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, August 02, 2004

1932 Ballot Discussion

1932 (August 15)—elect 2
WS W3 Rookie Name-Pos (Died)

266 65.1 1913 Wilbur Cooper-P (1973)
215 70.4 1913 Hooks Dauss-P (1963)
243 60.1 1909 Babe Adams-P (1968)
227 51.1 1910 Stuffy McInnis-1B (1960)
206 59.6 1918 Ross Youngs-RF (1927)
142 38.5 1914 Everett Scott-SS (1960)
134 37.0 1912 Hank Severeid-C (1968)
157 32.0 1914 Jimmy Johnston-3B/RF (1967)
106 25.8 1914 Bill Wambsganss-2B (1985)
105 28.2 1916 Whitey Witt-CF/SS (1988)
120 27.2 1916 Carson Bigbee-LF (1964)
Negro Lg 1910 Louis Santop-C (1942)
Negro Lg 1908 Jose Mendez-P (1928)
Negro Lg 1920 Dobie Moore-SS ()

Thanks to DanG for the necrology:

Players Passing Away in 1931

HoMers
Age Elected

75 1905 Hardy Richardson-2B/LF
73 1903 Roger Connor-1B

Candidates
Age Eligible

79 1894 Jack Burdock-2B
79 1890 George Bradley-P
74 1896 Joe Hornung-LF
72 1900 Charlie Comiskey-1B/Mgr
66 1904 Jimmy McAleer-CF
62 1913 Jack McCarthy-LF
57 1915 Jack Chesbro-P

Joey Numbaz (Scruff) Posted: August 02, 2004 at 10:17 AM | 328 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 3 of 4 pages  < 1 2 3 4 > 
   201. John (You Can Call Me Grandma) Murphy Posted: August 06, 2004 at 08:08 PM (#781772)
This question is bothering me this week. Can someone explain the gulf that separates Mickey Welch from Vic Willis? To me, they come off as quite similar when you adjust for era differences.

I think Willis is a notch better than Welch, but I do agree that the gulf shouldn't be that great as expressed through our ballots so far, regardless. I think some voters are mesmerized by the 300 WINS!!! in regard to Welch. Well, Willis would have been a 300 gamer if he had pitched during the 1880s, too.
   202. John (You Can Call Me Grandma) Murphy Posted: August 06, 2004 at 08:13 PM (#781782)
Wait... I just found this

Even though it's a cached page, I have included it as part of the Hall of Merit links. I was having problems resurrecting that thread from the BTF site, so that will be better than nothing. Maybe Joe can help us create a thread so that we can post to it (and maybe he can show me how to do it! :-).
   203. Kelly in SD Posted: August 06, 2004 at 08:34 PM (#781842)
I am working on a Willis profile right now. Maybe this will help for comparisons.
   204. OCF Posted: August 06, 2004 at 09:13 PM (#781904)
Willis in one of the two pitchers for whom I've put in a quality of team defense adjustment in the RA+/pythag system. (The other was Mordecai Brown.) Without this adjustment, Willis would look a lot better than what's listed below. Here are his equivalent records, year by year:

1898 18-16 (105 def adj RA+)
1899 24-14 (138)
1900 12-14 (092)
1901 21-13 (133)
1902 26-19 (119)
1903 19-12 (128)
1904 20-19 (101)
1905 17-21 (092)
1906 24-11 (156) (His best year)
1907 19-14 (119)
1908 18-16 (106)
1909 18-14 (120)
1910 11-12 (094)

That adds up to 248-196. Put that into FWP and it's 192, which is essentially tied with Vance (194), Waddell (191), and the defense-adjusted Brown (193) and ahead of Cooper, Powell, and Griffith (all 175-180).
   205. Paul Wendt Posted: August 06, 2004 at 11:26 PM (#782107)
David Foss:
Anyone have a list of those who were "grandfathered" and thus allowed to throw spitballs past 1920?

http://world.std.com/~pgw/Deadball/spitball.html
The Spitball Pitch: legal denouement
   206. Paul Wendt Posted: August 07, 2004 at 12:16 AM (#782291)
David Foss:
Its possible that the AL stars may not have affected the "base-level" of play for the league. Is there any evidence that would support that the AL also had more weak players and weak teams... not enough to reverse the discount, but enough to balance things?

