evidence as to the standard deviations for fielding and pitching numbers

Yes, but I don't have it at my fingertips. It was created over twenty years ago.

John -- I see a 113 OPS+ vs. a 92 OPS+. That's a pretty big difference. I'm pretty sure my offensive W-L records show the same thing, although I don't have access to them right now. I don't even think the two are remotely close. Ed was well above the league in both OBP and SLG, Latham was a little below the league in OBP and way (.028) below the league in SLG. And Latham played a bunch in the AA.

Williamson 68.4 WARP3, Latham 39.1.

Joe:
Do you have any idea why Win Shares treats both of them as very close? If not, I'm about ready to dump the whole book. Unless it's picking up Latham's huge difference in stolen bases. Hmm.... :-)

I think the AA reduction is offset by the 90's NL competitive boost for Latham, BTW.

John wrote:

"Was Caruthers ever the best at his position? No. Was Foutz? Arguably yes in 1890 (for all of baseball) at first.

I have both of them about equal. They both don't belong."

I don't see how they could be considered equal: Caruthers WARP 1= 88.3, Foutz WARP 1 = 67.0. Caruthers comes out about 25% better on career here.

Caruthers was also the best pitcher in 1886 (16.0 WARP1) and 1887 (16.2) when hitting is included. That seems clear to me. Foutz was fairly close in 1886 (15.4). Ed Morris had a 15.1 in 1885. Clarkson had a 15.1 in the NL in 1885, which could be close if you think the NL was more than 3 or 4 percent better that year, but 1885 was a high point for the AA, so I don't see it.

Also, I don't see how Dave Foutz was the best 1bman in 1890, Roger Connor was by a longshot. Cap Anson was second best. Foutz would be fighting for a close third with Dave Orr and Henry Larkin.

I am curious: what possible metric has Caruthers and Foutz about equal?

Even the oft-maligned Win Shares gives Caruthers a decent edge over Foutz - 337 to 292, though both are good numbers.

I currently have Foutz one spot behind Caruthers, but Foutz may be dropping.

WS doesn't adjust for league quality, and the AA was worse.

The AA quality issue is overstated, IMO. We're talking about a 4% reduction (but it does need to be factored in).

Latham is not a serious candidate for the Hall, but you can make a case for him at #15. Of course, the same could be said for about ten other players. I picked his name out of the hat for the prelim ballot this week. Whether he stays on my final ballot is another question.

John wrote:

"This is 100% wrong. Caruthers and Foutz are nowhere near McGriff and Martinez."

Perhaps it was an error to bring cross-generational players into the equation. In terms of value added, though, the time period should be irrelevant. Nonetheless, I will amend for clarity of intent:

Bob Caruthers = (2/3 Tim Keefe -- career pitching) + (1/3 Jim O'Rourke -- career hitting)

Dave Foutz = (2/5 Tim Keefe -- career pitching) + (1/5 John Morrill -- hitting early in career) + (Sid Farrar -- hitting late in career)

I don't think Foutz looks any better with contemporary anologies.

I do think it casts a better light on Caruthers, especially f you like Keefe and O'Rourke.

Also, I don't see how Dave Foutz was the best 1bman in 1890, Roger Connor was by a longshot. Cap Anson was second best. Foutz would be fighting for a close third with Dave Orr and Henry Larkin.

I would place Connor over Foutz, but Foutz still has the advantage over everybody else. Foutz was a better baserunner and fielder than both Orr and Larkin. When you add the great, but extremely limited, pitching, you could make the case that Foutz was the best player who played a majority of games at first.

I did said it was arguable, you know. :-)

I am curious: what possible metric has Caruthers and Foutz about equal?

I use a variety of metrics, mix in a blender and "Presto!" :-)

Seriously, I just take all the available information and try to weigh it carefully.

Caruthers was also the best pitcher in 1886 (16.0 WARP1) and 1887 (16.2) when hitting is included.

I was referring to the best at a position in all of baseball. Lady Baldwin or John Clarkson were better in '86. Remember: the AA played more games than the NL. That hurts the counting stats for the NL players when compared to the AA.

Caruthers and Foutz were about tied when taking in their whole package.

Let me say that Caruthers was some player for a couple of seasons, but I feel he falls short when compared to Clarkson, Keefe, Galvin, Radbourn, and Welch. Too short of a career.

With regard to Latham and the AA, don't forget that while the AA might be -4 or -5 %, some of those NL years in the late 70s and early 80s were actually + a couple of % over the average, which would add to the difference between Latham and Williamson.

I am factoring that in. I'm also taking into account Ed's time at short. That's why I have him at #12, while Fresh Man makes the bottom of the list.

FWIW, Bill James had Williamson at #45 and Latham at #54.

Joe wrote:

"Caruthers had the plate appearances of a guy that played about 5-6 years as a regular, O'Rourke played for 22 years. He's about a quarter of O'Rourke.

Same with Keefe, when you adjust for their leagues, he was about 56% of Keefe, not 2/3."

Fair enough. Using Joe's more conservative numbers, 56% of Keefe + 27% of O'Rourke = 83% of a consensus top-ballot HoMer (O'Rourke is in, and Keefe was a first runner up last year).

Unlike Ward, however, who was first a great pitcher for a brief time, and was then an above average shortstop for a while, Caruthers was significant fractions of Keefe and O'Rourke AT THE SAME TIME.

The reason that I weigh peak somewhat over career is the extra value that 40 Win Shares (or whatever) gives you in one season, compared to 20 Win Shares for two seasons. By doing it all at one time, rather than all spread out, you create additional value in your contribution to any given pennant race. And when you see that, between 1885 and 1990, Caruthers' team won 5 out of 6 pennants, you can see the advantages of players like Caruthers (and to a less extent Foutz), who not only could both hit well and pitch well, but could do it in the same season. The value is at least additive, and in some ways exponential.

The criticism appears to be that his career was "short", but a career is only short if you don't do enough within that time period to make your case. Caruthers' case would certainly be stronger with a few more years of decline factored in, but his case doesn't need it.

Even if Caruthers is only 83% of the Keefe/O'Rourke hydra, he makes up the extra 17% by doing in only 7 full seasons 83% of what Keefe/O'Rourke did in in about 17 (taking the average of the two).

Fair enough. Using Joe's more conservative numbers, 56% of Keefe + 27% of O'Rourke = 83% of a consensus top-ballot HoMer (O'Rourke is in, and Keefe was a first runner up last year).

