Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Monday, June 09, 2003

1903 Ballot Discussion

I’ve sent an email to the list explaining why I’m late, if someone can post the newly eligible list in the discussion here, that’d be great too. I figure there’s no point waiting to get this going though . . .

I should have the 1902 final result up in the next few hours.

Joey Numbaz (Scruff) Posted: June 09, 2003 at 09:18 PM | 155 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 1 of 2 pages  1 2 > 
   1. RobC Posted: June 09, 2003 at 09:30 PM (#513944)
Here is the list from the 1902 discussion, with Stivetts left off, since he isnt eligible yet.

1903:
   2. jimd Posted: June 09, 2003 at 09:54 PM (#513945)
This seems like a no-brainer election. I know that was said last time, though not by me, but this time it may even be true.

I have Connor #1 and Anson #2. I think Connor has a much better peak, outweighing Anson's never-ending career. That may be the big debate this week; who's #1?

I don't think anybody else even makes my waiting list. If I had no AA discount, maybe Lyons could replace Williamson. Pfeffer over Dunlap? I don't think so. Silver King is Caruthers lite without the bat.
   3. jimd Posted: June 09, 2003 at 10:03 PM (#513946)
I suppose another possibility is that the Anson first ballot boycott may provide an opportunity for the Ewing/Glasscock loser to sneak in early.
   4. Marc Posted: June 10, 2003 at 12:41 AM (#513949)
This is undoubtedly my first, last and only chance to post the first prelim (unless somebody is sneaking one in as I speak).

1. Spalding (up from 2)
   5. MattB Posted: June 10, 2003 at 01:02 AM (#513950)
Provisional ballot. Connor's the only new addition. Others bounce around a little.

1. Roger Connor -- #1 on my ballot with a 1906 start date, so obviously #1 in any individual year before that.

2. Joe Start (2) ? Holding strong at #2.

3. Pud Galvin (4) ? the more I think about Galvin, the more I think he's the best pitcher on the board. Better than Spalding, better than Radbourn, and probably better than the rest of the hitters.

4. Jack Glasscock (6) -- Head to head against Sutton, it looks like Glassock wins, so I'm giving him a bump.

5. Ezra Sutton (3)

6. Al Spalding (7) ? best early pitcher

7. Charlie Bennett (11) ? No doubt now. He's the best catcher on the board. Solid comparison to a HoMer (Ewing) helps his standing.

8. Bob Caruthers (5)
   6. DanG Posted: June 10, 2003 at 02:56 AM (#513952)
The SABR Bio Project has just published a good bio of Jack Glasscock at the page above.
   7. Rusty Priske Posted: June 10, 2003 at 03:17 AM (#513953)
I will post my pre-vote tomorrow but right now I see Anson as #1 and Connor as #2. I also think Silver King will make my ballot, but obviously lower.

It pains me to take Old Hoss out of my top2, however, as I think he should have made it in already.
   8. DanG Posted: June 10, 2003 at 04:16 AM (#513954)
Here's the Bioproject entry for Lip Pike.
   9. DanG Posted: June 10, 2003 at 04:26 AM (#513955)
Here's the Bioproject entry for Pete Browning.
   10. Rusty Priske Posted: June 10, 2003 at 12:19 PM (#513956)
Doesn't "boycotting" Anson for his racist attitudes (I assume) go directly against the purpose of the Hall of Merit and it's Constitution?
   11. Rusty Priske Posted: June 10, 2003 at 12:23 PM (#513957)
Here is the Top 4 I would like to put in this year:

1. Old Hoss Radbourne
   12. MattB Posted: June 10, 2003 at 12:54 PM (#513959)
Rusty,

The boycott section was specifically put into the Constitution (read it in the Constitution thread to the right) to strike a balance between those who wanted to consider personality as part of the mix and those who felt personality had no place in the discussion at all.

The relevant section of the Constitution reads:

"A player?s ?personality? is to be considered only to the extent that it affected the outcomes of the player?s games (e.g., via his positive or negative effect on his teammates). In rare and extreme cases, a voter may opt to exclude a player on ?personality? grounds on the first ballot on which the player appears. If that player does not get elected on his first ballot, the voter shall give the player full consideration in all subsequent ballots, regardless of the ?personality? factors."

Those who choose to boycott Anson may do so for only one year. Then, he must be placed on the ballot if not elected by the others (unless, that is, you somehow don't consider him in the Top 15, irrespective of personality issues, which doesn't seem likely).
   13. Rusty Priske Posted: June 10, 2003 at 12:59 PM (#513960)
Okay, I didn't catch that. A token head shake, basically.
   14. RobC Posted: June 10, 2003 at 02:12 PM (#513962)
My prelim ballot:

1. Cap Anson
   15. Marc Posted: June 10, 2003 at 02:50 PM (#513964)
I can't believe that an OF who hits like Thompson could possibly jump from 5th to 14th or 14th to 5th based on defense. 5th to 6th maybe. WS has an OFers' value pegged usually around 15-20% defense. Is this appropriate? Does anybody have a method of calibrating that?
   16. Chris Cobb Posted: June 10, 2003 at 03:37 PM (#513965)
Since Thompson is under discussion, this seems a good time to post this longer comparison of Thompson and Harry Stovey. It doesn't try to answer questions about defense, but it may be useful.

Taking on the mantle of FOHS, now the most underrated player on the ballot, I'm going to compare Stovey to Sam Thompson. I think it's readily demonstrable that Stovey was significantly better, both at peak and over the course of his career. Here's how their careers line up, as figured in adjusted Win Shares:

Age St Th
   17. dan b Posted: June 10, 2003 at 03:50 PM (#513967)
Marc - 5 prelim ballots up and 3 no-votes for Anson. Have you started to think about where you will place him on your 1904 ballot? :-)
   18. MattB Posted: June 10, 2003 at 03:55 PM (#513968)
Chris,

I agree with you on Stovey v. Thompson. I just don't have either of them ranked very high.

In terms of WARP numbers, Thompson does have an advantage.