Michael Schell (Biostatistics, UNC) believes that "standard deviation measures league talent" inversely. He uses a version of league SD as a measure, which is not an argument, but it is clear that he believes it. [Baseball's All-Time Best Hitters, Princeton U P, 1999, chapter 4.]

For example, "Many of Cobb's American League compatriots would likely have ridden the National League bench during the same era." That era is
1910-1914, SD batting average was AL .043, NL .030. [Page 91]
   207. Kelly in SD Posted: August 07, 2004 at 12:22 AM (#782311)
Vic Willis:
Teams: Bos 1898-1905, Pit 1906-09, StLN 1910.
Record: 249-205 2.63 era/3.67 runsallowed, K/W 1.36, WH9IP=10.88
Win Shares: Career 293; 3 yrs cons 84; 7 best yrs 199; per 40 starts 24. Seasons with 20+: 8. Seasons with 30+: 2.
AllStars: STATS 2, WS 4
Fibonacci WinPoints: 180
ERA+: 118
Black Ink/Grey Ink: 25/204
Bill James Rank: 84
Top 10s: Wins and Innings 9 times with 1 first in IP. SHo 9 times with 2 firsts. Ks 7 times. W%age 6 times. ERA and H/9IP 5 times with a first in each. aERA+ 5 times. Most walks 8 times.
Other/Unique: 67-19 against under .500 teams while with Pit while 29-27 against .500 or better. Lost final game of 1908 season against TF Brown, 5-2. If he won the Pirates would have finished 99-55, NY 98-55, Chi 97-56. That would put the make-up game in a totally different light. Lost Game 6 of 1909 WS. Agreed to jump to the AL in 1901 but changed mind. Agreed to jump to Detroit in 1902, but changed mind.
Teams he played for had losing records only in 1903, 04, 05, and 10.

Career Breakdowns:
Against HoMers:
Young      0-1
Nichols    1-2
McGinnity  2-9
Mathewson  7-3
Brown      3-4
total:    13-19
Also: 
Griffith   1-2
Against other STATS AllStars from 98-10: 10-15
Record depending on opponent's finish:
finish   Willis
1st      20-34
         .370
2nd      22-24
         .478
3rd      26-27
         .491
4th      31-27
         .535
5th      32-28
         .533
6th      36-26
         .563
7th      32-12
         .727
8th      39-22
         .639
9-12     20-8
         .714

Records against over/under .500 teams
Record  Rusie
.500+   111-125
         .470
.500-   147-84
         .636

% of career dec'n vs. .500+teams
.505
   208. Chris Cobb Posted: August 07, 2004 at 01:38 AM (#782606)
Here's my take on Welch vs. Willis.

Dan's question was highly interesting. As I set out to answer it, I came to a new understanding of why one might see them as alike, and why one might see them as significantly different.

Some of it depends on how you count Willis's two bad years.

For his career, Welch was (by my reckoning) about 28 support-neutral wins above average in 4800 ip. (Might be more like 30 -- I'm in the process of revising the way I treat fielders' error rates in my defensive support formula, and I haven't formally recalculated Welch, or anybody else from the 1880s, yet). He was above average in innings pitched in 6 of 11 seasons.

For his career, Willis was, by my reckoning, about 11 support-neautral wins above average in 4000 ip. He was above average in innings pitched in 11 of 12 seasons.

Looks like Willis was more durable, but less valuable on a per-inning basis than Welch. Of course, even though he was more durable, he threw proportionately fewer of his team's innings than Welch did, because of changes in pitcher usage post 1893.

However, Willis had two very bad seasons in which he threw about 600 innings and was about 9 wins below average between them. He was pretty close to a replacement level pitcher in both of those years. _If_ you want to throw them out, he becomes a pitcher who was about 20 wins above average in 3400 innings pitched, throwing an above average number of innings in 10 of 10 years as a full-time starter.