Which makes Caruthers an above average, but not a great player. I thought I got him right. :-)

10.4 9.1 8.6 5.3 5.2 4.2 3.8 3.6 0.5 0.4 -0.1
11.1 9.3 9.0 8.5 8.4 8.0 7.6 5.7 5.5 4.1 4.0 3.9 3.7 2.9 2.7 2.1 0.9

Here are the seasonal warp3 numbers for Caruthers and Ward, sorted from best to worst. While I tend to distrust the warp3 for pitchers, its a reasonable starting point. Caruthers had a great peak because he put the hitting and pitching together. 3 of Ward's best 4 are during his pitching days, although 1 of them got a healthy hitting boost. Caruthers is on my ballot due to his peak, but I dont know how anyone considers them close in value. Caruthers had an era+ advantage (123 to 118) in 350 more innings, but it doesnt make up the difference.

Caruthers isnt being hurt by putting the hitting and pitchers together at the same time, he is being helped by it. If he had all the hitting after he was done pitching, his career value would be the same, but his peak would be low enough that he wouldnt even be making my ballot.

Should we really be docking him for having his great career pitching and hitting numbers at the same time?

Who's docking him? I know I'm not.

Of course, after he was more or less washed up as a pitcher, he played the 1892 season as an outfielder (122 of 143 games), got 604 PAs, and hit .277/.386/.357 in a league that hit a park-adjusted .246/.318/.329.

And he was about the sixth best rightfielder that year. If you want to compare him to all outfielders, he falls further down the line. He was nothing special that year.

He made extraordinary contributions to 5 teams that won pennants, usually by small margins.

a LITTLE bit of overstatement here, perhaps. In the 3 years Caruthers was with the Browns, they won by 16, 12, and 14 games. After he and Foutz left, they slipped from 95 to 92 wins (in a 140 game schedule); the Browns didn't miss them; there were no irreplaceable players on that AA juggernaut. OTOH, he probably was the MVP of the 1889 Bridegrooms; they edged the Browns by 2 games; without Caruthers they don't win. In 1890 they won by 6 games; Caruthers also was past his peak; WARP1 says if all the Grooms could get was a replacement player for him, they would have been in a dog-fight.

Bottom line, he was essential to 1 pennant winner, while making significant contributions to 4 others.

Caruthers was a great pitcher for about 5 years (the usual shelf-life, just like Spalding, Radbourn, Clarkson, etc.), but in the AA. He was a great hitter for two years, but in the AA. He was a good hitter for a few more. He failed to adapt to the new mound, was a defensive liability in RF, and couldn't hit enough in the condensed NL to cover his defensive sins. He was washed up at 29. He's not a great hitter, anymore than Whitney was; I think he probably had a weakness, and once they figured out how to pitch to him, he failed to make the adjustments.

OTOH, Foutz was 7 years older than Caruthers and didn't play in the majors until age 27. (Just guessing, but he shows up the year they allow side-arm; maybe he was pitching somewhere it was legal before 1884.) He turned himself into a good defensive 1b man, and was still starting at age 37, backup at 38, cut at 39. I don't think he's HOM material, but it's an interesting might-have-been career.

There has been some debate about the quality of the AA compared to the NL. A 4% difference has been mentioned. I?m not sure who came up with that statistic, or even what it means. It does seem to me, however, that the NL had much more frontline talent than the AA. Let?s look at it position by position:

Catcher?The NL had Buck Ewing and Charlie Bennett. The best AA catcher, as far as I can tell, is Jocko Milligan, or maybe Fred Carroll. Neither are as good as Bennett, certainly not as good as Ewing. Advantage: NL.

First base?The NL had Dan Brouthers, who was in a class by himself as a hitter. The NL also had Cap Anson, and Roger Connor; a couple of notches below those guys is Joe Start. The AA had Harry Stovey, if you count him as a first baseman, and Dave Orr. You could argue those guys are better than Start, but they?re not in the ABC category. Brouthers did have one year in the AA, but was primarily a NL player. Advantage: NL.

Second base?Bid McPhee played for the AA; he was the best fielder at his position in the nineteenth century, one of the best ever. In the NL, Hardy Richardson also fielded the position well, and was a better hitter than McPhee. The NL also had Fred Pfeffer, who was a notch below both of these guys. To give the AA a draw is a little generous, but I?ll do it. Advantage: even.

Third base?The NL had Ezra Sutton and Ned Williamson; the AA had Arlie Latham. There is already a debate about whether Latham was better than Williamson; I don?t think you can argue that he was better than Sutton. Advantage: NL

Shortstop?The NL had Jack Glasscock and Monte Ward. IN the AA, Frank Fennelly had a few good years with the bat; Oyster Burns played the position a couple of seasons, but couldn?t field. Herman Long also played for the AA in it?s last couple of seasons. I?d still give this one to the NL, although it?s closer than other positions. Advantage: NL

Outfield?The NL had considerable talent here: Jim O?Rourke, King Kelly, George Gore, Sam Thompson, Mike Tiernan; they also had Paul Hines in the early 80s, Billy Hamilton in the early 90s. The AA had Pete Browning, who was a tremendous hitter but a lousy fielder. After him there?s Oyster Burns, Charley Jones, and Tip O?Neill, all of whom had some very good seasons. Even so, the NL had more top line outfielders. Advantage: NL.

Pitchers?NL top four pitchers: John Clarkson, Charley Radbourn, Tim Keefe (who did have two AA seasons), and Mickey Welch. AA top four pitchers: Bob Caruthers, Tony Mullane, Dave Foutz, Silver King. Which rotation are you going to pick? Advantage: NL.

Maybe you disagree with my picks; maybe I?m looking at this the wrong way. Still, from what I can see, the NL had more and better frontline talent at every position except maybe second base. That has to make the NL a significantly better league. I?m not arguing that the AA is the UA; the league had guys who could play with anybody. McPhee is an obvious Hom?r, Stovey has good argument, and you can (and a few have) build decent cases for Caruthers and Browning. Burns, Jones, Mullane, Foutz?these are all good players. But the league simply was not as strong as the NL, and it deserves a serious discount.

favre, excellent in-depth analysis.

I've been wondering where that 5% figure came from (and to which years it applies). I do know that if you examine the WARP3 numbers for the peak pitchers during that era (e.g. Clarkson vs Caruthers), Davenport at BPs has a much harsher view of the AA then a mere 5% discount.

A little history of the AA. Unlike the other new leagues that started up afterwards (UA, AL, FL), the AA did not arrive on the scene invading NL cities with new teams conducting player raids. It didn't have to, because the NL had committed a nearly fatal blunder.

When the NL started it had teams in 6 of the 8 largest markets; it was a "national" league (or as close as was practical in 1876). By 1882, it had devolved into a regional league; Chicago (3) and Boston (4) were the only large markets; the other teams were in smaller cities convenient to the rail lines between them (nowadays it would be the Interstate-90 league). The AA set up as a similar regional league between Philadelphia (2) and St.Louis (5). However, the cities strung along in between included Baltimore (6), Cincinnati (7), and Pittsburgh/Allegheny (8). (History lesson: in 1880 Detroit is smaller than Louisville; Henry Ford did make a difference. :-)

The AA didn't need to invade NL cities to get into big markets. They didn't need to conduct player raids to convince the fans of the invaded cities that their teams were just as good. All they needed to do was outbid the NL on young talent and they would eventually become the dominant league. It was a long-term plan, and it might have worked.

