Prelim:

1) Ezra Sutton (2): I love McPhee as you can see, but how does he wind up before Sutton? If we had Win Shares for the NA, he would have around the same number as Bid (at least), plus Sutton's peak was much greater than McPhee's. I'm curious to hear the arguments for Bid over Ezra.

2) Bid McPhee (2):

3) Al Spalding (3):

4) Cal McVey (4):

5) Dickey Pearce (5):

Everybody moves up one after Pearce:

6) Joe Start (7):

7) Charlie Bennett (8):

8) Billy Nash (9):

9) Jack Clement (10):

10) Ed Williamson (11):

11) Fred Dunlap (12):

12) Levi Meyerle (13):

13) Lip Pike (14):

14) Pud Galvin (15a):

15) Mike Tiernan (n/a): I think he's slightly better than Sam Thompson, but close enough to argue the other way. Best major league rightfielder for 1888 and 1889 (close in '91). Best major league centerfielder for 1890. About equal with Stovey for best NL rightfielder in 1891. .

My very-PRELIM ballot. I have some research to do to clean up the bottom of the ballot. Sutton is close. I currently have Weyhing as the only newcomer, but I am not sure he will stay this high, or even on the ballot. He is not HoM reagrdless, but none of the guys that low are, imo.

1. Galvin

I know there's no one that good that's newly eligible, but does someone have a list of the 'newbies' anyway?

So, I'm thinking about Bid McPhee, and I'm looking at his WARP card, and I'm thinking, "Hey, it looks like McPhee had a late peak."

What I'm noticing is that McPhee's WARP-3 in his last 4 years in the AA was (on average) 6.2, while in his first four years in the NL, it jumps to (on average) 7.9. Pretty big improvement for ages 30-33 compared to 26-29.

But then I look at his WARP-1 numbers, and I see that, actually, his numbers unadjusted numbers for 1889-9 average 9.55 Wins, and for 1890-3 also average 9.55 wins. He has essentially played exactly as well, but got docked more for his competition during the AA years, making it appear that he improved in the NL.

So, now I'm thinking about why Bid McPhee is playing about the same irrespective of what league he plays in, rather than regressing against stronger NL competition, and I start thinking about league discounts.

Now, it makes sense to discount McPhee's hitting if you believe that he was facing poorer pitchers in the AA. (McPhee, in fact, has a 111 OPS+ in his last four AA years, and a 114 OPS+ in his first four NL years).

But why, exactly, does it make sense to discount McPhee's fielding stats for playing in an inferior league? Did lower calibre AA hitters hit more weak grounders (if so, wouldn't it be HARDER to turn all those double plays)? Probably not, because they were batting against lower calibre pitchers, too.

So, it struck me that maybe defense is a little more context-neutral than hitting or pitching is. I mean, certainly if your team is full of bad defenders, you'll get more defensive plays and stuff, but that's not really relevant for McPhee, because he was playing with the same group of guys in Cincinnati irrespective of the league (the whole team changed leagues). It was just the opponents that were changing.

So, the question is, does a discounting stat like WARP-3 systematically undervalue fielding accomplishments by discounting them just like hitting accomplishments?

Win Shares, we know, systematically undervalues early defense by giving too much of the credit to pitchers. Does WARP-3 do the same, but discounting hitting and fielding stats the same?

Cribbed from someone else, not perfect but probably close...

1906

My very-PRELIM ballot. I have some research to do to clean up the bottom of the ballot. Sutton is close.

Here's my question from above:

I love McPhee as you can see, but how does he wind up before Sutton? If we had Win Shares for the NA, he would have around the same number as Bid (at least), plus Sutton's peak was much greater than McPhee's. I'm curious to hear the arguments for Bid over Ezra.

Sutton has to be near McPhee.

12. Weyhing

Huh?!? :-)

Seriously, no real peak to speak of, while his career numbers are nothing special. I questioned Rusie as a HoMer, but he was unquestionably a great pitcher at his peak. Weyhing is half of Rusie.

Cribbed from someone else, not perfect but probably close...

1906

When we say "1906 election", we're voting in January or February of 1906, right? Radbourn and Richardson were 50 and 49 years old when when elected them. (Well, Radbourn was dead, but you get the idea.) We elected Rusie at the age of 32, and that made some of us uneasy. We shouldn't be letting players gain an advantage in an election by retiring young and thus avoiding comparisons with some contemporaries.

What follows is a list of players who have appeared on ballots or in arguments, listed by age as of Jan. 1, 1906. I would not want to mandate any lower age limit, but I would suggest that we exercise caution with everyone who is younger than about 45.

Age Players

Preliminary:

Spalding's batting - I haven't in the past given him enough credit for that. I'm a FOPG, but he kicks Pud's butt on that front, and I'll have to weigh that in. These guys got a LOT of at-bats, so it mattered a lot..

1906 Very Preliminary

Since the elections drew from the middle of my list, the top five are unchanged. I think these five should go in, in this order, as soon as it can be suitably arranged with the arrival of new candidates :-) .

1. Pud Galvin

Fixing my last post.

I accidentally left off McVey and Browning. In January 1906, McVey is 55 (same age as Spalding, Sutton, Barnes, and O'Rourke) and Browning is 44. Well, Browning died just this year, but he would have been 44.

Bert Cunningham is 40, Ted Breitenstein is 36, and Frank Killen is 35. They're all too young to take seriously yet, which isn't the only reason not to take them seriously. But they're all older than Rusie.

Preliminary ballot ? Thanks and a tip of the hat to Andrew Siegel for a great way to format a ballot:

The guys it would be a mistake to exclude:

None ? To an outsider looking in, there would be no glaring omissions if none of these guys ever made it.

Guys who should probably be in:

Does anybody know what's up with WARP's fielding runs for pitchers? I haven't looked systematically, but every pitcher I have looked at closely looks pretty much like a replacement level fielder. Have I just not looked at the right players, or does WARP basically treat replacement level and average level as the same for pitchers, and then deduct runs if they look like bad fielders??