20 wins above average in 3400 innings is the same rate of value as 28 wins above average in 4800 innings. If pitching 340 innings in 1905 is about like pitching 480 innings in 1885, we have a Vic Willis who looks a lot like Mickey Welch.

And that is, I surmise, what Dan G is looking at when he asked the question (which I had not seen before).

Have I argued myself into seeing the two as indeed quite similar? No, although I now understand why one would, and could, see them as similar. I suggest two reasons why one should see Welch's performance in his context as quite a bit more impressive than Willis in his.

1) Welch's value above average is more impressive. In the pitching conditions of the 1880s, there was not all that much per-inning difference in value between a great pitcher and an average pitcher, so Welch's 28 wins above average put him pretty close to the other top pitchers of the time. I have Galvin at 37 W+ in 6000 innings, Clarkson at 35 in 4500 innings, Keefe at 30 in 5000 innings, Radbourn at 30 in 4500 innings.

Willis, on the other hand, at 20 W+ in 3400 innings, trails his contemporaries quite significantly. Walsh is at 39 W+ in 3000 ip, Plank at 41 W+ in 4500 ip, Brown at 30 W+ in 3300 ip, McGinnity at 36 W+ in 3400 ip, Mathewson at around 75 W+ in 4800 ip, Waddell at 30 W+ in 3000 ip, Joss at 30 W+ in 2300 ip. When the top pitchers were dominating like that, Willis’s achievements appear less impressive.

2) Willis’s durability is less meaningful than Welch’s. Since pitchers threw proportionately fewer innings and the per inning difference between a great pitcher and a good pitcher was more in Willis’s era, throwing a lot of average innings just didn’t matter as much to a team as it would in the 1880s.

So, in context, it’s easy to see that Welch was both more valuable and more impressive. Attempting to find some all-time way of comparing the pitchers, I look at it this way. Willis might have been great if he had started his career twenty years earlier, and Welch might not if he had started his twenty years later. However, Welch was regarded as a smart pitcher and a great competitor, so I think he would have found ways to win in era. Don’t know how Willis was regarded, but his talents would have only had great value in the very early game.

In the end, the decisive factor for me is that according to the numbers I have, in comparison to peers, Welch was clearly among the best of his time and Willis clearly was not. If one translates their performances into one another’s eras, their accomplishments do look similar, _if_ one drops Willis’s bad years out of the picture. That’s too many ifs for me to make a compelling case for Willis.

Others with different numbers and different approaches to comparing different periods may reach different conclusions.
   209. jimd Posted: August 07, 2004 at 02:11 AM (#782739)
believes that "standard deviation measures league talent" inversely.

There are two reasons for a lower standard deviation. One is the reason cited by Michael Schell, that is, a higher mean that reduces the standard deviation of the outlier. The other is a lack of outliers to stretch the standard deviation.

I prefer the latter explanation for this case, unless someone wants to argue equivalances like Wheat==Cobb, Roush==Speaker, Doyle==E.Collins, but we just don't realize it because the NL was so much better.

IIRC, Michael Schell's argument can also be used to show that the AA was better than the NL, in which case we've messed up the 1880's pretty badly. Again, I think it's more likely the AA lacked the outliers.
   210. Chris Cobb Posted: August 07, 2004 at 02:42 AM (#782823)
On his prelim ballot, Daryn wrote the following:

Andrew Foster – On the Foster thread Chris Cobb has a great detailed analysis of Rube’s career – 241-176 MLE, 314 MLE Win Shares. We have made those kind of pitchers first ballot HoMers. Plus, like Frank Chance and Roger Bresnahan, we are allowed/encouraged to take into account his managing as a player manager. Plus, he was a good hitter.

Three comments.

First, I should remind everyone that Joe D. mentioned an election or two back that the Constitution is in error on the taking into account managing as a player manager and that that passage should be revised. So, by the letter of the Constitution we are allowed to take account of managing by player-managers, but we are certainly not encouraged to do so.