The AA's blunder was a clause in the "non-aggression" treaty signed in 1883; it allowed teams to switch leagues if the other league would accept them. The NL used this to take advantage of internal dissension in the AA to persuade Pittsburgh (1887), Cleveland (1889), Cincinnati (1890), and defending champion Brooklyn (1890) to switch leagues. After the last two, it was all over except for a desperate, last-ditch battle in 1891 over player re-assignments post Players League. Then a merger on the NL's terms. The NL of the 1890's is 12 teams, 8 of which had their roots in the AA.

The early AA is clearly not the match of the NL in talent. In 1884, the AA expanded to confront the UA which invaded and raided more AA cities than NL ones, diluting the AA. I don't know how close they got to the NL during 1885-1886. After that, they were losing teams to the NL and replacing them with minor league teams; it's quality has to be declining during this period.

I was told the following came from The Hidden Game. It compares BA.

1800s lg BWA
76 N -.013
77 N -.014
78 N -.005
79 N -.004
80 N .002
81 N .000 - I referenced all to 1881 NL
82 N .002 AA -.037
83 N -.003 AA -.027
84 N -.008 AA -.026 U -.065
85 N -.007 AA -.015
86 N -.009 AA -.008
87 N -.001 AA -.007
88 N -.002 AA -.009
89 N .004 AA -.005
90 N -.005 AA -.036 P .001
91 N .009 AA -.024
92 N .010
93 N .011
94 N .011
95 N .010
96 N .012
97 N .015
98 N .020
99 N .021

The average difference between AA and NL is 17 points but the range is from 39 to +1. It matters a lot which year(s) you're talking about. And how you translate, say, 10 points of BA to a % discount.

In my view--and I'm open to some help here--a 10 point difference in batting average (i.e. a player who hit .300 in the AA would hit .290 in the NL) is not 1/30th, but rather it is 1/x where x is the standard deviation. So 10 points is more like, what, 10 to 20 percent?

That's my take.

PS. I calculate a player discount as follows. First, I take a 1 to 1 correspondence of BA from the chart above to the percent discount. In '82 the AA was 39 points weaker--I discount the player 39%. Second, I average the discount for those specific seasons a player was in the AA. Say, Pete Browning, '82-'89. The discount for those 8 seasons averages out to 11 percent. Third, what percentage of the career in question is that? For Browning it is 8 of 12 years, 67 percent. So .11 x .67 = 7.4%, that's Browning's discount off of WS or OPS+ or whatever.

Crude, but it's where I'm at.

Like I say, help me with a better way to use the numbers above. But at least they're something.

I've been using the data that Marc has from The Hidden Game of Baseball, too. The only difference is that I use the numbers for slugging percentage instead of BA.

jimd:

Do you know how Davenport is compiling his numbers?

I'm tentatively looking at:

1. Wright
2. Keefe

3. Start
4. Richardson
5. Sutton
6. Glasscock
7. Radbourn
8. Galvin
9. McVey
10. Stovey
11. Caruthers
12. Spalding
13. Bennett
14. Welch
15. Latham

>>It does seem to me, however, that the NL had much more frontline talent than the AA.<<

I'm not arguing in favor of AA (or anything else really), but I have a questions. Couldn't this simply mean that talent was more evenly distributed in the AA, and more polarized in the NL? After all, we didn't see these guys play. Most people seem to be using OPS+ (surprisingly, to me). Has anyone looked at an OPS+ distribution among the two leagues? Just curious. Seems possible that the OPS+ range for all players in the AA was narrower than the NL, or maybe vice versa.

Jeff, I can't really support this, but I'm pretty sure that the weaker teams in the AA were weaker than the weaker teams in the NL--on average across their 10 year history, that is. Again, the variation from year to year would be significant. The Temple Cup suggests that the top teams in the AA were pretty close to the NL, but looked at team to team (rather than position by position as in Brett's post--I assume that's Brett Favre) I think the lower level teams in the AA would look pretty weak. Just a theory. I suppose I could look at a few rosters but that really sounds like work.

RICHARDSON V. SUTTON - A mini study

A couple of notes to preface this:

A few months ago, I knew nothing about Hardy Richardson. I stuff know virtually nothing biographically about him. Because he has ended up high on my ballot (no one has voted him higher than me) I feel like the defacto champion for Richardson.

I think the both Richardson and Sutton will eventually make the HoM. I have Sutton 9th on my most recent ballot, but a case can be made for him being higher, in fact, I have had him higher. I will be making some arguments below for putting Sutton above Richardson, not that they persuade me. What I dont understand, is having Sutton significantly above Richardson (I dont know if anyone actually has done this, I havent looked).

Anyway, a thought the pitcher debate went so well, I would start a 2B/3B one. I will probably throw in some Dunalp/Williamson numbers too, probably in a 2nd post. Im not presenting any raw numbers, if anyone sees any raw numbers that are important to the discussion, please post them. Win share numbers come from the positional thread from way back whenever. The warp3 peak is best 5 years, not necessarily consecutive. Hopefully this chart is readable:

Richardson Sutton
WS-Hit 259 198
WS-Field 71 74
WS Total 332 273
WS Peak 153 146

W3-BRARP 321 210
W3-FRAR 281 201
W3-FRAA 44 -32
Warp3 85.1 68.7
W3 Peak 45.6 34.8

OPS+ 130 119

Some possible problems:

1. Does warp-3 adjust for short seasons properly? Sutton had more short seasons, is he being given enough extra credit.
2. What did Sutton do before 1871? I realize I dont know the answer to this, I may need to give him a little bump.
3. Do either of these methods adjust for defense properly? It looks like warp3 gives a larger boost to Richardson between fraa and frar, which wouldnt be right with 3B being the harder defensive position. Same for these WS fielding numbers, although I think they may have it right. It looks like Sutton was a worse defensive player relative to his position, but due to position, was slightly more valuable defensively. Is it enough to make up the hitting difference? I dont think so.

With the longer career, you would think that Sutton would have been able to make up some ground on the career numbers, but, Richardson is just that much of a better hitter.

My conclusion: I may be underrating Sutton, he may move up to 5-7ish on my ballot, I dont know yet. What do others think, what have I missed, what do I have wrong?

"I stuff know"? I need to lay off the stuff. That should be "I still know" in sentence 2.

2. What did Sutton do before 1871? I realize I dont know the answer to this, I may need to give him a little bump.

He played in Forest City for the 1870 season at third. I have no idea how well he did.

Re: Win Shares.

The numbers from the positional thread don't include the NA years. Sutton would sail by True Blue if he had them. If you then give Sutton a positional boost, the difference would be that much greater.