Sorry. Post #10 is from me, not Matt B. I meant to address it to Matt B and somehow put his name instead of mine in the "Name" field.

Brain cramp at work.

Does anybody know what's up with WARP's fielding runs for pitchers?

As far as I can tell, they don't evaluate the fielding of pitchers as pitchers (there are no ratings or fielding runs associated with the raw fielding stats in the pitcher's section). IMO all of the fielding runs are the results from fielding other positions, which for most pitchers in most years are just token appearances.

Hopefully this is the correct thread to ask this:

Not to sound like an EOPG, but why is it that quite a few people think he's HOM material?

1. Yes, he pitched ALOT of innings for a long time, and that certainly counts. But Jesse Orosco pitched a lot of games, and Jack Quinn and Charlie Hough pitched a lot of innings in a later context.

2. I keep seeing "he didn't have good defense behind him", but where is this coming from? Over half the teams he played for were winning teams, and with defense being so important in the 19th century, it's hard to think that these teams were pathetic enough on defense relative to every other team to make a huge difference.

3. Overall, his ERA+ is 108, and when you consider that defense contributes more value in the 19th century to preventing runs than today, maybe that's equiv. to 104 for a modern pitcher?

Again, I'm not arguing against him, but I'm wanting to understand why he's being considered as more than just a good player that played for a long time and had one really great year (1884).

Continuing the thought on demographics and date of birth:

Some good players born 1861-1865:

J.Clarkson, P.Browning, B.Caruthers, J.Clements, M.Griffin, B.Nash, B.Joyce, B.Cunningham, M.Baldwin, C.Buffinton, H.Boyle, D.Casey, J.Ryan, D.Hoy, D.Johnston, B.Sanders, F.Carroll, H.Collins, J.McTamany, T.McCarthy, T.Ramsey, T.Tucker, G.Pinkney, O.Burns.

Some good players born 1866-1870:

B.Hamilton, B.Dahlen, G.Davis, C.Childs, C.Young, K.Nichols, E.Delahanty, J.Burkett, D.Lyons, G.Weyhing, M.Tiernan, J.Stivetts, T.Breitenstein, F.Killen, G.Van Haltren, B.Rhines, J.Beckley, S.King, S.Stratton, S.McMahon, E.Smith, M.Kilroy.

I apologize for this list shorting infielders and catchers - I was just quickly scanning offensive and pitching league leader lists on bbref.

It would seem that the 66-70 group was stronger and had more and better HoM candidates than the 61-65 group. Given what was happening in the U.S. in those years, that's not too surprising. But from where we are now, we've already elected Clarkson are considering most of the rest of the candidates from that time, with perhaps only Jimmy Ryan and Dummy Hoy not yet eligible. By contrast, the bulk of the good candidates born 1866-1870 are not eligible for consideration yet, and I'm not inclined to look very hard at the likes of Tiernan and Stivetts just yet.

Overall, his ERA+ is 108, and when you consider that defense contributes more value in the 19th century to preventing runs than today, maybe that's equiv. to 104 for a modern pitcher?

No, I think the influence of the error rate would be on the confidence that one would place on ERA and ERA+. "ERA fielding support" becomes kinda like "offensive run support", some years you get a lot (ample errors at "opportune" times so that oodles of runs become unearned), some years you don't (many errors that don't impact ERA at all). It's an issue with single-season ERA and ERA+ crowns, but a career should even that out, I would think.

108 ERA+ still mean 8% less earned runs than the average pitcher. I don't think it's that valid for comparisons because the average pitcher of 1880 is about the 6th best pitcher in the majors, while in 1888 it's about the 25th best pitcher in the majors; expansion made a huge quality difference in a very short period of time.

Notes:

John C wrote: <i>I don't see how you can argue that Cal McVey has career value close to McPhee.

Adjusted WARP3- McPhee has 108, McVey has 62.4

John,

Thanks for pointing out the faulty ratio I was using. It's clear I'm using one that's too high. I'll look at the numbers more closely, and pick a new multiplier. Since WARP really doesn't give McVey much defensive value (they rate him a pretty poor catcher!), I think the multiplier will end up closer to 2.75. If that turns out to be about right, I'll be dropping McVey down a bit; I'll need to look again at Spalding and Start as well, whom I've placed in the rankings in part upon conversions from WARP to WS, as well as estimates for undocumented portions of careers.

<i>14. (Tie) Ezra Sutton

And, I am contemplating voting for Jack Stivetts at 15.

Well, he's better than Gus Weyhing. Not too much of a recommendation. :-)

Prelim: 1906

1. Ezra Sutton (1)--Ahead of the field on both career and peak value. I?ve never been able to understand why one wouldn?t at least put him on the ballot.

2. Bid McPhee (2)--Defense and career value trumps the AA discount and the lack of a tremendous peak.

3. Joe Start (4)--More career value than McVey. But a lower (documented) peak.

4. Cal McVey (5)--I like the Ross Barnes comparison a lot.

5. Harry Stovey (6)--I think some people have been applying an awfully harsh AA discount to him. He was a tremendous hitter and looks great in WS pennants added and in the baserunning info that?s been posted. More career value than any of the other ?hitters? on the ballot.

6. Pud Galvin (9)--Bumped ahead of Pike this year, giving more credit for career value over peak.

7. Lip Pike (7)--Not as good in the NA as McVey, but better before.

8. Charlie Bennett (10)--Great defense at catcher keeps him in the middle of the Outfielder/Pitcher Glut.

9. Al Spalding (11)--Here for his hitting and the adulation of his peers. This low because of the defense behind him, the hitters on his team compared to the competition and the amount of credit I give pitching vs. fielding in the pre-93 era.

10. Pete Browning (12)--AA discount brings him down to Thompson and Tiernan and Griffin?s level. Browning still has the higher peak though.

11. Mike Tiernan (13)--I don?t think 3 players could be any more equal than Thompson and Tiernan and Griffin. Tiernan has a slight peak advantage over Thompson.