Stovey has a 108.2 to 102.4 advantage is WARP-1 (unadjusted), but Thompson leads in WARP-3 (adjusted for season length and level of difficulty) 86.3 to 76.2. The problem seems to be that Stovey's best years were '83, '84, '88, and '89, with weaker performances in the relatively stronger '85-'87 AA.

On the other hand, a lot of Thompson's value is tied up in his defense in WARP (562 batting runs above replacement and 334 fielding runs above replacement). This strikes me as odd, as Tom H describes above. If the defensive numbers are put aside, or looked at from a Wins Shares perspective, Stovey is the clearly superior player.
   19. Marc Posted: June 10, 2003 at 04:20 PM (#513969)
danb, small sample ;-) I still assume Cappy will get elected this year. But if not, forgetting for the moment the specific new eligibles in '04 (whom I have indeed forgotten), but just placing him among the '03 cohort, I do have him above Connor so the only question in my mind is where he goes relative to Spalding.

BTW, re Thompson Matt inadvertently answered my question. Does it seem odd to anybody else that a guy whose whole career says "slugger" would have 500+ BR and 300+ FR? IOW 3/8 of his value is defense? I'm not as familiar with WARP as a lot of you, but this seems totally inappropriate. What gives?

I also think the AA discount is not the only issue, BTW, but the fact that, Thompson played in a 12 team environment for part of his peak and Stovey peaked in 16 and (even) 25 team environments. I don't timeline, per se, and this isn't the same as the AA discount (vs. a contemporary league) but this is a different consideration and I think a significant one.
   20. Chris Cobb Posted: June 10, 2003 at 04:48 PM (#513970)
Re: number of teams:

Doesn't analysis of later expansions suggest that there is less competitive balance in an expansion year, but that after that differences disappear?

Thompson played for his first seven seasons, ages 25-32, under 16-25 team conditions. If competition was easier then, wouldn't Thompson's earlier years, including his single best season (the peculiar year 1887), need to be re-assessed downwards also?
   21. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 05:07 PM (#513972)
I have Thompson slightly ahead of Stovey, but it's extremely close.

I can't see either of them in the top ten of a ballot, IMO.
   22. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 06:07 PM (#513975)
John C:

If we compare Start and Stovey after the age of 28 (the age Start played his first game in the NA), Old Reliable's career is almost twice the length of Harry. I think it's reasonable to conclude that if he had those missing pre-NA years attached to his record, Start would flat out destroy Stovey in value as a player.
   23. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 06:19 PM (#513976)
<i>AdjWS (not including NA for Start)
   24. MattB Posted: June 10, 2003 at 06:23 PM (#513977)
Exactly, John.

Start's numbers are from age 33 on. If Start had a better career from age 33 on than Stovey had from age 28 on, it's pretty clear to me that the two are not even in the same ballpark, value-wise.
   25. jimd Posted: June 10, 2003 at 06:44 PM (#513979)
Where has the idea come from that the AA was competitive with the NL in 1883 and 1884?

According to BP (Davenport) the AA of 1882 was not much better than the UA in 1884. The AA of 1883-84 is somewhat improved but still around AAA quality (20% discount from NL).

If you look closely at the AA of 1882, it's low quality rating makes sense. Only 14 of 54 regulars were ever regulars in the NL/NA at any previous time, and only 3 had been regulars in the NL of 1881. The AA never attempted raids on the NL like the AL is "currently" doing. It didn't have to because it occupied larger markets than the NL did; dominance was inevitable in the longterm if it used its market power to compete for young talent. I think that the NL realized that and made sure that it didn't happen by aggressively persuading AA teams to switch leagues.

It's hard to see how the AA could improve much during 1883-84 without player raids on the NL because the AA was expanding during this period. The AA went from 6 teams in 1882 to 8 in 1883 to 12 in 1884. Only when it contracted back to 8 teams in 1885 did it start to approach the NL in quality. Davenport's numbers contend that it never reached parity; the gap was somewhat worse than the NL/AL gap of the early 60's, so OTOH there is also no question that the AA was a major league from 1885-89.

After that, it's all downhill. The AA of 1890 was hard hit by 1) loss of 3 of its top 5 teams, Brooklyn and Cincinnati to the NL and Baltimore just quit, 2) player defections of its top stars to the PL, 3) player raids by the NL; the AA was forced to add minor league teams and was by far the weakest of the three leagues. 1891 brought a desperation war with the NL, and merger on NL terms afterwards. The four AA clubs that survived finished 9-12 in the NL of 1892. The NL of 1892 consisted of four old-line NL teams (Bos, Chi, NY, Phi), four AA "turncoats" (Pit, Bro, Cin, Cle), and four AA survivors (StL, Bal, Lou, Was), so in a sense, it was the NL in name only.
   26. MattB Posted: June 10, 2003 at 06:44 PM (#513980)
"How much value do we assign players for their career before 1871, since there isn't the gold standard of "pennant"?"

Now, wait a minute.

The National Association existed before 1871, but it was an amateur association rather than a professional one.

Joe Start played for the Brooklyn Atlantics, which won the amateur national championship (the highest level of play available at the time) in 1861, 1864, and 1865. In the latter two years, Start was the team's leader.

Extra credit if you can find Joe Start in this picture of the Pennant-Winning 1865 Brooklyn Atlantics. Try telling these guys they didn't just win their second consecutive pennant.

http://rs6.loc.gov/pnp/cph/3b40000/3b40000/3b40700/3b40721r.jpg
   27. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 06:45 PM (#513981)
<i>This "if" is the crux of the issue. That's a really big "if" to me, because I don't know why I should consider Start's pre-1871 value.
   28. jimd Posted: June 10, 2003 at 06:50 PM (#513983)
the amateur national championship

This is an honest question, as I've never noticed any discussion of this.

How was this championship determined? Was it by best record? A playoff system of some kind? A newspaper poll (like college football today)?
   29. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 06:54 PM (#513984)
Obviously, Thomas and Palmeiro are exceptions, and Thomas could decide to play for 10 more years and obliterate Palmeiro in value. However, the example still points out that not all career paths are the same, and post age 33 value is not always fully indicative of pre age 33 value.