Numbers for Dunlap and Williamson in format same as Richardson/Sutton:

Dunlap Williamson
WS-Hit 191 192
WS-Field 58 80
WS Total 249 278
WS Peak 168 143

W3-BRARP 238 192
W3-FRAR 266 285
W3-FRAA 64 46
Warp3 70.3 68.4
W3 Peak 45.1 40.9

OPS+ 133 113

Rob,

Good analysis. I think, though, that a positional adjustment has to consider more than just the fact that Sutton "played third" and Richardson "played second".

Sutton played 1263 games, over 1100 of which were at 3rd (880) or the similar shortstop (245).

Richardson played in 1331 games, but only 784 at 2B (585), 3B (178) or SS (21). A huge chunk of his games (544, or 41%) were in the outfield.

Now, this isn't all bad. In 1886, for example, Richardson was the best left-fielder in all of baseball (second best outfielder to King Kelly, in my estimation). In 1880, I think he was the second best third baseman to Roger Connor (who is not often thought of as a third-baseman either). Richardson should get the same positional boost as Sutton for playing third that year.

But the point is that a comparison of the two using the headings "best third baseman" versus "best second baseman" misses the fact that Richardson was really only a second baseman in less than half of his games. He's better described as a 2B/LF, and as such loses some of the boost that a second baseman with a 130 OPS+ would be entitled.

Note: I'm not sure how that jibes with Rob's point that Richardson is getting more fielding credit. It seems to be that it should be the opposite.

Good analysis. I think, though, that a positional adjustment has to consider more than just the fact that Sutton "played third" and Richardson "played second".

Absolutely. Hardy doesn't get the same boost in my rankings as Fred Dunlap does.

Even if you discount two years (more than I am willing to, I admit), he pitched for fourteen years! Can two years make enough of a difference to drop him below any of these other guys?

Personally, I think he is the best player on the ballot this year.

(Of course, I thought he was the best player on the ballot last year as well.)

From the BP cards, the difference in league quality is measured for pitchers by adjusting their ERA by some measurement that supposedly takes "everything" into account.

Is there a link to it on the Web?

Do you know how Davenport is compiling his numbers?

Unfortunately, he doesn't document this at anywhere near the level of detail that I'd like. I would assume it's similar to the Davenport Translations that he uses for minor-league -> major-league numbers in BP. There may be other stuff mixed in too, like the "Spalding effect" (playing on the dominant team means that you play against weaker competition) which also affects players like Caruthers who spend their entire peaks playing for a dominant team.

I'm sure that most of you know this, but a Stolen Base in the 1880's isn't the same thing as in the 20th century. There was a scoring change sometime during the 1890's that created the modern Stolen Base. Before that, it was a more general base-running metric; if you went from 1st-to-3rd on a single, it was a "stolen base", etc.

I'm sure that most of you know this, but a Stolen Base in the 1880's isn't the same thing as in the 20th century. There was a scoring change sometime during the 1890's that created the modern Stolen Base. Before that, it was a more general base-running metric; if you went from 1st-to-3rd on a single, it was a "stolen base", etc.

WHAT?!?!?!

Oh, wait... I do know this. :-D

Thanks for the Davenport info, Jim.

Tom:

Thanks to you, too. I wish he explained his system a little bit more, however.

name.......DERA...adjDERA...diff
Spalding....4.19......5.11.....0.92
Keefe........3.85......4.67.....0.82
Radbourn...3.91......4.56.....0.65
Caruthers...3.83......4.81.....0.97
Galvin.......3.86......4.52.....0.66

Clarkson...3.91......4.55.....0.64
Whitney.....3.99......4.62.....0.63
Ward.........4.06......4.66.....0.60
McCormick.3.80......4.49.....0.69
Mullane.....4.04......5.14.....1.10

A few more names. DERA is Davenport's "Defense-adjusted ERA. Not to be confused with Voros McCracken’s Defense-Neutral ERA."

Most of the NL guys are pretty consistent in the .63-.66 range;
McCormick is penalized some for his UA adventure, but still has the best DERA; Ward is rewarded because he stopped pitching before the league quality was diluted. Mullane is penalized the most because his workload-peak is in the AA's early years.

I have a feeling Davenport is adjusting way too much for his league differentials. Dick Cramer's section in The Hidden Game of Baseball included every single comparison of batting between players throughout history (min. 20 PA) up to 1980. His study was an attempt to see if hitting had improved through the years. While it failed at doing that (Ty Cobb a .289 hitter?, Babe Ruth: .260?), the built-in problem wouldn't affect yearly differences.

Palmer and Thorn didn't include pitching or fielding, but they stated that results already published at the time correlated well with the offense numbers.

You're right about discounting Glasscock's 1890 NL. WARP3 figures that to be his 5th best season, after 1882, 89, 86, and 83. His peak is scattered, in part because his defense and offense peaked at different times. His defensive peak seems to be in 82-3 when he was 22-3. His offensive peak is from 86-90 when he was 26-30 (just about where we'd expect). Like many middle IFs, he went downhill quickly in his 30's; he came up young enough that he still had an excellent career.

I think he's near the top of my ballot, possibly replacing Ward.

Also, for Dunlap, you MUST account for his 50+ WS season coming in the UA. He's got to be knocked down to about 30 WS in my opinion, even that might be a little high. When you do that, his resume really slips, as his peak looks a lot more normal. Also, his career value, which was already low, takes a hit. For me it's enough to say nice player, no vote.

A 40% deduction? That seems extremely harsh. I shaved 20% off his '84 season. I'm not saying you're wrong, but only that I have no idea which measure is correct now.

How much of an outliner is his 1884 "season for the ages" ?
I am sure someone has done the math.

If there is a study that indicates that the UA was 40% less competitive than the NL (and we can see how the numbers were compiled), that's great. Then I would have to reevaluate my rankings.

However, if we are just deducting a huge chunk off of Fred Dunlap's season because it doesn't look right compared to the rest of his career, I'm not buying. There have been numerous freak seasons that had nothing to do with the quality of the league in question (Radbourn, Norm Cash, Brady Anderson, Heinie Zimmerman, etc.)

At any rate, some deduction obviously needs to be made.

Dunlap downgrade in 1884 - here are the numbers from BP.

Warp1 18.2
Warp2 7.4
Warp3 9.6

For those unfamiliar, warp2 adjusts for league difficulty, warp3 for
season length also. Dunlap gets knocked down enormously for 1884. His 1880 is 7.8,6.3,9.8. Warp3 has '80 as his best season, with '84 2nd. 9.6 is still a damn good year.

John, I think the problem is that you can't apply the 'discount' equally . . . I think the benefit from playing in a bad league is exponential, not linear. The worse the competition, the larger the outlier years become.

What you're describing is the difference in standard deviation for the league. Obviously, this is true. McGwire and Sosa needed that extra push called expansion to help them get over that 60 homerun level. I'm not sure McGwire would have made 70 without it (which makes what Bonds did even more impressive).