12. Sam Thompson (14)--Lower peak than Tiernan, higher peak than Griffin.

13. Mike Griffin (15)--Defense brings his (relatively) low peak slightly below the rest of the outfielder glut.

14. Bob Caruthers (-)--Lack of newbies brings Parisian Bob back to the ballot. . .

15. Jim McCormick (-)--Along with McCormick, who hasn?t made my ballot since ?98.

It seems odd to be defending someone I DON'T think belongs in the HoM, but that's what happens when you are talking about people on the bottom of the ballot.

Weyhing did have a peak. He had an adjERA+ of 133 early in his career. He also had a decent career length and was solid throughout.

I will reiterate though, that even if he appears on my ballot, he is well below what I consider the Hall cut-off point.

As for Ezra Sutton, I'm still working on him, but it looks like he is moving up.

Does anyone who did NOT start with rankings off WARP1 or WARP3 or VORP or anything else believe that Start, McVey, or Sutton don't belong on the ballot?

'So, there's no real consistent number to convert between adjWARP1 and adjWS'

Could this be because you're trying to go from a system that is based on Replacement level and converting it to a system based on a guy who didn't play? I'm not an expert on this sort of thing, but wouldn't you have to select a RP for WS for each season. Then convert adjWarp-1 to adjWS above RP. Then you could add in the RP to get adjWS. Maybe there's a more consistent ratio if you do it that way. If there is a reasonably accurate way for this to work, I would be interested because I would eventually like to convert to using WS. I just don't feel comfortable with WS for 19th century players, particularly those who played in the NA. If I could convert WARP into WS for 19th Century and then use actual WS for 20th century, that would be ideal. I'll stop babbling now.

1. Pud Galvin (1) ? Is this his year? He was first runner-up in 1905. If he makes it, it will be an impressive climb from coming in #12 on the opening ballot, and a point in favor of deliberative democracy.

2. Joe Start (2) -- Still holding strong. One day we'll get there.

3. Al Spalding (3) ? best pitcher, 1867-1876 with no close second. I simply refuse to believe that pitching was so unimportant that no candidate is worthy.

4. Bid McPhee (4) -- I try to be as conservative as possible with the new guys, but putting McPhee any lower than third goes beyond conservative to reactionary. A close study of McPhee has led me, however, to move Hardy Richardson up in my estimation.

5. Ezra Sutton (8) -- I've had him this high before, but had a change of heart. The heart is changing back now.

6. Bud Fowler (6) - the best Negro league player to retire in the 19th century gets precedence over the fifth best first baseman/ left fielder until I hear evidence to the contrary.

7. Harry Stovey (7) ? a great player, but at deep positions. Still not sure about him, but he moves up in my estimation this week.

8. Cal McVey (10) -- flipped McVey and Bennett, which is no slight at all to Bennett.

9. Charlie Bennett (9) ? should go in eventually, I think, but I was oversold on his defense before. The truth is likely somewhere between WS and WARP.

10. Sam Thompson (15) -- better than I though.

11. Bob Caruthers (12)

12. Pete Browning (13)

13. Mike Tiernan (15)

14. Lip Pike (off) -- bouncing around the in/out line.

15. Mickey Welch (off) -- first time on my ballot, but worthy of the 15th spot today.

Dropping off for now: Mike Griffin goes down to 16th.

It seems odd to be defending someone I DON'T think belongs in the HoM, but that's what happens when you are talking about people on the bottom of the ballot.

Sorry for being a pain in the butt, but THIS TIME IT COUNTS! :-)

Seriously, this is going to be the toughest election yet, so we need to make sure we have the right guys placed correctly (however that is).

Weyhing did have a peak. He had an of adjERA+133 early in his career. He also had a decent career length and was solid throughout.

With all due respect, no to both points. Was he ever close to being the best pitcher for any particular season? No. Was his career numbers close to being the best of his time? No.

When he had his 133 ERA+, the top guys were averaging 179. The highest ranking he ever attained for that statistic was the fifth position. Very good, but not great.

As for Ezra Sutton, I'm still working on him, but it looks like he is moving up.

I'm glad you are recognizing him more.

I don't know. Even though the article reads like a refutation, it actually is very supportive of DIPS. On the other hand, since it only uses data from 1913 to the present, I don't see how it could relate to Galvin one way or the other.

Just so you know there is no strategic voting going on:

Because I am adding Sutton to my ballot (possibly as high as 5th, as I am starting to be convinced of his worth), if I decide to keep Weyhing on the ballot, it is at the expense of Bud Fowler, and not someone aho has a chance of being elected this year.

I think anyone who leaves one of those 5 candidates off their ballot should have to justify it

Whew! I'm safe! :-)

Just so you know there is no strategic voting going on:

I didn't think you were doing that at all. You had Sutton tied at #15 for the last election, so you were consistent for your prelim.

BTW, I'm not above being challenged on any of my picks. I have changed my rankings in the past, so none of my picks are carved in stone. I welcome any input from anyone here.

Because I am adding Sutton to my ballot (possibly as high as 5th, as I am starting to be convinced of his worth), if I decide to keep Weyhing on the ballot, it is at the expense of Bud Fowler, and not someone aho has a chance of being elected this year.

If you honestly feel Weyhing belongs, then do what you feel is right. My name is not Frankie Frisch, Jr. I don't have all the answers.

I think your multiplier of 3*WARP1 = WS is too high.

I've been playing with a factor of 3 and have the following observations:

1) WS favors offensive players relative to WARP-1 because of it's low "replacement level".

2) WARP-1 favors players at key fielding positions; this is what you'd expect when the pitcher's value is redistributed around the team due to the different pitching/fielding balance. Catcher gets a big boost.

3) Large discrepancies exist when a player has a fabulous fielding season (see Richardson 1881). I guess this is due to all of the cutoffs built into WS, causing it to refuse to credit what it sees as extremely good (or bad) fielding numbers that lie outside of the limits imposed on each of the individual fielding skills.

4) BP uses 5 year averages for park factors; WS uses single-season factors but "regresses" them by adding 100 runs and 10 games to the raw home and away numbers. There are problems with both techniques but there is no commonly accepted solution to this problem either (that I know of).