Agreed. That's why I had him at #9 on my ballot. I think we need to be conservative with our analysis of him.

However, we do know what he did after 1870. We also know he was considered the best first baseman of the '60s. He played 27 years. That has to have more value than the 14 year career of Stovey.
   30. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 07:02 PM (#513985)
According to BP (Davenport) the AA of 1882 was not much better than the UA in 1884. The AA of 1883-84 is somewhat improved but still around AAA quality (20% discount from NL).

jimd:
   31. Chris Cobb Posted: June 10, 2003 at 07:02 PM (#513986)
jimd -- thank you for more specific info on Davenport's evaluation of AA vs. NL. Would it be too much trouble for you (or someone else) to list the year-by-year differences, as he assesses them? (These may have been posted before, but I haven't been able to find them among the thousands of messages lying around . . . )
   32. MattB Posted: June 10, 2003 at 07:09 PM (#513988)
"How was this championship determined? Was it by best record? A playoff system of some kind? A newspaper poll (like college football today)?"

The National Championship was generally the team with the best record among those teams that had joined the National Association. In practice, it was a mess, since, for example, in 1867 there were over 300, so everyone did not play everyone else, of course. The professional NA was set up in 1871 specifically to deal with the problems of deciding who was "National Champion" that had occurred in the 1860s.

By general acclaimation, there was a known pecking order. Teams that thought they were worthy travelled to Brooklyn to play the Atlantics, the Washington Nationals, and The Philadelphia Athletics (the stars came to the big cities to play).

In 1864 and 1865, the Atlantics beat all comers. They were undefeated, and generally crushed the competition. There was no question.

In 1866, the Philadelphia Athletics were a real challenge to the top spot. A Championship Game was played between the two teams and Philadelphia won.

The coming of the professional Cincinnati team in 1869 put them on top, but their winning streak was ended in 1870 by Joe Start and the Atlantics (Start had the game winning 11th innning triple, over right fielder Cal McVey's head).
   33. Chris Cobb Posted: June 10, 2003 at 07:12 PM (#513989)
jimd -- thank you for more specific info on Davenport's evaluation of AA vs. NL. Would it be too much trouble for you (or someone else) to list the year-by-year differences, as he assesses them? (These may have been posted before, but I haven't been able to find them among the thousands of messages lying around . . . )
   34. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 07:16 PM (#513990)
I was going over Al Spalding's BP card and noticed the huge discrepancy between the Boston defense of 1871 and 1872. Except for Andy Leonard and Fraley Rogers, the starting lineup of '72 is the same as '71. Does this make sense to anyone?
   35. jimd Posted: June 10, 2003 at 07:41 PM (#513991)
George Wright missed half the 1871 season (a knee injury, I think). Ross Barnes played out-of-position at SS to replace him, and Sam Jackson replaced Barnes at 2B.
   36. John (You Can Call Me Grandma) Murphy Posted: June 10, 2003 at 07:46 PM (#513992)
George Wright missed half the 1871 season (a knee injury, I think). Ross Barnes played out-of-position at SS to replace him, and Sam Jackson replaced Barnes at 2B.

Barnes was originally a shortstop before 1871, so he knew the position (and appears to have done a great job that year).

Jackson wasn't Barnes at second, but looks like he didn't hurt them there.
   37. Carl Goetz Posted: June 10, 2003 at 07:57 PM (#513993)
Warp-3 analysis Age 28 and after:
   38. Marc Posted: June 10, 2003 at 07:57 PM (#513994)
1800s lg BWA
   39. Marc Posted: June 10, 2003 at 08:03 PM (#513996)
PS. In the Stovey-Thompson debate, note that we are not talking about an expansion but a contraction. I don't know of any study of the impact of contraction but I don't see how expansion models contraction. With an expansion, you assume that a growing population of baseball playing boys and men gradually fills the expanded ML with players of roughly equal skill and ability. With contraction both trends (more men, fewer teams) would both push in the same direction--i.e. better play, tougher competition.

Also, perhaps Thompson's late '80s seasons should in some sense be "discounted," but then not as much as Stovey's (league difference). But more to the point, how you treat his '80s seasons have nothing to do with how you treat his great years in the '90s.
   40. jimd Posted: June 10, 2003 at 08:04 PM (#513997)
Using BP fielding ratings: in 1871 Wright was 123 at SS, Barnes 122, Barnes filled Wright's shoes; at 2B Barnes was 173, Jackson was 100 or league average, Jackson didn't fill Barnes'. In 1872, both HOMers were phenomenal; Wright was 176 at SS and Barnes was 220 at 2B.
   41. jimd Posted: June 10, 2003 at 08:18 PM (#513998)
Here are my two previous posts giving data on this subject:

"19th-Century DERA offset"

"AA/AL vs NL as percent"

The last sentence should read "between NL and the AA at its best." (? Where did that correction go; I remember posting something to that effect earlier. ?)
   42. Devin has a deep burning passion for fuzzy socks Posted: June 10, 2003 at 08:58 PM (#514000)
Hey, don't the Atlantics' 1864 & 1865 championships have to be discounted due to the wartime level of competition?

(Ducks to avoid thrown objects)

On a more serious note, FOHS have referenced a SABR group determining him to be the 19th-Century player most deserving of Hall of Fame induction. Does anyone have any more information on this (who the runners up were, how they made the decision, etc.)?
   43. Howie Menckel Posted: June 10, 2003 at 11:49 PM (#514001)
No bouquets from Anson from me, but he's in my top 2 for sure.
   44. Howie Menckel Posted: June 11, 2003 at 01:26 AM (#514003)
HOM New York bias? Kinda kidding, but this is getting interesting. Taking a wild guess that Connor gets in this time will leave us with this....