How about Carew in '77? Norm Cash in '61? I've mentioned this phenomena many times on this and other threads. I was taking this into account with my rankings. The question is still how much we should deduct?

I never realized we were using batting averages for those discounts people are posting.

Nobody was using BA per se, but the league deduction for it (I was using the correction for slugging percentage).

I'm not sure how to convert them to 'wins' (I'm sure we could figure it out easily enough), but Davenport's discount seems about right to me. It still turns out to be his best year, except for 1880, which was in one of the toughest years we've covered (pre-AA NL).

Again, I'm not saying that the Davenport conversions are wrong. However, I need to know exactly what is being done to create them. I can't assume they are correct without understanding the nuts and bolts of the process.

Prelim Ballot. Note most of this is consistent with my 1900 ballot. Ward and Clarkson are off and Glasscock and McVey(He's Back!) are on. The other 13 remain on the ballot from last time and I've kept their order relative to each other the same.

1)George Wright(2)
2)Old Hoss Radbourn(3)
3)Tim Keefe(4)
4)Jack Glasscock(Newly Eligible)
5)Joe Start(6)
6)Pud Galvin(7)
7)Hardy Richardson(8)
8)Charley Bennett(9)
9)Ezra Sutton(10)
10)Al Spalding(11)
11)Ned Williamson(14)-OK I lied in that 1st sentence. This is more a statement of my downgrading Stovey and Browning than it is upgrading Williamson
12)Cal McVey(off)- The Triumphant return of Cal McVey! OK that may be overstating things a bit, but I don't have any big changes in my ballot this year and this qualifies as the biggest. I have trouble with McVey because I keep waffling on whether to give him credit for his minor league career after the majors. He would have moved onto this ballot regardless, but the reason he is 12 instead of 15 is that I am currently giving him this 'extra credit'
13)Harry Stovey(12)-I just think he's overrated when his league's are taken into account.
14)Pete Browning(13)-Ditto
15)Jim McCormick(15)

Jack Glasscock earned more Win Shares than any other SS in the UA; he only played 35% of their season, after jumping late from Cleveland NL. If he had played the whole UA season at that pace, he would have topped Dunlap for the league's best player. I don't think Dunlap's 1884 season is a fluke where he suddenly became great; I think the league really was that bad. (James dedicates a few pages to trashing the UA in the NBJHBA.)

Jack Glasscock earned more Win Shares than any other SS in the UA; he only played 35% of their season, after jumping late from Cleveland NL. If he had played the whole UA season at that pace, he would have topped Dunlap for the league's best player. I don't think Dunlap's 1884 season is a fluke where he suddenly became great; I think the league really was that bad. (James dedicates a few pages to trashing the UA in the NBJHBA.)

Nobody is saying it wasn't bad. There is no doubt about it. I deducted 20% off his season. What I need to know is how the Davenport deductions are being created. I don't take any metric at face value, that's all. It would be silly to claim a formula is the proper one when you can't duplicate the results.

If, after reviewing the system, they make sense, then I'll jump on board.

BTW, Dunlap was a terrific player before the UA and was highly regarded as such. He was the best second baseman for the first half of the eighties. He falls behind McPhee and Richardson, though.

John, I use 35% for the UA. Glasscock loses proportionally for the part of the season he was there. I think 35-40% is a better number than 20%. Just compare Dunlap '84 to his otherwise peak; his '84 is stratospheric. Note that I have Fred on my ballot based on the balance. But 35-40% is fair, IMNSHO.

Another Push for Poor "Old Hoss":

In James' New Historical Abstract he identifies the outstanding Pitchers for each season the old fashioned way (like the Cy Young award was before it was "expanded") and awards 1 per year for ALL leagues combined. The results for our leading candidates:

1876 Spalding
1877 --
1878 Bond
1879 Galvin
1880 McCormick
1881 Whitney
1882 RADBORNE
1883 RADBORNE
1884 RADBORNE
1885 Clarkson
1886 --
1887 Carruthers
1888 --
1889 Clarkson

Obviously one takes these things (along with 300 wins) with however much salt it requires to go down.

But, I do find it hard to believe that Radborne doesn't at least show up on all of our ballots. Also, as Pitchers go, is there is even good argument that anyone elgible other than perhaps Keefe, Spalding or maybe even Galvin were even as good as Old Hoss ?
Can someone who disagrees please enlighten me ?
( and I'm sorry if I've been a bit harsh here)

I posted the leaders in accumulated WS over the previous 3 and 5 seasons and career to date on the pitchers thread. Radbourn does about what you'd expect. The guy who surprised me was McCormick.

This is using James WS (no NA).

Here are the top position players in accumulated CAREER WS at the end of each season:

1876-Barnes 20
1877-O'Rourke 32
1878-O'Rourke 44
1879-O'Rourke 61
1880-O'Rourke 78
1881-O'Rourke 92
1882-O'Rourke and Anson 103
1883-O'Rourke 120 Anson 118
1884-O'Rourke 145-137
1885-O'Rourke 169-160 (Ward, no longer pitching 261 and in the lead through retirement in '94 if you want to include him, otherwise:)
1886-O'Rourke 193-190
1887-Anson 209-206
1888-Anson 238-223
1889-Anson 259-242, Connor and Brouthers coming up fast 234-224

1890-Anson 283, O'Rourke 262, Connor 259, Brouthers 243
1891-Anson 304, Connor 282, O'Rourke slipping back a bit
1892-Anson 323, Brouthers and Connor 306
1893-Anson 334, Connor and Brouthers 322
1894-Anson 345, Brouthers 343
1895-Anson 355, Connor 348, Brouthers 346
1896-Anson 371, Connor 362, Brouthers 355
1897-Anson 381, Connor 363
1898-McPhee 293, Hamilton 283
1899-McPhee 305, Hamilton and Van Haltren over 290
1900-Hamilton 321, Van Haltren 312
1901-Hamilton over VH 337-335, Delahanty moving up 318

Just compare Dunlap '84 to his otherwise peak; his '84 is stratospheric.

This is beside the point. Let's take my Norm Cash example from a previous post. While we deduct somewhat because expansion in '61 caused the standard deviation between the best and worst hitters to increase, nobody whittles his numbers down to his otherwise peak. His season still stand outs to a great degree as part of his career.

Radbourn took advantage of a change in the rules for his monster season. While he was a great pitcher before '84, does anyone think he would have had that season without the benefit of overhand pitching? Yet, we still give him the proper credit for his achievement.

With that said, you still might have a point. I have to study the issue further.

At any rate, I'm not really concerned about Dunlap and Glasscock because a further 20% reduction for one season is not going to change my rankings. I'm more concerned that I'm overstating Bid McPhee's achievements in the AA when he's eligible in a couple of "years."