I thought I would share the process I have been going through with older players.

I really like stats. I have a problem when it comes to analyzing players without proper stats. I refuse to ignore players that have too much time before the time things like WS kick in, but I also refuse to give them standing until I am comfortable with them.

I never voted for George Wright (actually I think I had him 15th once.) but he was voted in before I got on track with him.

I initially didn't vote for Al Spalding or Joe Start.

Start was first for me, as I got his worth straight in my head. I now have him 3rd on my ballot.

Then came Spalding. I now have him 2nd.

This week it is Sutton. I recognize that I was using a "partial" career for him, but that is all I was willing to use until I was comfortable with the rest of him.

This week I have him 5th.

So who is left?

Well, I would be happy to hear convincing arguments for Cal McVey, Lip Pike, Dickey Pearce, and Levi Meyerle because they are getting votes from people and I don't even have them on my radar.

In short, I am happy to hear criticisms because my opinions are like a constantly changing river: you can take a snap-shot but you are only getting that moment in time. :)

Carl Goetz wrote:

Which leads into the question whether Win Shares or WARP are even the correct metrics when you're trying to identify the BEST players. I think measuring "ABOVE AVERAGE" or something even higher is probably the way to go.

For the most part, I'm going with a combination of Win Shares and WS per season. By doing this, it tends not to overstate the contributions of either the long (Pete Rose) or short (Sandy Koufax) guys.

short career guys

Which leads into the question whether Win Shares or WARP are even the correct metrics when you're trying to identify the BEST players. I think measuring "ABOVE AVERAGE" or something even higher is probably the way to go.

KJOK, It seems to me that these measures do a fine job of identifying the best players -- they have the most wins or win shares each season, and if you want to use them to measure "above average" play, you can figure out an average value and subtract it out of the totals.

I think the difference in ratings is not a product of the statistical systems folks use, but a product of different ideas they have about how "above average" is to be valued (not how it is to be measured).

VERY PRELIMINARY BALLOT, SUBJECT TO CHANGE:

1. CHARLIE BENNETT, C, Comp is Roy Campanella. Better hitter and fielder than Clements. Until Roger Bresnahan, only Ewing was a better Catcher. Unless you?re taking the position that almost NO catchers were among the most valuable players in the 19th century, I find it hard not to advocate Bennett.

<i>Counting only seasons of 10 G or more, I have it at:

1. Ezra Sutton

jmid wrote:

"HOM All-Star team of the 1870's

<i>HOM All-Star team of the 1870's

P. vacant (A. Spalding?)

3B/C Deacon White*

Good try, KJOK, but White didn't play third until the eighties. You're not knocking Sutton off that easily. :-)

BTW, I'm surprised you don't like Sutton since he had a great peak.

11 (-) Ned Williamson--And here's where it gets interesting. I believe I severely underrated 3rd basemen, but I'm not sure it mattered. I picture Ned as the better all-around player in the Ezra v. Ned debate. I put Ned 15th in the 1898 ballot as a nod to that, but frankly I don't see Ned as being a HoMer.

I'm not using this to beat you up, Mark, but as a springboard for some thoughts on Sutton/Williamson.

Without adjusting for schedule or including his NA seasons, Ezra follows Ed by only 14 Win Shares. However, he beats Williamson in WS per season 24.98 to 23.34. Since their WS seasons encompass the same league and era, their respective competition is not a factor (except for maybe 1876).

The funny thing is the NBHA still has Sutton with the better peak: top three seasons - 71 to 60; top five seasons - 98 to 77. Those are unadjusted numbers.

Now baseballreference.com and WARP don't adjust for Williamson's 1884 season for the crazy park effect that year, which leaves him ahead of Sutton (which is ridiculous). Lake Front Park had one of the major park adjustments of all-time that year. James takes it into account.

If you factor in Sutton's NA seasons, he zips by Williamson easily career-wise. If you take peak and career, Sutton tops Williamson.

I can't see Ed above him.

Do you mean that the discount for the (N)ed's home park is not severe enough to account for the 1884 season?

Bingo.

How did you determine that EQA/WARP doesn't discount enough for the home park?

From baseballreference.com:

<i>We use a three-year average Park Factor for players and teams unless they change home parks. Then a two-year average is used, unless the park existed for only one year. Then a one-year mark is used. If a team started up in Year 1, played two years in the first park, one in the next, and three in the park after that and then stopped play, the average would be as follows (where Fn is the one-year park factor for year n):

Year 1 and 2 = (F1 + F2)/2 Year 4 = (F4 + F5)/2

Some of BP's stuff is explained in their glossary. It requires some careful reading though and not all is revealed.

The upshot on BP's park factors is, 5 year average. The link contains another link to a table showing the park factors they use for each year. Enjoy.

Thanks, Jim. I don't know how I missed that.

1883 CHI-N 1084 1062

Sorry Joe. It's Win Shares that uses "regressed" single-season park factors before 1909. Here's a quick sample for Louisville, which is nearly neutral in 1876 and a hitter's heaven in 1877. First number is a single season factor (ready to use), second number is the average (which is actually used).

BProspectus

John - I think you need to rethink your park adjustment issue with Ned Williamson. Even if a method uses a 3 year adjustment instead of a single season, Williamson played ALL OF THE YEARS surrounding 1884 with Chicago, so if he's not getting "penalized" enough by the factor in 1884, he's getting "overpenalized" by the same amount in the surrounding years. And since his # of plate appearances in those years are almost identical, the net effect on his actual adjusted stats is practically ZERO!

John - I think you need to rethink your park adjustment issue with Ned Williamson. Even if a method uses a 3 year adjustment instead of a single season, Williamson played ALL OF THE YEARS surrounding 1884 with Chicago, so if he's not getting "penalized" enough by the factor in 1884, he's getting "overpenalized" by the same amount in the surrounding years. And since his # of plate appearances in those years are almost identical, the net effect on his actual adjusted stats is practically ZERO!

You're 100% right - career-wise. The problem is BP and baseballreference.com, by using the 3 year average, distorts the peaks and valleys of his career.