TEAMS WITH SIX HALL OF MERIT MEN
   45. Marc Posted: June 11, 2003 at 01:30 AM (#514004)
Here's an interesting thought. I was looking at candidates known mostly for playing the "hitter" position--e.g. Anson, Browning, Connor, Jones, O'Neill, Orr, Stovey, Thompson. Which of them has the LOWEST career OPS+?
   46. Brian H Posted: June 11, 2003 at 02:52 AM (#514005)
Mark --
   47. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 05:45 AM (#514006)
Here are my two previous posts giving data on this subject:

Thanks, Jim! I have them bookmarked for later use.
   48. MattB Posted: June 11, 2003 at 01:40 PM (#514008)
"And the 1887 team came in FOURTH!!"

Well, the fact that the team had a lot of HoMers doesn't necessarily mean it was a GOOD team that year. George Gore and Jim O'Rourke were both having an off year.

Roger Connor was having a good year, but Dan Brouthers' year for first place Detroit was better.

Ward had a good year, but not as good as Sam Wise's at short.

Buck Ewing had a good year, but he wasn't catching, and his defense at third base was questionable.

And, of course, Tim Keefe wasn't the best pitcher in the league. Either Jim Clarkson or one-year-wonder Dan Casey was.

It is, of course, a problem with over-valuing career-value. Six players having a-little-above-average years may all by HoMers if they keep it up for 15 or 20 years. In any given year, though, they're likely to finish 68-55, 10.5 games behind the first place team that has a bunch of guys with little career value all having great peaks. (First place Detroit won with Fred Dunlap, Sam Thompson, and Hardy Richardson, along with HoMers White and Brouthers).
   49. Jeff M Posted: June 11, 2003 at 02:12 PM (#514009)
I'll reiterate someone's comment from before about the AA discount. We have absolutely no idea how Davenport has derived his numbers, so it is in my opinion suspect for us to assume they are correct.

Also, I'm not sure that the DERA adjustment percentages posted by jimd work for the hitters. For example, I believe the DERA adjustments would say the PL was stronger than the NL in 1990 by approximately 3% (which is also supported by Cramer's numbers). But Pete Browning's Warp2 and Warp3 for the 1890 PL season (adjusted for all-time) reflect a larger discount from Warp 1 than the discount given to Jack Glasscock's Warp1 for his 1890 season in the NL. Just an example that we don't actually understand Davenport's numbers.

I saw on this thread a re-post of Cramer's Hidden Game adjustments to BWA. I like the concept, because at least Cramer was building from the ground up instead of pre-conceiving a discount. However, I thought these adjustments had been largely discredited as an effective tool to discount the statistics of players. Cramer himself acknowledged that the numbers don't work in most contexts without further adjustments (which he never explained).
   50. MattB Posted: June 11, 2003 at 02:22 PM (#514012)
And, as I posted earlier, there is practically no possible reliable data on the relative merits of the AA and NL between 1885 and 1889. Many players shifted leagues between 1884 and 1885, and many more changed between 1889 and 1890, but during the 5 year "heart" of the AA, no more than 2 or 3 players changed from one league to the other, certainly not enough to draw conclusions about relative merit.

It would be like trying to compare Japanese and American baseball if your only data points are Ichiro Suzuki and Tuffy Rhodes. Ichiro plays well here, maybe Japanese baseball's pretty good. Tuffy becomes a home-run champion, maybe Japan is weak.

I can simply see no way to come to any statistically significant conclusion that the AA is 5% worse or 10% worse (whatever that means) than the NL from 1885-1889. There just was not enough movement.
   51. MattB Posted: June 11, 2003 at 02:34 PM (#514014)
I wrote this in the 1901 ballot discussion thread, and am reposting it because new people seem to be asking about AA discounts. My conclusion is in the previous post:

Question: How did the AA compare to the NL during the AA's "Peak Years" (1885-1889, inclusive).

Hypothesis: There was no statistically significant distinction between the two, as could be meaningfully measured by available statistics.

Method: Look at every hitter who played in at least 70 games in either the AA or NL between 1885 and 1889, inclusive, and then played at least 70 games in the other league in the next year (assuming that the next year was also between 1885 and 1889, inclusive). I then compared their unadjusted OPS and OPS+ in each year.

Why only 1885-1889? There was lots of player movement between 1884 and 1885 due to the UA and also between 1889 and 1890 (due to the PL and whole franchises switching leagues). These confounding factors make a direct comparison less valuable.

Raw data:

The study has only 20 data points (13 until I lowered the threshhold from 100 games to 70 games). Paul Radford and Sam Barkley each constitute 2 data points, as they both changed leagues and then changed back. There was very little movement between the AA and NL during these five years among players who could be considered at least "semi-regular." That, in itself, is relevant to any comparison.

Of these 20 data points, 12 moved from the AA to the NL, and 8 moved from the NL to the AA.

Of the 12 that moved from the AA to the NL, 8 moved between 1886 and 1887 and 4 moved between 1888 and 1889. None changed leagues between 1885 and 1886 or between 1887 and 1888.

Of the 8 that moved from the NL to the AA, 1 moved between 1885 and 1886, 3 moved between 1886 and 1887, 0 moved between 1887 and 1888, and 2 moved between 1888 and 1889.

Here are the numbers:

AA ----> NL

NAME AAYEAR OPS OPS+ NLYEAR OPS OPS+
   52. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 03:08 PM (#514017)
Though apparently meant in jest, this is a serious point if consideration is to be given to Start's pre-1871 career. A large proportion of young able-bodied men were not able to play "organized" amateur baseball in 1864 and 1865. Moreover, a significant proportion of young men didn't come back from that war, which affects the talent pool at least into the early 70s. Add that to the incohate nature of the "league" in the 1860s, and any alleged performance during that period has to be viewed with some skepticism, especially with the paucity of hard data.

Well, if we're not going to vote for players who played before 1871, we should institute some type of election reform or we're going to wind up with a greater percentage of players from the 1880s than from the 1980s. Is there anybody here who feels the 19th century decade was more competitive than the 20th century version?
   53. MattB Posted: June 11, 2003 at 03:19 PM (#514018)
I presented the raw data, but the point was really that there were only 20 data points. If the Cramer study and Davenport adjustments for the years in question are based on just these data points, they are not statistically significant.

"Most of the players who were great after contraction were former NL players too, right?"