I don't think Norm Cash is a great example if your trying to say a player can have 1 season that stands head and shoulders above the rest of his career. He did cork his bat after all. And, yes, I do deduct his '61 season more because of that.

I don't think Norm Cash is a great example if your trying to say a player can have 1 season that stands head and shoulders above the rest of his career. He did cork his bat after all. And, yes, I do deduct his '61 season more because of that.

If he was using a corked bat in '61, why did he stop? Nobody caught him using it that season so it doesn't make sense that he would have ended something that would put more money in his pocket. Unless it really had nothing to do with his success...

But if you don't like that example, how about Rico Petrocelli in '69?

I assume that you excluded George Wright because you are so convinced of his imminent induction that he didn't even warrant a mention?

Andrew, who's grading the grader, to paraphrase (whom?). In re:

>NA: (1) Wright; (2)McVey; (3) Spalding; (4) Pike; (5) Meyerle; (6) Bond
>AA: (1) Caruthers; (2) Stovey; (3) Browning; (4) O'Neil; (5) Mullane; (6) Orr; (7) Foutz
>All-Stars: (1) Glasscock; (2) Sutton; (3) Richardson; (4) Bennett; (5) Williamson; (6) Dunlap
>NL Pitchers: (1) Keefe; (2) Radbourne; (3) Welch; (4) Galvin; (5) McCormick; (6) Whitney

In the NA, I like Spalding, Wright, McVey, Pike, Meyerle. I don't consider Bond in that category, his peaks came after '76.

In AA, it would be Caruthers, Browning, Stovey and who cares. OK, Mullane fourth. I can't imagine spending any time on guys who clearly trail Harry Stovey.

As for NL pitchers, Keefe and Radbourn are pretty easy. If Welch and Galvin hadn't won 300, nobody would remember them today, they are freaks of their time. McCormick is the one who looks better every time I double back to consider him.

As to "the rest" (not begging the question of whether they were All-Stars or MVP candidates or whatever), Glasscock and Richardson stand out for me. I guess I'd go with Sutton third, then Bennett. I need to look at Dunlap again. I like Fred over Ed except for that UA thing.

I like your method. It is hard enough to compare 2 or 3 players from this era (these eras) much less the whole bunch. But in this bunch I still see Spalding, Wright and Keefe ahead of the pack. Then it gets hard.

Re. McCormick, I was shocked to find that for four years from 1884-87, he had the most accumulated career Win Shares. I had thought of him as a short career but he was around long enough to lead the pack, and then to stay there for four years. That is the longest of any post-'76 pitcher until Kid Nichols '98-'01.

Couple of lists for DIPS-lovers. Don't know how important DIPS stats were then comparatively, but they were certainly important to some degree. Through in upcoming HOFers Kid Nichols and Amos Rusie for comparison:

Most K/9 IP career

Rusie 4.65
Keefe 4.57
Whitney
Clarkson 3.92
Radbourn 3.63
McCormick 3.58
Welch 3.46
Ward 3.36
Nichols 3.32
Caruthers 2.86
Galvin 2.71
Bond 2.18
Spalding 0.44

Fewest BB/9

Spalding 0.49
Bond 0.49
Ward 0.92
Whitney 1.06
Galvin 1.12
McCormick 1.58
Radbourn 1.74
Caruthers 1.90
Keefe 2.18
Nichols 2.26
Clarkson 2.36
Welch 2.43
Rusie 4.10

Fewest HR/9

Spalding 0.05
Ward 0.10
Bond 0.13
Keefe 0.14
McCormick 0.18
Galvin 0.18
Rusie 0.18
Welch 0.20
Whitney 0.20
Radbourn 0.23
Caruthers 0.25
Nichols 0.28
Clarkson 0.32

Leaving aside the 1870s NA guys (who are extreme outliers everywhere), Whitney looks pretty good, judging by Ks and BBs. Keefe and Clarkson both look like 19th century Nolan Ryan lites (lots of K's, lots of BB's). And Galvin was your "keep the ball in play" kind of guy who didn't hurt himself with tons of walks or HRs.

Sounds like lots of us are using The Hidden Game chart above to make league comparisons. I looked at pp. 130-133 of my copy of The Hidden Game, and thought some additional explanation would be helpful.

First, the chart was originally generated by Dick Cramer, using his Batter Win Average formula. Then he converted the BWAs to BA and SLG. His study was questioned by Rubinstein in The Baseball Research Journal, when application of the chart showed Ty Cobb having a .289 batting average if he played in 1979. The gist of the article is that the adjustments might be fine for the average player, but they should not be applied to every individual player, or disparate results will occur. Cramer responded, and noted that his BA correction presumed a plyer drew walks and hit for extra bases at the league average, so that any above-average offense is reflected in the chart BA only as additional singles. He acknowledged that Cobb would hit much higher and that for players like that, you'd have to "fix up" his record by adjusting BA upward and SLG downward.

Based on the above, I wonder if the adjustments are valid when applied to hitters. Since they are league averages, they may validate average hitting ability in a league (and thus how easy or tough life was for all pitchers), but the adjustments don't seem to justify a one to one correlative discount to hitting numbers.

Second, I question whether we want to use BA numbers as a measure of differences in league quality. Very few of us would use BA to measure players against each other, so why use it to evaluate leagues?

Finally, the numbers in Cramer's chart on pp. 132-133 of The Hidden Game, if adjusted for 1881 National League as a baseline, seem to produce different numbers than the charts posted in these threads -- often significantly so. For example, Cramer uses 1976 NL as his baseline. The 1881 NL BA is -.107 with respect to the 1976 NL. The 1885 AA is -.107 with respect to the 1976 NL. Doesn't that mean that if the 1881NL is made the baseline, the difference between the 1885 AA and the 1881 NL is .000? Yet the posted chart shows -.015 for the 1885 AA.

I'll post a few relevant BA and SLG numbers from The Hidden Game if anyone is interested, just to compare the AA vs. NL for BA and SLG. I can probably scan the chart in and e-mail it if you send me a request.

BA:
81 N .000 reference year
82 N .011 AA -.020
83 N -.004 AA -.017
84 N .005 AA -.004 U -.039
85 N -.010 AA .000
86 N .000 AA .011
87 N -.010 AA -.017
88 N .019 AA .015
89 N .000 AA -.005
90 N .003 AA -.025 P -.001
91 N .016 AA -.020

SLG:
81 N .000 reference year
82 N -.003 AA -.010
83 N -.024 AA -.018
84 N -.010 AA -.011 U -.032
85 N .009 AA -.004
86 N -.013 AA .009
87 N -.044 AA -.033
88 N .011 AA .016
89 N -.017 AA -.019
90 N -.007 AA -.026 P -.027
91 N .004 AA -.031

'If he was using a corked bat in '61, why did he stop? Nobody caught him using it that season so it doesn't make sense that he would have ended something that would put more money in his pocket.'