Second, I should mention that I have posted a small downward revision of my MLE win-share estimate for Foster on the Rube Foster thread, from 314 to 301, based on problems with the estimate system I identified when working on Jose Mendez this week (along with some other comments on how I treat this win-share total in my rankings).

Third, I would say that a 241-176 support/neutral w-l record (or 240-182 as Chris J. surely more accurately estimates it) has not generally been first-ballot HoM territory for pitchers. It’s probably an HoM-worthy record, but it’s not slam-dunk by any means.
   211. DanG Posted: August 07, 2004 at 03:40 AM (#782911)
Chris Cobb:

As usual, a valuable, yeoman effort. Since I abhor writing lengthy responses, let me toss these out there:

jimd wrote The other is a lack of outliers to stretch the standard deviation.

I prefer the latter explanation for this case,


Isn't this similar to your presentation, showing Welch's era lacking in outliers compared to Willis'? Could we say that the fact that Welch is closer to the best of his time is due to the lack of outliers, an era devoid of dominant pitchers? Welch is near his HoMer contemporaries, but seems clearly below them. How far above the next guys is Welch? Guys like McCormick, Mullane, maybe someone else? You have to draw the line somewhere--I like to find a gap to draw the line--so is the gap between Welch and the HoMers less than the gap between Welch and the others? If so, I like his case better.

Another thought takes me back to revisiting some very early ideas in this project. The old question of When did pitching become valuable? We know that 1860's pitchers basically had the job of tossing it up there just to get the action started. There were few strikeouts and next to no walks. Defense was key and a Pitcher's fielding was about as important as his pitching. Certainly his hitting ability was more important. Like a slow pitch softball pitcher, a good pitcher could have some effect but it was essentially up to his defense to make or break. All of this justifies our nearly complete dismissal of pre-1876 hurlers (except for Spalding).

Gradually this changed and by Willis' time pitchers could dominate a game. It mattered little how they hit and pitchers could win without a great defense behind them.

The question is what about Welch? My impression is he falls closer to the early "action-starter" pitchers than the subsequent "dominator" pitchers. Or does he? How valuable was Welch? Was he a more easily replaced commodity than Willis? Was it easier to be a Welch or a Willis? Or am I going up a blind alley with this line of thought?
   212. OCF Posted: August 07, 2004 at 03:52 AM (#782929)
However, Willis had two very bad seasons

I saw that and my first question was, "What two years is Chris talking about?" Clearly, one of those years was 1900. I checked a little further, and I found an intriguing discrepancy. I've been using the park factors printed in the STATS, inc. All-Time Major League Handbook, first edition (copyright 2000). For 1900, this book shows Boston having the rather startling park factor of 121, sandwiched between a 103 in 1899 and a 109 in 1901. It's not just a misprint in Willis's entry - it shows the same way for Nichols. The Baseball-Reference park factors for all three years cluster around 110. Nichols did not take a big hit in his ERA/RA in 1900 but Willis did. Is that 121 believable? And why are the two sources so very different?

With the 121 park factor, Willis's RA+ is 92 and his equivalent record is 12-14. Below average, yes, but clearly above replacement. If you were to change that park factor to 111, the RA+ would go down to 85 and the equivalent record to 11-15.

The flip side of this calculation is that if you were to change the 1899 park factor from the 103 that STATS has for it to 108, Willis's RA+ would go from 138 to 145 and his equivalent record from 24-14 to 25-13.

What's the other bad year? 1905, I suppose. Willis pitched 350 innings in 1904, 342 in 1905. In 1904, he allowed 111 ER, which jumped up to 122 ER in 1905. However his runs allowed went went from 174 to 174. He was a worse pitcher in 1905 than in 1904 - the same runs in slightly fewer innings in a lower rund environment - but I have the RA+ going only from 101 to 92, and his equivalent record in those two years 20-19 and 17-21. Again, below average in 1905 but above replacement. (His actual records for both of those years were horrid: 18-25 and 12-29.)
   213. John (You Can Call Me Grandma) Murphy Posted: August 07, 2004 at 03:55 AM (#782934)
First, I should remind everyone that Joe D. mentioned an election or two back that the Constitution is in error on the taking into account managing as a player manager and that that passage should be revised.