That is true, but only to a point. First, Brooklyn won the AA championship in 1889. Then, the whole jumped to the NL in 1890 and won the NL championship. Cincinnati finished 4th in the AA in 1889, jumped leagues, and finished 4th in the NL in 1890. I think winning a pennant (or finishing fourth) is different from winning a short post-season series (which the AA won a lot of, too.) In the best possible test case (actually moving a team from one league to another!), the teams in the AA performed identically in the NL. (This is, of course, confounded by the PL, which hurt both leagues.)

Second, note that 2/3 of the players who went from the AA to the NL had their raw OPS go up, and half of the players who went the other way had their raw OPS go down. Did the NL have a higher replacement level than the AA? Did the AA have better pitchers? Did the AA have more pitchers parks? It's hard to say, because many of these points aren't directly measurable.

I'll try to look at pitchers who jumped leagues, too, but I'm guessing I'll have even fewer data points in the years in question.
   54. Marc Posted: June 11, 2003 at 03:20 PM (#514019)
Good discussion about peak and career value. All of us balance the two in our own way, I don't think there is a statistically valid and right way and a wrong way to do that. I consider peak value very strongly, specifically because I disagree with Joe's comment as follows:

>We make a big deal about the pennant impact of 3 or 4 monster years, but 15 pretty good years can still impact a number of pennant races, and cumulatively probably more than the guy that had 3 or 4 big years.

Speaking as a Twins fan, it is obvious to me that most players who are/were not Yankees are lucky to get 1 much less 2 or 3 world championships in a career. Most fans not in New York are lucky to experience 1 or 2 championships in a lifetime. Yet it is exactly those championships that are the ultimate objective of the game. The difference between 1st place and 2nd place is not at all similar in kind much less degree as the difference between 2nd and 3rd.

The 15 years of "pretty good" are (conceptually, cannot prove) the difference between 2nd and 3rd. The 3 or 4 monster years are the difference between 1st and 2nd. So peak value is a very big deal to me.

Of course, all statistical analysis breaks achievements down into smaller pieces. We don't award 1 World Championship Win Share to 1 player. We break up the season into games and award 3 WS per game. In the real world, runs are scored in integers; statistically if a guy hits a single or even works a BB, he gets a fraction of a run. Our methods do not reflect reality in significant ways. So statistically the difference between 15 pretty good years vs. 3-4 monster years and 11-12 years of average/slightly above average may be negligible, but in reality whatever makes a pennant or a world championship possible represents all the difference in the world.

This cannot finally be quantified. But I will continue to regard peak value very very highly.
   55. Rusty Priske Posted: June 11, 2003 at 03:34 PM (#514021)
I have done a complete 180 on Start. I feel that I am finally comfortable with him, thus he will start get good votes from me. I am still unclear on Spalding, and I still don't understand why some people have Hoss so low.

Here is my revised prelim ballot:

1. Radbourne
   56. Howie Menckel Posted: June 11, 2003 at 03:36 PM (#514023)
MattB,
   57. Rusty Priske Posted: June 11, 2003 at 03:54 PM (#514028)
Oh, I left off the comment from my initial prelim ballot. My mistake.

I'm not really putting Radbourne first. My real top 4 are Anson, Connor, Radbourne, Galvin. I am just putting him up there as a comment that I think he should have made it already. I am actually realy disappointed that I two guys popped up this year that are forcing me to move him down.

I'm not so stubborn as to ignore reality however. :)
   58. MattB Posted: June 11, 2003 at 04:13 PM (#514029)
A couple of weeks ago I started looking at a comparison of defensive plays per game by 2B/SS/3B between the 19th century and today to get a handle on the "second base then is like third base today" argument. Here are some numbers. Note that statistics are not my strong point, so I apologize in advance if I screwed up my interpretations of the standard deviations.

I looked at the three positions defensively in the NL of 1886("A"), the AA of 1886 ("B"), and the NL of 1986("C"). I looked only at the "regulars" at each position. I have labelled them A, B, and C to make it easier to present. The number in parenthesis is the standard deviation.

Let's start with second base:

Put Outs Per Game:

A: 2.63 (0.44)
   59. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 04:18 PM (#514030)
This is the result of moving elections back from 1906. While I didn't agree with that move, it's too late to change now. People shouldn't be forced to vote for players from the 1860s just because we're electing too many players from the 1880s. Electing too many players from the 1880s may be a problem with our election system, not with the electors.

I'm not asking anyone to elect players that they don't want. I am asking the electors not to ignore certain players because it's more difficult to analyze their achievements.

The only way that Start can be ignored is if you give him zip credit prior to 1871. Some may deny this, but how do you ignore a career of his length where more than half of it is documented?

Dickey Pearce (he was a star five years before the Civil War, BTW), Start, etc., were great players for their time. Were they better than the best players of 1900? Who knows? Who cares? I know damn well the 19th century players we have been electing probably don't deserve to be in the same room with the best players of today (any doubts about this?) That doesn't mean we should ignore them and elect a Mark Grace to replace them.

BTW, I'm not surly, just passionate. :-)
   60. Jeff M Posted: June 11, 2003 at 04:38 PM (#514032)
Jason, the voter list never closes. Welcome aboard.
   61. Carl Goetz Posted: June 11, 2003 at 04:38 PM (#514033)
Jason Koral,
   62. karlmagnus Posted: June 11, 2003 at 04:41 PM (#514034)
I think I?ve figured out a way to handle the 1870s/1880s comparison. If you ?normalize? all careers to modern standards, assuming that each player played 140 games per season in the ?core? of his career (knock off most ?breaking in? seasons and any that are obviously ?playing out the string? and then add them back after normalizing) then you get figures for hits that compare reasonably with modern hits figures, and with those from 1890 on. This still doesn?t solve the problem of Harry Wright and Dickey Pearce, who were 35-36 in 1871, but it gives a good metric for Start, Pike and McVey.
   63. Carl Goetz Posted: June 11, 2003 at 04:49 PM (#514035)
Prelim ballot-Basically 1902 ballot with Brouthers and Ewing gone and Anson and Connor added.
   64. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 04:50 PM (#514036)
Jason, the voter list never closes. Welcome aboard.