If we were to take his best 3 OPS+ seasons with baseballreference.com, they would be 1879 (151+), 1884 (169+) and 1882 (132+). If we adjusted the '84 season, we would still have the '79 season to use for his 3-year peak, but the other two would probably be no greater than 130+. His peak is totally distorted by not adjusting for the huge park factor of '84.

As we're all trying to draw the firmest distinctions we can among players we've been looking at for a while, there's a lot of talk about peak (career value, I guess, being more straightforward with only matters of season length, league quality, time-line adjustments, and estimates for undocumented play to consider . . . ) but I often find myself unclear about just what posters mean when they talk about peak. How do they measure it, actually? How do they value it?

I ask this question particularly because I sometimes get the idea that, in dealing with peak and peak vs. career, there's a lot of "eyeball measurement" going on. Maybe that's necessary, but I'd find it useful, as I think about how to measure and to value peak, to see what others really do. I suspect that there's a lot of variety.

Since I've raised the question, I'll state what I do. I started off thinking about top three seasons and best five consecutive seasons and WS/162 games, as per the BJ NHBA, but I'm now feeling like these ways of selecting information are too arbitrary, so I'm calculating what I think of as "total peak performance." I've calculated an average win share level for position players under the conditions that apply for a particular group of seasons, and subtract that total from a player's WS each year that the player was above average to find the total number of WS a player was above average in a career.

I then use differences in "total peak" to give me a suggestion about when to shift players out of a basic rank order established by career win shares. If the difference between career peaks for any two players is about the same as the difference between career win shares and the player with the lower career total has the higher peak, I consider whether to switch the players in the rankings.

Question for you all: I've been trying to come up added information when trying to place Sam Thompson in the voting ranks. Courtesy of mlb.com, here's some material about the Philadelphia home park starting in 1887.

"After holding their first spring training at Recreation Park, the Phillies opened their maiden season there, losing to the Providence Grays, 4-3, on May 1, 1883. The Phillies remained at Recreation Park for four seasons, but in need of a park to accommodate larger crowds, finally moved out after the 1887 season. The Phillies moved into a brand new stadium at Broad Street and Lehigh Avenue. Officially called National League Park, it was informally called Huntingdon Street Grounds and Philadelphia Base Ball Park, it was built at a cost of $101,000. The original park seated 12,500.

Soon after its construction, National League Park was being hailed as the finest and most modern stadium in the nation. Using brick instead of wood for the outside part of the structure, it included two 75-foot high turrets at either end of the 5,000-seat pavilion behind home plate and a 165-foot high turret at the main entrance at 15th and Huntingdon Streets. Sheds for 55 horse-drawn carriages were located under the grandstands.

The park also had some highly unusual features, foremost of which was the high outfield wall that extended from right to center field. Just 272 feet down the right field line through most of the years (originally, it was 300 feet), the wall was notorious for turning pop flies into home runs and screaming line drives into singles. National League Park opened on April 30, 1887 with the Phillies defeating the New York Giants, 19-10. In 1894, a major fire destroyed much of the park, and it was rebuilt using mostly steel, a radical new technique in stadium construction. Capacity was increased to 18,800, and clubhouses were installed in center field. The Phillies clubhouse contained a swimming pool. "

The new park eventually became the Baker Bowl and we know the rest. So my question is Sam Thompson's accomplishments due to him being a left handed hitter in a park whose park factor is not extreme but looks to greatly benefit a LH pull hitter? Is he a product of his ballpark (as well as a good hitters enviroment in general in the 1890s)?

Regarding peak. I measure peak as best 5 years(not necessarily consecutive) and best 8 years(also NNC). I determine who the bes players are by this measure and I compare that to my list of best players career value only. I then make a hybrid list that attempts to value both equally. I admit that 5 and 8 years are an arbitrary number, but that is what works for me. I don't use consecutive years because I think that discriminates against players who don't follow a typical career progression for one reason or another. Right now, I am using W1 with an adjustment to 162 games and an AA discount as well. I eventually plan on using both WS and W3 once season lengths and leagues stabilize into there more modern form- ie when I start to trust that WS is accurately evaluating pitching and defense. I think that I will feel comfortable with this once we get to players who played their entire career after the Turn of the Century.

The new park eventually became the Baker Bowl and we know the rest. So my question is Sam Thompson's accomplishments due to him being a left handed hitter in a park whose park factor is not extreme but looks to greatly benefit a LH pull hitter? Is he a product of his ballpark (as well as a good hitters enviroment in general in the 1890s)?

His homerun home/away breakdown is 84/42. He definitely received an advantage, but not of Gavvy Cravath or Ed Williamson proportions.

"The new park eventually became the Baker Bowl and we know the rest. So my question is Sam Thompson's accomplishments due to him being a left handed hitter in a park whose park factor is not extreme but looks to greatly benefit a LH pull hitter? Is he a product of his ballpark (as well as a good hitters enviroment in general in the 1890s)? "

Al:

The ballpark probably turned some of Thompson's pop flies into home runs -- although he only hit 4, 7, and 9 HR in 90, 91 and 92, so it didn't happen too often in those years.

On the other hand, it is just as likely that it turned a fair number of hits that would otherwise be doubles and triples into singles if he was banging them off the wall. I'm not sure the net effect would be tremendous in Thompson's case.

Re Sam Thompson and being a left-handed hitter in Baker Bowl: Billy Hamilton, Jack Clements, and Elmer Flick may have some relevance to that discussion, as well as Chuck Klein from a later era.

Maybe it's the obvious point, but the effects of the Green Monster in Fenway should provide some examples of the effect of the Baker Bowl's wall, though at 272' down the line it was obviously more extreme.

Billy Hamilton -- a left-handed Wade Boggs with speed?

Wade has a good point. Hitting to the opposite field also gets its benefit when lazy fly balls bounce off the wall for doubles.

though at 272' down the line it was obviously more extreme.

It's at least 20 feet closer. :-)

I think I'm more in Joe's camp in terms of what to do with Thompson. We're comparing player vs. player and environment has to be considered, albeit to a small degree, when doing so.