I didn't realize it was in doubt. I believe Cash admitted to doing it himself. Why would he say he corked his bat if he didn't? But, your point is taken with Petrocelli.

I didn't realize it was in doubt. I believe Cash admitted to doing it himself. Why would he say he corked his bat if he didn't?

Unless he was a liar (why?), I don't doubt that he used the corked bat that year. I'm only skeptical of the bat's affect on his season's numbers.

IIRC James pointed out that the biggest flaw of the study is that you can't apply it for too many years, it breaks down. This is because the small errors when you are comparing years that are close together grow when you apply them for too long a period. So the study might be fine for comparing 1881 to 1887, but you can't apply it to 1976 because the cumulative error dwarfs the results.

Correct. I was very disappointed when James pointed this out in the HBA in 1985. However, I'm starting to be convinced that there are also problems in Cramer's project with the outlier seasons that Joe and the others have skillfully championed on this thread. Does anyone have the non-Hidden Game league conversions for the AA (and also for all the leagues in 1890)? I think I need to do some more downsizing. :-)

Thanks in advance!

1901 prelim ballot

1. George Wright
2. Jack Glasscock
3. Ezra Sutton
4. Hardy Richardson
5. Al Spalding
6. Tim Keefe
7. Harry Stovey
8. Pud Galvin
9. Hoss Radbourn
10. Ed Williamson
11. Joe Start
12. Cal McVey
13. Bob Caruthers
14. Charlie Bennett
15. Fred Dunlap

I think that Wright and Glasscock are the cream of the crop and deserve high placements from all voters this week. I cannot believe that some voters are leaving Wright off their ballots completely.

Continuing with some WS numbers, here are the top 3 and 5 year accumulated WS as of the end of each year, for position players.

1878-O'Rourke for 3 at 44
1879-O'Rourke and White 44
1880-Hines 56, O'Rourke 78
1881-Hines 57 and 78
1882-Anson 60, Hines 90
1883-Brouthers 59, Hines 92
1884-*Dunlap 68, Hines 98
1885-Brouthers and Connor 72, Brouthers 107
1886-Connor 89 and 127
1887-Connor 87, Connor and Brouthers 129
1888-*O'Neill 91, Connor 142
1889-O'Neill 91, Connor 145

1890-Connor 83 and 140
1891-Hamilton 85, Brouthers 129
1892-Hamilton 86, Ryan 129
1893-Duffy 85, Hamilton 130
1894-Duffy 90 and 144
1895-Duffy 84 and 141
1896-Hamilton and Jennings 89, Hamilton 134
1897-Jennings 94, Hamilton 137
1898-Hamilton 91 and 150
1899-Delahanty 97 and 161
1900-Delahanty 93 and 149
1901-Wagner 97, Delahanty 151

*If you take the league discounts for Dunlap and O'Neill ('88 but not '89) each would trail--Brouthers with 66, and Connor with 89, respectively. O'Neill's lead is large enough to stand up to the discount in '89).

No player was good enough, long enough to lead in accumulated career WS while still at their "peak"--i.e. ahead on the 3 or 5 year list as well. Well, OK, there were three--O'Rourke in '78 and '79 though that is a special circumstance as nobody had more than a 3 or 4 year career at that time. Anson who tied O'Rourke for the career lead in 1882 then fell back again until 1887. And Billy Hamilton in 1898. With the retirement of Anson, Hamilton came in second to McPhee by a mere 293-283. Actually if you apply the discount to McPhee, then I think Hamilton moves ahead in both '98 and '99 when he trails McPhee 305-298 but of course by '99 Hamilton himself is past his peak.

The next closes call is Connor who trails Anson 259-234 in '89 when Connor is the top 5 year man. In '91 Brouthers trails Anson 304-272 but both Connor and O'Rourke are in between.

I guess a good definition of an all-time great would be a player who was good enough long enough to lead all active players in accumulated career WS while still good enough to be the top current peak player at the same time. I wonder if anybody ever did that ;-) Well, of course, some pitcher probably did it in the 19th century.

PS. Yes, Bond in '80, Ward '81 and Nichols in '98-'99-'00, did it, though Nichols in '99 was tied with Delahanty in the 3 year category. It was only about then that position players began to compete with pitchers in these measures.

Well, while we're taking veiled shots at people's credibility, I'd like to say that I cannot believe that some voters are taking evidence that George Wright hit .600+ in 186x and then applying it to the ballot as though that were indicative of anything.

Who is doing that? I know I'm not. I haven't read another post even suggesting this. That would be beyond idiocy.

I do think he was the best shortstop in the sixties. There is no evidence available to refute this. If there is, I'd be happy to see it.

For all I know, he may not be major league material today, but how many players on your ballot or mine would make it today either? Who cares? We're not ranking the greatest players of all-time from 1 to 100. We're trying to highlight the best players from each generation.

Mark Mc,
No doubt from anyone here that you know your stuff.
But how much "boost" does Wright need from 1860s play to get into the HOM? He probably was the best SS for several years there, but even "good" might get him in, no?
Assuming he was A-Rod of the time might be speculative. But assuming no value, or little value, is equally speculative (and more likely to be inaccurate).
THIS contest wants us to try to make the most reasoned estimates of a player from that time. If you ran a contest, I doubt that aspect would be featured. But we can only go by what THIS contest asks.
Now, if it's Dickey Pearce we're talking about, a couple of voter's refusal to list him simply costs him 15th place or whatever.
But with Wright, it's keeping him out of the HOM at the expense of players whom a majority of voters consider slightly less attractive choices. Doesn't require you to change your vote, but does suggest a consideration of the specific rules involved.
Is that fair?

My view on right is pretty clear. He was among the best statistically among shortstops (and often among all positions) during eight of the first nine years of organized pro baseball.

He was anecdotally the best for two years before that -- anecdotes that I put more credence in because of his subsequent performance. A player who is the best at his position for a decade -- and among the best for every single year of that decade -- deserves a ballot position irrespective of the position or time period. When, in each of those years he was playing against the best opponents available, and for several of those years he was among the best players in baseball, he moves right to the top.

Now, I'm certainly not trying to gang up on Mark, who has a fair ballot -- his last one did not include any NA players. Obviously he is using a steeper timeline adjustments than most others.

But the way I see it, George Wright was better than Ross Barnes, and Barnes is in. Barnes is also the only player that got in so far without being named on everyone's ballot (although he was first on Mark's!) I'm guessing that Wright will be the second (and not the last).

The prelims make it pretty clear that it's a three-man race with Wright and Glasscock among the three. It seems to me that the two major differences (not the only ones, certainly) are that 1) we have Glasscock's complete record but not Wright's, and 2) a timeline difference of about 13 years.