Welcome!
   65. karlmagnus Posted: June 11, 2003 at 04:55 PM (#514037)
Incidentally, if you repeat the process described in #91 above for Paul Hines, George Gore and Ross Barnes, Hines (3333 adjusted hits) makes it but Gore (2342) and Barnes (2363) look pretty dubious HOM selections to me. Barnes gets some additional points for being a 2B, but the others don't. Not much we can do about it now, though.
   66. Marc Posted: June 11, 2003 at 05:01 PM (#514038)
>Marc, you are missing my point. What I'm saying is that the player that had 15 pretty good years is much more likely to play a bigger role in pennant races that the player who was great for 3 or 4 years.

Joe, I didn't miss your point, I disagreed with it. Again, as a Twins fan, two words: Zoilo Versalles. Not that he's a HoMer by any means, but the odd blockbuster year is a huge factor in real world pennant races. Versalles' problem is he only had one; but a guy who has 3 or 4 of them in fact has more impact on pennant races than a whole lot of pretty good. Your statistical argument is fine as far as it goes. But that .2 chance works out to 1 or 2 pennants, same as what a Dave Orr or a Zoilo Versalles might also help to create. We just disagree.
   67. karlmagnus Posted: June 11, 2003 at 05:05 PM (#514039)
Incidentally, if you repeat the process described in #91 above for Paul Hines, George Gore and Ross Barnes, Hines (3333 adjusted hits) makes it but Gore (2342) and Barnes (2363) look pretty dubious HOM selections to me. Barnes gets some additional points for being a 2B, but the others don't. Not much we can do about it now, though.
   68. karlmagnus Posted: June 11, 2003 at 05:07 PM (#514040)
Incidentally, if you repeat the process described in #91 above for Paul Hines, George Gore and Ross Barnes, Hines (3333 adjusted hits) makes it but Gore (2342) and Barnes (2363) look pretty dubious HOM selections to me. Barnes gets some additional points for being a 2B, but the others don't. Not much we can do about it now, though.
   69. karlmagnus Posted: June 11, 2003 at 05:08 PM (#514042)
Sorry for triple post -- VERY confusing system
   70. jimd Posted: June 11, 2003 at 05:11 PM (#514045)
We have absolutely no idea how Davenport has derived his numbers, so it is in my opinion suspect for us to assume they are correct.

Yes and no. Unfortunately, his method is not public so we can't replicate the numbers. However, he has published some articles in Baseball Prospectus on his minor/major league translations, enough so we may be able to sketch out the major league process.

IMO, the backbone of these translations would be the year-to-year major league translations. This would be based on the largest sample you'll ever get, how players in year N perform in year N+1 of the same league. These answer the question, "Is the league getting better or worse, and by how much?" for each year. With this superstructure in place, the smaller samples of player transitions from league to league can be used to place the leagues in relation to each other. Most of that AA data will come from the years 1888-92 when 7 AA teams entered the NL and many AA and NL players jumped back and forth during the PL war of 1890 and the AA war of 1891. The rest of the calculations should be based mostly on the AA's internal evolution back to its beginning, with the sparse interleague data averaged in.

I have some faith in these numbers, largely because the AL/NL data corroborates much of what I've read over the years re contemporaneous opinions about the relationship between the AL and the NL. That the NL was that much better in the 30's came as a bit of a surprise, as was magnitude of the AL superiority over the NL in the late 1990's, but otherwise, spot on.

If this doesn't make any sense, then let me have it, both barrels :-)
   71. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 05:20 PM (#514046)
<i> As an "enemy" of Joe Start, if someone can explain to me why I should consider a 1st baseman who only garners 5 black ink and 102 gray ink points in ~15 "official" seasons, I'd be much obliged.,/i>

Forgetting the obvious age issue again, Start didn't have the luxury of playing during a two-league era for the seventies as the next generation did for the '80s. Without a doubt, if the leagues had been combined into one super league during the 1880s, the black and gray ink results for many a player would have been reduced.
   72. Marc Posted: June 11, 2003 at 06:40 PM (#514055)
> Marc, maybe I'm misinterpreting your argument, but Zoilo Versailles did not create a pennant all by
   73. Chris Cobb Posted: June 11, 2003 at 06:43 PM (#514056)
In interpreting the significance of extrapolated hits from the NA years, wouldn't it also be important to keep in mind the absence of walks from the game? Even if you figure the adjustment based on plate appearances, the ratio of hits to plate appearances will be higher because batters got pitches to hit, and hit them they did. The walk was basically not part of the game. So if you compare adjusted hit totals from the 1870s to hit totals from other eras, the 1870s players are going, proportionately, to have more.

The hit totals thus can give you a better sense of what a player's career would look like, but (for so many reasons) they're unreliable for direct comparisons between players.
   74. Chris Cobb Posted: June 11, 2003 at 06:45 PM (#514057)
In interpreting the significance of extrapolated hits from the NA years, wouldn't it also be important to keep in mind the absence of walks from the game? Even if you figure the adjustment based on plate appearances, the ratio of hits to plate appearances will be higher because batters got pitches to hit, and hit them they did. The walk was basically not part of the game. So if you compare adjusted hit totals from the 1870s to hit totals from other eras, the 1870s players are going, proportionately, to have more.

The hit totals thus can give you a better sense of what a player's career would look like, but (for so many reasons) they're unreliable for direct comparisons between players.
   75. John (You Can Call Me Grandma) Murphy Posted: June 11, 2003 at 06:45 PM (#514058)
First, there is no obvious age issue. I don't think it's too much to demand for a Hall of Merit level 1st baseman to be towards the top of some league leaderboards past the age of 30.

First of all, there is an obvious age issue. He entered the NA at the age of 28. He missed his prime years while playing in the professional leagues.

Secondly, there are major problems with the black and gray ink methods that Joe and I addressed to you.