The short porch in Philadelphia might explain some of Thompson's fielding numbers as well. Not sure how to get my arms around that issue yet.

One thing that troubles me about discounting Thompson's production because he got the most out of his home ballpark -- isn't that similar to (but not quite the same as) discounting Barnes for making use of the fair foul rule. To a large extent both players made the best of what they were presented with and the results were actual runs/wins/pennants or whatever.

Having finally elected Old Hoss, I'm busy re-jigging my ballot for next week. 1870s players in 4 of the top 5 positions, which reflects the fact that we've elected everybody really good except Galvin from the 1880s. However, my no. 6 is contrary to conventional wisdom, and I thought it worth running up the flagpole prior to registering it definitvely:

6. Mickey Welch ? 307-210 comes to impress me more and more (yes, I know it was mostly with the strong Giants.)1885 looks like a pretty good peak too; 44-11 with a 1.67 ERA is pretty impressive, compared for example to Clarkson?s 49-19 at 2.73 in 1889 or his other peak of 53-16 with a higher ERA of 1.85 in Welch?s peak year of 1885. Clarkson gets ERA+ of 165 for 1885, compared to Welch?s 161, the difference presumably due to ?park effects? between Chicago and NY. Well, I don?t think we have any idea what the ?park effects? should be for the 1885 National League. If it?s calculated by looking at runs for/against in each park during the season, then there is room for huge random fluctuations as you?re talking sample sizes of 60 games or so. Welch not as good as Clarkson, but not that far off.

" One thing that troubles me about discounting Thompson's production because he got the most out of his home ballpark -- isn't that similar to (but not quite the same as) discounting Barnes for making use of the fair foul rule. To a large extent both players made the best of what they were presented with and the results were actual runs/wins/pennants or whatever."

This is my position also. I'm 100% in favor of discounting or marking up a player's raw stats based on their run scoring environment.

However, NO player played in a neutral, vacuum environment. Every player's stats are impacted by their home park and their league environment (Barnes, etc.). Unless you have sufficient data to "adjust" all players to a neutral environment, I don't think it's correct to cherry pick certain players to adjust up or down in this manner.

"Brian, I see your point, but Barnes took advantage of something that was available to every player in the league (a rule). Thompson (if the hypothesis presented earlier is correct) took advantage of something that was available to LHB in his park. That separates the two for me."

I think this is not the best way to go. For an example, if Joe Dimaggio either didn't or couldn't adjust to the deep LF in Yankee Stadium while batting, then he was costing his team REAL outs, regardless of whether he might have performed better in a park that was more favorable to RH batters.

And if Jim Rice was somehow better able to take advantage of the Green Monster than other players, then he was providing his team with REAL HR's, REAL WINS, etc. that should not be "adjusted out" of his performance.

My take on Thompson's defense, WS vs WARP, as that they're probably both right, given their point of view. Win Shares treats all Outfielders alike; Thompson apparently didn't make as many plays as the average OF'er, therefore he's not so hot (B-R.com backs this up; his Range Factor is .23 plays per game less than the average OF'er). WARP considers LF, CR, and RF each as a separate position (and attempts to estimate breakdowns when the guy plays multiple OF positions); Thompson apparently made more plays (particularly assists) than the typical RF'er, therefore he's above average.

Like with catching (Ewing vs Bennett), to me fielding WS appears to be finely tuned to the modern game and has trouble dealing with stats that are out-of-whack for a modern fielder. This doesn't mean that WARP-1 is necessarily better, but when they give significantly different answers, it starts one thinking.

Outfield play, now and then:

Fielding stats for modern outfielders per game:

jimd -- this is fabulous! Thanks!

No analysis to contribute yet, but on King Kelly I can report that James in NHBA states that Kelly's assist rates are almost twice that of any other outfielder. James doesn't know quite what to make of it -- he suggests that Kelly may have been listed in RF but played as a fifth infielder much of the time. Perhaps Richardson was playing similarly, in imitation?

Could be. There are 10 player-seasons between 1878 and 1884 where an OF played at least 30G and posted an assist rate of at least .5 per game. Kelly has 3 of them (1878, 80, 82), Hugh Nicol (82, 84); Richardson 81, Cassidy 78, Ward 83, Holbert 78, and Shaffer 79. Shaffer has the highest assist rate (.694 per game); he's also an RF (I checked him, not the others). Almost all the PO rates are below 1.5 however; Richardson and Ward are big exceptions though, over 2.0. James is right; these stats are odd. OTOH, I wonder if Richardson '81 is BP's analog to Total Baseball's Lajoie?

Since I did a work-up of Cal McVey's estimated WS 1871-75, converted from WARP (posted on the McVey-Start thread), I thought I would do the same for Ezra Sutton. If you want to see the full methodology of the conversion, it's in the McVey post.

I conclude that, if Sutton's NA years are given weight, he's clearly the best position player on the ballot. I also have some educated guesses about the causes of his curious mid-career trough.

Ezra Sutton Study

157 raw win shares (273 adjusted) 76-88 117 batting, 41 fielding

289 BRAR 76-88

replacement level = 3.75 BWS / 162 games

269 FRAR -- > 26.9 FWAR FWS/FWAR = 1.52

Year -- BWAR/FWAR --> BWS / FWS = total --> adj. 162 games (from x games in season)

Once more for Al Spalding...

1) Retired at age 26. He had the opportunity to make more money outside of baseball and he took it (speculation: his arm was done and he knew it, hence the move to 1B, and he then realized he was just another above-average hitter if he wasn't pitching). He was one of the most famous ballplayers at that time, if not the most famous, and he cashed in on that fame with a sporting goods business that made uniforms and baseballs (its still in business in 2003, making basketballs and other stuff) and landed the NL contract for baseballs, and with a publishing business that did the official NL publications as well as popular guides to baseball. He also served as Chicago's club secretary (sort of like GM but nowhere near that busy) and bought the team when League president and Chicago owner William Hulbert died in 1882.