I'll take care of that, Chris.
   214. OCF Posted: August 07, 2004 at 04:57 AM (#782990)
the rather startling park factor of 121

The park factors for 7 years of Vinny Castilla's career, 1993-1999, from the same source: 122, 117, 127, 129, 122, 127, 127.
   215. Chris Cobb Posted: August 07, 2004 at 05:05 AM (#783002)
Dan G. wrote:

Could we say that the fact that Welch is closer to the best of his time is due to the lack of outliers, an era devoid of dominant pitchers?

Interesting idea!

As with the exchange between KJOK and Kelly in SD about great catchers, it’s an issue of interpretation. Are we seeing a lack of great players, or are we seeing an effect of particular playing conditions? Since (1) dominant pitchers appear right around 1893, (2) I don’t believe we will see any stretch of subsequent baseball history without dominant outlying pitchers, (3) there are plenty of batting outliers in the 1880s, and (4) people certainly thought that Charley Radbourn was a dominating pitcher in 1884, I lean towards the view that the nature of the pitching enterprise at that time set a lower limit on dominance than was later achievable. The alternate interpretation, that there are simply no great pitchers, is also plausible, but I see fewer arguments in its favor.

You asked Dan, if your concluding thoughts were leading to a blind alley -- I don't think they are, but I can't suggest a way out of it yet. All I can say is that pitchers were dominating enough by the 1880s that their batting was nearly irrelevant to their value, but they weren't dominating enough to win games all by themselves.
   216. Chris Cobb Posted: August 07, 2004 at 05:18 AM (#783015)
OCF:

You picked the two years right out. Interesting weirdness with the 1900 park factor; a difference between using 1-year and 3-year, I guess, compounded by the 1900 contraction season's inherent weirdness.
   217. OCF Posted: August 07, 2004 at 06:24 AM (#783044)
a difference between using 1-year and 3-year, I guess

That line sent me to the explanation in the introduction to the STATS book. It turns out that they use centrally-weighted 5-year averages, except when there's a clear change in park - from 1909 to the present. Before 1909, they use single-year park factors. James explains that by invoking the impermanence of parks before the age of "modern" steel and concrete parks built mostly between 1909 and 1920.
   218. Paul Wendt Posted: August 07, 2004 at 04:19 PM (#783232)
DanG
The old question of When did pitching become valuable? [Granted.] We know that 1860's pitchers basically had the job of tossing it up there just to get the action started.

Pitchers were probably trying to put batters out when the NABBP was established in 1857-1858. (There was substantial agreement at a SABR34 special session on 19c base ball.) Bill Ryczek says that many teams in the 1860s played with seven amateurs and a professional battery, which shows perceived value. We know that Jim Creighton was putting batters out ~1860 and was extraordinarily influential. The only question regarding competitive clubs seems to be how soon how many pitchers were how valuable.

From: John Thorn <jthorn@newworldsports.org>
Date: Fri Jul 23, 2004 9:23 am
Subject: Finding Frank Pidgeon
. . . a story I wrote for the Woodstock Times (and Saugerties Times) about the Eckfords' star pitcher of the 1850s, about whom I'll be glad to learn more from any of you.

John Thorn on Frank Pidgeon
   219. Paul Wendt Posted: August 07, 2004 at 04:24 PM (#783235)
   220. OCF Posted: August 07, 2004 at 04:25 PM (#783238)
I haven't usually been posting preliminary ballots, but this one's ready to go so I might as well put it out there.