Thirdly, the man played 26 years. Unless you're ranking your players on 100% peak, I can't see him not making the top ten of everyone's ballot (I have him at #9, BTW).
   76. RobC Posted: June 11, 2003 at 06:56 PM (#514061)
Marc - if you are going to do that make sure you dont give the credit for the monster season on crappy teams. Walter Johnson's 1912-1913 were only good enough to push the Senators into 2nd. And his 1916 only pushed them into 7th. Fortunately his 1924-1925 were good enough to push them into 1st, so he is twice Zoilo Versailles (sticking with the same franchise).

Based on the Washington/Minnesota franchise a 7-9 WARP season will push you into 1st, but a 16 WARP season will only push you into 2nd.

I think you have to reduce it to odds, neither Zoilo or Walter could control their teammates. Not that Johnson had any for most of his career.
   77. MattB Posted: June 11, 2003 at 06:56 PM (#514062)
Mark,

While you're regressing 1871, make sure you don't forget to "progess" 1872 and 1873. Those below-average years of 50-some games each were more likely the equivalent of a bad month or two today. Looking at surrounding evidence, Start's true ability for those two years are likely higher than the numbers show.

In 1871 goes down toward the mean, the next two years will have to go up toward it.
   78. MattB Posted: June 11, 2003 at 07:21 PM (#514064)
Mark,

I think there is a problem with your regressing to the mean technique if you take a guy with below average numbers, regress him, and he turns out worse.
   79. karlmagnus Posted: June 11, 2003 at 07:28 PM (#514065)
You shouldn't and don't need to do the "normalization" season by season. If you do it for the career as a whole you average out good and bad years and avoid the "small sample size" problem. The only problem then is that you still tend to underestimate players like Anson and Start who were better in the 70s than the 80s but played more games in the 80s -- you weight by number of games played, so Start's mega-1871 is weighted only by his 33 games in 1871.

I'm not claiming it's accurate, but I still think it's valid. And moderate inaccuracy doesn't change the conclusions: Anson was a galactic star, Start was a certain HOM'er, Sutton was indeed a top player and McVey and Pike were at least as good as the 1880s players that others have been promoting. Gore and Barnes, OTOH, were nothing all that special.
   80. Chris Cobb Posted: June 11, 2003 at 07:51 PM (#514066)
"I'm not claiming it's accurate, but I still think it's valid. And moderate inaccuracy doesn't change the conclusions."

With so many great players on the ballot, small differences in estimations of value can move players 5 or 6 spots on the ballot. I need an estimation I can really trust, then, if I'm going to move Joe Start from 9 or 10 on my ballot to 4 or 5, jumping over players like Sutton and Glasscock.
   81. karlmagnus Posted: June 11, 2003 at 08:02 PM (#514067)
What I mean by "Valid but not necessarily accurate" is that if there's no major conceptual flaw, then the only question is whether you normalize to 130 or 140 games (130 is too low, but takes account of errors effect on at-bats. At 130 Start is at 2,636 hits after age 28 1/2, or perhaps 3,700 for career adding good but not great 1864-70. Way ahead of Sutton (2,460 normalized to 130 games) or Glasscock (2,382)
   82. karlmagnus Posted: June 11, 2003 at 08:31 PM (#514074)
In Start's first 3 seasons, 1871-73, including the 2 "bad" seasons 1872-3, he played 141 games (just about a season) and made 201 hits -- up there on the leaderboard in any modern season. My "normalization" says that if they'd played "normal" seasons back then he'd have played 390-420 games and got 201x(390-420)/141 = 556-599 hits. That seems a reasonable, and not "structurally" inaccurate way to benchmark his career as a whole.
   83. Marc Posted: June 12, 2003 at 02:09 AM (#514077)
Hey Joe, et al, what is the procedure for a proxy? I am going out of town for a week. I maybe could vote a week from Monday but I'd rather send in a ballot tomorrow or Friday. I could post it here or email it to Joe...?
   84. jimd Posted: June 12, 2003 at 02:35 AM (#514078)
I don't think anyone can accurately take a 30 game season and prorate it to a 140,154,or,162 game season.

It all depends on what you're trying to measure. As a projection, you're right; it's bull****. OTOH, if you're trying to measure value using some win-based metric (Win Shares, WARP, etc.), there's no doubt that Meyerle had a huge impact; he hit .492 for the SEASON (whatever the length), and his OPS+ was similar to Ted's in 1941. It was instrumental in his team winning the first pro league pennant, and would make him a front-runner for its MVP. (He also played in a league where the average team scored over 10 runs per game, each; compared to 1871 all other seasons are low-scoring.)

Putting the seasonal value onto a common scale makes comparisons easier. If it makes you more comfortable, you could scale all seasons thereafter down to 30 games, but it's less work to go the other direction.
   85. DanG Posted: June 12, 2003 at 03:20 AM (#514079)
I'm not around alot this week (okay, quit cheering), but a couple things I wanted to touch upon.

Welcome to Jason Koral. As with all recent arrivals, I urge you to read as much of the archives on this blog as time allows. For instance, much of your questioning of Joe Start's qualifications has been hammered at for awhile; the 1902 Ballot Discussion thread had lots of good stuff.

Another example of this is someone's questioning of our voting procedures on another thread. If anyone is truly curious, there was a lengthy discussion before we settled on the current system. Suffice to say, there are very good reasons why it's set up as it is.

Devin asked about the SABR 19th century survey. I haven't found that one yet. But, four years ago, SABR conducted a survey of its general membership as to who were the greatest 19th century players. FWIW, here's how they voted:

1-Anson
   86. Brian H Posted: June 12, 2003 at 04:58 AM (#514080)
I like that list ! It has "(Poor)Old Hoss" ahead of all of the Pitchers elgible to date.
   87. Philip Posted: June 12, 2003 at 11:06 AM (#514081)
"I don't think anyone can accurately take a 30 game season and prorate it to a 140,154,or,162 game season. Only actual at bats, hits etc. should ever be considered."

I don't agree with this. I think you can extrapolate because over the course of a few seasons the extremes will even out. Of course you will see some bizarre seasons for reasons mentioned often enough, but this will work both for and against a player if you look at a span of several years.
   88. Marc Posted: June 12, 2003 at 01:05 PM (#514083)
Yes, that SABR list is disappointing in that it is so...so...conventional. Where's the insight? Glasscock and Dahlen in the 30s behind Duffy?