2) Teenage phenom. Bob Feller struck out 15 in his major league debut at the age of 17. These things happen.

3) DIPS. First cut; the circumstantial evidence. If pitchers had no impact at all, then they'd have somebody like Dan Brouthers out there tossing meatballs so they could get his bat in the lineup; give these guys some credit for knowing their game well enough to know where they can grab an edge. Harry Wright thought that pitching mattered enough that he had to have Spalding, and paid him accordingly, and Harry proved that he knew what he was doing (built teams that won 8 championships in 10 years 1869-78) far more than most of us have (:-). Spalding's hitting and the fielding aren't enough alone to warrant that kind of attention and salary unless he's also having a perceptable impact as a pitcher (and Harry noticed this before Spalding joined Harry's championship team). As far as doing an actual DIPS study during this period, the rules are constantly changing so you'd have to normalize the rates each year, the sample sizes are about 6-10 pitchers per year (really only one pitcher per team), the park factors are unreliable due to that visiting-team-supplies-the-nonstandard-ball rule; IMO it'd be very difficult to do anything credible (but I'm not a statistician, so what do I know).

4) Own team. How big a factor is this? In an 8 team league, 7/8ths of the season is directly comparable when comparing players on two different teams. Any differences arise from their head-to-head games. Nobody seems to be bothered by this factor for Caruthers or Clarkson or Nichols, other pitchers who spent practically their entire careers on pennant-winners/contenders, or for the hitters in similar positions who never faced their own championship pitching staffs (e.g. Gore, Wright, Barnes).

5) Pitcher Fielding. Far better minds than mine (Bill James, Davenport) have punted on the question of the fielding value of a pitcher. All I can say is that the total package would seem to indicate that he was a great athlete; his fielding stats at pitcher are arguably better than Ward's. Spalding made 25% more plays with 29% less errors compared to the average pitcher fielding rates of his league; Ward made 20% more plays with 31% less errors in his. It's a huge leap to say that Spalding also could have been a SS, but BP rates him better than league average in a 13 game sample at 2B, and gives him a "Gold Glove" at 1B in 1877.

Other FOAS chime in if I missed anything.

Since I did a work-up of Cal McVey's estimated WS 1871-75, converted from WARP (posted on the McVey-Start thread), I thought I would do the same for Ezra Sutton. If you want to see the full methodology of the conversion, it's in the McVey post.

I conclude that, if Sutton's NA years are given weight, he's clearly the best position player on the ballot. I also have some educated guesses about the causes of his curious mid-career trough.

Ezra Sutton Study

157 raw win shares (273 adjusted) 76-88 117 batting, 41 fielding

289 BRAR 76-88

replacement level = 3.75 BWS / 162 games

269 FRAR -- > 26.9 FWAR FWS/FWAR = 1.52

Year -- BWAR/FWAR --> BWS / FWS = total --> adj. 162 games (from x games in season)

re: Thompson debate

Baker Bowl Estimated Impacts vs. LH an RH batters, 1921-1927

LEFT HANDED BATTERS RIGHT HANDED BATTERS

Thompson only played around 17% of his career in the Baker Bowl.

He played around 25% of his career at Recreation Park in Detroit, dimensions unknown.

He played around 58% of his career in the Philadelphia Baseball Grounds. It had a 25 ft high fence, was 500 ft in LF, 500 ft in CF, and 310 ft in RF. That may explain a lot about his fielding stats, but the shorter RF distance wouldn't NECESSARILY be an advantage to LH hitters in that era as the longer distances in LF and CF would probably result in more singles, doubles, triples and inside the park HR's for RH batters.

jimd wrote:

It looks likely to me that Sutton suffered some sort of an arm injury in 76 , necessitating his shift to second and first base, and that the injury lingered for several years, causing him to play more SS and less third base.

He definitely had an arm injury in '76 and was moved to first because of it (according to Nineteenth Century Stars). Whether or not his arm problems carried over to those other seasons is a mystery.

Great work on Sutton, Chris! The case for Ezra builds and builds!

My concern starts with the fact that he retired at age 26. How many guys would have been legit HoMers if they had retired at age 26?

He blew out his arm. That's why he was being used as a first baseman in '77.

2) To the extent that people point to Spalding's performance as a teenager, that only seems to beg the question about the true quality of the competition he faced.

Actually, it begs the question how young Ted Williams, Roger Clemens and Honus Wagner would have been playing with the best competition of that time? Most people didn't have the time or monetary resources to keep playing in their twenties. No wonder there were more teenagers during that era.

New Prelim Ballot

Well, I said Sutton was moving up, and I meant it. I have been spending time on the McVey thread and he gets a lofty spot on my ballot for the first time. Here goes:

1. Galvin (2)

Is the '314' for McVey with or without adjusting the fielding WS upward? If the score is 415 to 314 in favor of Ezra, it may come down to whether I credit McVey for 100+ WS in his Western and pre-NA careers.

Here's the info filled into Tom's chart:

player .....years.. WScalc WS(w/ def +30%)

'And if Jim Rice was somehow better able to take advantage of the Green Monster than other players, then he was providing his team with REAL HR's, REAL WINS, etc. that should not be "adjusted out" of his performance. '

Could a Friend of Dickey Pearce(probably John Murphy) run down his accomplishments prior to the NA? I am trying to give him some more consideration, but don't feel I currently have enough information.

tentative:

1. Joe Start

Here is my preliminary ballot. There are quite a few changes from my previous ballots. I recently finished reading up on Win Shares and just got seasonal Win Shares data recently. (Better late than never). I also took a closer look at the data. (Both WS and WARP). Also, this doesn't really affect this election much, but I will be taking a closer look at a players contemporaries, both on the ballot and those not eligible yet. I neglected to do that when the contemporaries were not on the ballot, and I believe that was a mistake.

1906 List

1. Al Spalding (1) - Best 1870?s pitcher.

Spalding, WARP-1, WARP-2, and WARP-3.

Spalding added extra value to his position as a hitter. But his predominant value is still from being a pitcher, at least according to BP and WARP-1. Be patient with me because this will be a bit technical. Here's his player card if you want to try to follow my confusing explanation.