Both played about 17 years. Wright's documented OPS+ is 123, Glasscock's 112. Whether Wright's 1866-70 OPS+ is closer to 123 or 112 we don't know but it was probably somewhere in there. Based on all the evidence (not just this one number), one would be very hard pressed to say Glasscock was a more valuable offensive contributor per PA or whatever denominator.

On defense WS has Glasscock at A- but does not rate Wright. Wright in his decline phase after 1876 earned about 19 defensive WS in 329 games or one every 17.3 games. Glasscock earns 91 defensive WS in 1700+ games, or one every 19 games. (On FR, it's Wright 1 FR for every 6.4 games and Glasscock 1 every 7.2 games.) Again I don't see among the available evidence anything that says Glasscock was clearly better on the rates and some reason to think maybe Wright was better per opportunity.

Which brings us to season length. On career value, 1700 games versus 600 documented--but probably another 1000 undocumented for Wright--but ignoring that--if Wright was 10% better per game but in 1/3 as many games, advantage Glasscock. But if you consider peak value--if you do--it certainly closes the gap, depending on how you value career and peak.

So Wright leads on DOCUMENTED peak and rate, Glasscock on career. So finally it comes down to:

1) how you value career vs. peak vs. rate stats, and there is no moral authority on that, and

2) how you value short seasons. Some people crank them up to 162, others obviously don't. I don't know of any overriding authority on that, and

3) timeline, but in this case we're only talking 13 years, and

4) finally how you consider undocumented (pre-'71) play--some appear to value it at zero which flies in the face of the documented evidence (lacking statistical dimension, of course) that Wright DID play and was one of the very best.

In short, Wright has clear advantages over Glasscock on documented peak and rate and there are no unknowns, I don't think, that could push Glasscock ahead. Glasscock's career advantage can on the other hand be argued depending on your view of pre-'71. In the end, one might reasonably discount Wright's advantages on the various bases (1-4) just mentioned. One might equally (I would say MORE) reasonably not discount Wright on every single variable and come up with George ahead.

If I completely ignored everything before 1871, Wright would still be no lower than 6th or 7th on my ballot. He didn't need to be A-Rod before that, just 3 or 4 years of Edgar Renteria. That's a conservative estimate IMO and is more than enough for me to move him to #1. To leave him off my ballot entirely, I would have to assume that he was not only a terrible player pre-1871, but that he got enough PT to significantly hurt his teams during that period. You're entitled to your opinion, Mark, but the 33.5 people who voted highly for Wright didn't require him to hit .600 in 186x to put him there. If people were assuming that, he'd already be in and this wouldn't even be an issue.

BTW, shouldn't there be a BALLOT around here somewhere?

>The statistical evidence as I see it presents a picture where much of Wright's value came
from playing with people that just weren't all that good.

This of course is the timeline argument. The question is how steep a timeline. I could argue the same point re. every 19th century player. Given that Wright is just 13 years older than Glasscock, we're basically talking Ted Williams vs. Mickey Mantle by analogy. I can't imagine Glasscock can rate very highly on your ballot by this logic. Or Keefe, or Stovey, or Richardson, or Radbourn, or....

PS Mark: I would guess that in total you rank AA players more highly than anybody. Stovey's and O'Neill's and Browning's and Caruther's et al competition may (or may not) have been better than George Wright's, but at least Wright from '69-'79 played against the best competition available anywhere. The AA guys clearly did not. One needs a timeline to discount Wright. One does not even need a timeline to discount the gentlemen mentioned above.

PPS. And just to close out the record--and to completely whip the dead horse's eyes and make him sleep... Mark M. voted Harry Stovey #1 in '00. Stovey is nine years younger than George Wright. Nine years! Wright played the best available competition, whom Mark designates as "people that just weren't all that good."

Nine years later playing against in the second best competition available, Stovey played against competition that was so much better that he ranks #1 and George Wright lower than #15???

Is it just me, or are all of your posts showing up with a time 1 hour and 15 minutes after you posted them?

BTW: I have not posted anything under the surname of "Horse".

BTotherW: I have posted my ballot and am immediately antsy. I've got Dan Brouthers and Buck Ewing entering my ballot next week. Am I missing anyone I should be considering?

I likewise have posted nothing under any other name than the one above. Twice as many at bats are nice, and I've said that how you want to deal with season length is your business. The question before the house was how much tougher the competition got in 9 years that Wright the elder is not on some ballots at all while Stovey the younger is rated #1 while not playing against the toughest competition of the day.

I think the clock is off. And with Brouthers and Ewing I don't think it makes any difference in the short term who else there is, but I don't see anybody else of substance.

My advice is hold on to your ballot a while ;-)

The following are what appears to be Davenport's "Adjustments for all time" that turn DERA to adjDERA. Per usual, 1881 is the base year. (It's also the strongest 19th century season except for the "super season" of 1900 when the 12 team NL contracted to 8 teams.)

71n .49
72n .53
73n .41
74n .39
75n .71
76N .45
77N .16
78N .35
79N .36
80N .03
81N .00
82N .06 82a 1.56
83N .12 83a 1.11
84N .51 84a 1.20 84u 1.83
85N .34 85a .81
86N .39 86a .74
87N .17 87a .62
88N .22 88a .53
89N .11 89a .49
90N .41 90a 1.22 90P .26
91N .01 91a .56
92N .10
93N .13
94N .10
95N .13
96N .18
97N .11
98N .04
99N .03
00N -.34

Note that in Davenport's "opinion", the AA never attains parity with the NL, though at its best, the AA does reach a rough parity with the NA (average .51). In its weaker years, it is midway between the NL and the UA (and in 1882, not much better than the UA).

Also, in his opinion, the Player's League is the best of the three 1890 leagues, though not by as much as I thought. The AA was badly hurt by both team and player defections, though it managed to recover for 1891, aided by many members of Boston's defending PL champions, who formed the nucleus for the AA's Boston entry.

It is clear, then, that Davenport is not using a narrow standard deviation of OPS as the definition of league quality. In the 1880s, the standard deviations of OPS among NL and AA were not always in the NL's favor on a year-by-year basis. The UA is also not as bad using the standard deviation analysis (I AM NOT arguing that the UA was good -- just that standard deviation of OPS does not accurately reflect how bad it was).

Another thought: All of the charts compare to 1881 NL as a baseline. Everyone is docking the AA for being worse than the 1881 NL, but nobody seems to be docking the 1882NL, 1883NL, etc. for being worse than the 1881 NL. In some years, the AA appears to be stronger than the NL in the same year, but I don't see anyone applying NL discounts for those years.

The Hidden Game numbers had the AA better by .001 one year (I think it was '86). The above numbers do not have the AA better, ever.

Besides, the "docking" is not vs. NL '81. It is vs. the better contemporaries. I don't discount AA '86.

Docking vs. '81 is more like a reverse timeline, and I do not use a timeline. A pennant is a pennant (a season is a season). It is clearly a different thing to compare a weak league to a strong league within a season.