My major moment of enlightenment came in the SABR Centennial Survey, the year they elected Ray Schalk. I wish I could remember who he beat out for the honor that year. But it clearly showed me that the average SABR member is not Bill James or Pete Palmer, but rather is captive to the same conventional thinking as the average baseball fan.

BTW, I think somebody said that the survey picked Stovey as the best player not in the HoF which turns out not to be accurate. Pls take note.

PS. Or is Cap Anson just Pete Rose?
   89. dan b Posted: June 12, 2003 at 03:02 PM (#514085)
Marc ?Re: Stovey ? different SABR survey. Stovey was named ?most deserving of HOF recognition? by SABR?s 19th Century Committee, not the general membership. My source for this information is the SABR publication ?Nineteenth Century Stars? which pre-dates the SABR-wide poll conducted 4 years ago.
   90. Howie Menckel Posted: June 12, 2003 at 03:45 PM (#514087)
what the heck, a tentative ballot..

1. Cap Anson
   91. Philip Posted: June 12, 2003 at 04:08 PM (#514091)
"If the standard measure of value is a pennant, why is an NL pennant worth more than an AA pennant?"

Because the NL was a better league. So yes, an NL pennant is worth more than an AA pennant. Since the NA was the only pro-league of its time the top players all competed in the same league and so an NA pennant is worth more than an AA pennant. This is the basis of assessing value which is, of course, the whole point of the HoM.
   92. Philip Posted: June 12, 2003 at 04:08 PM (#514092)
"If the standard measure of value is a pennant, why is an NL pennant worth more than an AA pennant?"

Because the NL was a better league. So yes, an NL pennant is worth more than an AA pennant. Since the NA was the only pro-league of its time the top players all competed in the same league and so an NA pennant is worth more than an AA pennant. This is the basis of assessing value which is, of course, the whole point of the HoM.
   93. Rick A. Posted: June 12, 2003 at 04:16 PM (#514094)
Prelim. ballot

1. Roger Connor
   94. Marc Posted: June 12, 2003 at 04:48 PM (#514097)
Here's another way to think about career value vs. peak value. If you were starting a baseball team (real, fantasy, whatever since this is all make believe anyway), what would you pick for your team? 27 years of Cap Anson or 18 years of Roger Connor (or 19 of Dan Brouthers)?

How about 17 years of Johnny Bench or 24 years of Pete Rose? 13 years of Joe DiMaggio or 19 years of Ted Williams or 22 years of Stan Musial? 12 years of Dizzy Dean or Sandy Koufax (or, really, only six of each) or oe 10+ years of Teddy Lyons or Don Sutton?

It's fun and easy to make up your own comp/not comp pairings in the privacy of your own home!
   95. John (You Can Call Me Grandma) Murphy Posted: June 12, 2003 at 05:24 PM (#514100)
Prelim:

1) Cap Anson (n/a): Far and away the most value of all eligible players since 1898.

As someone who originally wasn't going to vote for Joe Jackson in any election and supported the Anson snub, what is the point of doing it here? I can understand it for the Hall of Fame because of its place in baseball (I would never vote for Jackson there), but our tiny (but much better organized) hall is not really bestowing an honor but just highlighting the best players throughout baseball history. Just a thought.

2) Al Spalding (3):
   96. John (You Can Call Me Grandma) Murphy Posted: June 12, 2003 at 05:40 PM (#514102)
Here's another way to think about career value vs. peak value. If you were starting a baseball team (real, fantasy, whatever since this is all make believe anyway), what would you pick for your team? 27 years of Cap Anson or 18 years of Roger Connor (or 19 of Dan Brouthers)?

Here's another way to look at it:

Would you pick Al Rosen (now born during the forties) or Sal Bando to man third base for the Mets during the sixties? Rosen was by far the best player at his peak (but didn't last too long), while Bando had a very long career with a smaller peak. Since we know the Mets had crapola at the hot corner for decades, Bando seems much more appealing to me than Rosen.

Peak value has its limitations.
   97. John (You Can Call Me Grandma) Murphy Posted: June 12, 2003 at 05:43 PM (#514103)
Okay, maybe if Wagner had played more games there earlier in his career, he would have ranked higher.

Without thinking twice, Wagner over Glasscock. Easy decision. Not even debatable.
   98. jimd Posted: June 12, 2003 at 06:01 PM (#514104)
"If the standard measure of value is a pennant, why is an NL pennant worth more than an AA pennant?" etc.

OTOH, each of us appears to have our own way of integrating the data. And who's to say which way is right and which is wrong?

If we all agreed on everything, the process wouldn't be so much fun. We wouldn't have George Wright elected while not making the top 15 on two ballots. We wouldn't have these great debates about Joe Start or Al Spalding, each of whom has about half-a-dozen "Elect him now" advocates, and a similar number of people who can't find a ballot spot for him.
   99. Marc Posted: June 12, 2003 at 06:46 PM (#514106)
>>Because the NL was a better league. So yes, an NL pennant is worth more than an AA pennant.

>Philip, I assume then that as an extension of this reasoning, you are giving more weight to accomplishments in the 1880s than the 60s or 70s, since the leagues in the 80s were better than their predecessors.

Two points. 1) These are different issues. (A) is an in-season discount between better competition and poorer competition the same year. (B) is the timeline argument. I buy A but not B. As to B, I like the "pennant is a pennant" argument. I can buy that "a pennant is a pennant," in other words, relative to one but not the other. They're different issues.

2) It's not obvious that the caliber of play in the '80s (2 leagues, 16 teams) was better than the '70s, especially not the NL of the later '70s (1 league, 8 teams). Some have argued, persuasively, that the NL of '79-'80 was not improved upon until '92.
   100. Marc Posted: June 12, 2003 at 06:48 PM (#514107)
Or to put it another way, Ross Barnes played against the best available competition anywhere. Harry Stovey did not.
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Tuque
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

Syndicate

Page rendered in 0.7740 seconds
43 querie(s) executed