If you look at the "Advanced Pitching Statistics" section, there are a number of columns. The ones involving runs are:

Here are adj. WS for Lip Pike's NA seasons:

bWS / fWS = total (games) --> adj. WS (162 games)

More on Ezra Sutton's NA WS: when I was doing Lip Pike's translations, I realized that I had used the Boston team's season lengths for Sutton instead of the Cleveland and Philadelphia season lengths, which were shorter. So Sutton earned his WS in fewer games, leading to higher adj. WS. Here are his totals adjusted based the actual season length:

71 5.2 / 3.2 = 8.4 (30 games) --> 45

Could a Friend of Dickey Pearce(probably John Murphy) run down his accomplishments prior to the NA?

I'll be glad to make the attempt.

Pearce created the position of shortstop as we know it today in the mid-1850s. Before he came along, the position had the least defensive importance. Pearce was the first to roam into the outfield to find fly balls, mastered defensive positioning, and improved upon the art of making the throw to first.

For about ten years (1856-1866), he was the consensus for best shortstop in the game. When Wright came, that was the end of his dominance (though his fielding was still comparable). He was picked for two All Star games: one as a shortstop in 1858 and the other as catcher (then more important than even pitching) in 1861. He was primarily a catcher for a couple more seasons and was regarded as terrific there, too.

As for his hitting, he wasn't Wright, but he appears to have been the best at the height of his dominance. He was a contact hitter and was instrumental in the development of the bunt and fair/foul hit (he was the best at the latter until Ross Barnes).

He played 20 years at the best level of competition. His stats don't look impressive to the eye, but you have to remember that he was 35 in 1871. No shortstop of the nineteenth century had the value he had after that age. It's a testament to his durability in a time where your hands would be broken many times over from the lack of playing glove. Not until George Davis at the turn-of-the-last-century did shortstops have productive seasons after the age of 35.

Here's a summary of the WS translated from WARP for McVey, Sutton, Start, and Pike. All figures are seaonally adjusted WS.

Cal McVey 48, 24, 27, 38, 51 --188 total

Ezra Sutton 45, 18, 35, 30, 37 -- 165 total

Joe Start 21, 9, 11, 27, 25l -- 93 total

Lip Pike 43, 33, 27, 44, 49 -- 196 total

I'll have Dickey Pearce and some other folks to add before the weekend is out.

Ack! typos! Here the info with Start's 1875 WS correctly indicated:

Here's a summary of the WS translated from WARP for McVey, Sutton, Start, and Pike. All figures are seasonally adjusted WS.

Cal McVey 48, 24, 27, 38, 51 --188 total

Ezra Sutton 45, 18, 35, 30, 37 -- 165 total

Joe Start 21, 9, 11, 27, 25 -- 93 total

Lip Pike 43, 33, 27, 44, 49 -- 196 total

I'll have Dickey Pearce and some other folks to add before the weekend is out.

Chris:

Do you have time for Levi Meyerle and Dave Eggler? The latter was a terrific player in his own right but had a very short career.

Thanks!

Sometime Sunday I'll be posting NA WS for the following players:

Dickey Pearce

Chris Cobb, you're a sabermetric God! :-)

BTW, here's something for everyone here to beat me up on. I think Sutton was better than Frank Baker in total value at third. Baker trumps Sutton's peak, but Ezra has a healthy lead in durability (though part of Baker's problem is more due to missing two seasons than injuries). The Home Run guy is close, but not close enough.

I haven't looked at Traynor, Groh, Hack, Dandridge, Elliot or Clift yet to see how much farther into the last century Sutton was the preeminent third baseman.

John:

Won't beat you up, because there is by no means a "right" answer, and both should be HOMers, but I have Baker ahead of Sutton.

No question that there is lots of value in the length of Sutton's career, and if you take into account the greater defensive responsibility of a 3b during Sutton's era, the defense comes out about the same, IMO.

So I'm looking at a few other intangibles. I think Ezra was the best player at his position during his era, but I think Baker was the best player at his position AND at times, arguably the best player in baseball during his era. And while I realize that winning teams are rarely based on one player, Baker was at times the best player on one of the top 15 or so dynasties in baseball history.

Hopefully Sutton will be elected before Baker becomes eligible in 1927. He's number one on my ballot.

Jeff:

You're right that there are different ways to compare Sutton and Baker. You can make an excellent case for the latter over the former. What's amazing to me is that there is any comparison at all! Who heard of Sutton a year ago?

How does the voting work in a one-electee year like 1906?

A heads-up for those who worry that the pool is soon going to get too shallow: We've already elected 18 of the 33 players we'll elect "through the 1916 ballot," which in real time will take us to the end of 2003.

Incidentally, the "Dec. 22-Jan. 5" discussion/voting period shouldn't take too much time away from anyone's holiday plans: 1917 is a one-electee year, and Denton True Young will be joining the ballot.

I can't see how he'll miss being number one on every voter's ballot.

I don't know, John. Denton True only led his league in ERA and ERA+ twice, and he spend half of his career in a much weaker American League -- not the best competition of his time. And worst of all, his second Most Similar pitcher in only Pud Galvin!

Plus, I checked the record books thoroughly. He did not win a single Cy Young award! What kind of a pitching great has that kind of record?

Here are NA WS for other players whom you might be interested in. Since many of these players have been off our radar, I've included adj. WS for their post-NA careers and career WS totals in addition to their season-by-season adj. WS for the NA.

Please remember that these are estimates of what WS values would look like. For players with significant NL time, range of error is about 5%. For players with little or no NL time, range of error goes up to 15% or so, since I have to make an educated guess about the fielding translation by looking at how WS treats similar fielders.

If you want to see the fuller write-up on how the translation worked for any of these guys, let me know.

NA WS Summary

Dave Eggler, 1871-75

Plus, I checked the record books thoroughly. He did not win a single Cy Young award! What kind of a pitching great has that kind of record?

Okay, you convinced me. I won't place him on my ballot. :-)