1932 preliminary ballot. This isn't a week for big shakeups for me; that will start with the next ballot.
1. Louis Santop (new) There is a reasonably persuasive case for him - and face, it, the competition isn't exactly stiff.
2. George Van Haltren (10, 9, 8, 2, 1) No great enthusiasm for this choice - he gets here by attrition. As "peakless" careers go, he's got substantially more offensive peak than the likes of Beckley or Hooper. Not much pitching value (and it was a whole lot easier to be a pitcher-hitter before 1893 than after), but what little pitching there is serves as a tiebreaker among similar candidates.
3. Jimmy Ryan (8, 7, 6, 3, 2) Nearly indistinguishable from Van Haltren. 24rd year on my ballot.
4. Larry Doyle (3, 3, 2, 5, 3) Big hitter in low scoring times - nearly as good a hitter as the available outfielders. Mediocre defense, but occupied the position for a long time. I must not be a believer in the NL discount.
5. Hugh Duffy (9, 8, 7, 4, 4) 26th year on my ballot.
6. Rube Waddell (7, 6, 4, 6, 5) I continue to see him as the best available pitcher. This week's discussion doesn't change much for me, but it doesn't hurt him.
7. Roger Bresnahan (11, 10, 9, 8, 6) Very good offense for a catcher; not enough if we think of him as an outfielder.
8. Gavy Cravath (8, 12, 11, 10, 9, 7) In the system I use, the biggest offensive peak of anyone other than Chance (well, maybe Tiernan). Yes, he took unique advantage of his park, but real wins resulted from that. Seriously lacking in bulk unless you also consider his work in Minneapolis.
9. Mickey Welch (15, 14, 11, 5, 7, 8) The more I learn, the more confused I get. I'm noticing renewed support for McCormick and Mullane - recognition that there's not much between them and Welch and Caruthers?
10. Jake Beckley (20, 19, 18, 10, 11) No peak, long career. But still more peak than Hooper.
11. Frank Chance (16, 15, 13, 11, 10) I could have him higher; huge offensive seasons, discounted for his lack of playing time.
12. George J. Burns (----, 11) My new favorite leadoff hitter, ahead of Thomas and Hartsel. (See also the NBJHBA.)
13. Vic Willis (13, 12, 11, 12, 12) I could reasonably have him right up there with Waddell.
14. Andrew Foster (23, 21, 21, 19, 13) Maybe a longer effective career than I had been giving him credit for.
15. Harry Hooper (----, 14) Less peak than Van Haltren, less peak than Beckley - it's sheer length of career as a good player that gets him on this thin ballot.
16. Babe Adams (new) All the energy I spent this week talking him up, and this is the best I can do for him. Oh, well.
17. Herman Long (-, 25, 15, 14, 15) The key man in a great team defense. Scored some runs, too - twice over 100 R*.
18. Joe Tinker (19, 18, 14, 13, 16) Defense at SS does matter.
19. Clark Griffith (18, 17, 17, 17, 17) RA+ PythPat of 203-146.
20. Bobby Veach (----, 18)
21. Wilbur Cooper (new)
22. Roy Thomas (17, 16, 16, 15, 19)
23. Johnny Evers (15, 14, 12, 16, 20)
24. Spotswood Poles (-, 20, 18, 21)
25. Hughie Jennings (24, 22, 22, 20, 22)
   221. Paul Wendt Posted: August 07, 2004 at 04:36 PM (#783258)
Schell's argument implies that the AL was weaker 1910-1930; stronger 1901 and 1960-1975. That isn't plausible.
But he does make an argument re 1910-1914 that I merely hinted above, and Zach Wheat isn't on the radar screen. In AL1910-1914, 4% batted above .350 and 12% below .220 (mean-adjusted). In the contemporary NL and in both leagues 1980-1984, the shares were 0% and 6%.
   222. Paul Wendt Posted: August 07, 2004 at 05:27 PM (#783299)
Sorry, I missed that "Estimating League Quality" has been revived as a Hot Topic, so it shows in the right sidebar.
Estimating League Quality (Part I, the concept)
   223. Adam Schafer Posted: August 07, 2004 at 09:17 PM (#783597)
I am going to be out of town all week, and would appreciate it if someone could cast this ballot for me when the balloting begins for the 1932 election.

1. Mickey Welch (2) - From all of the discussion that we've had on him, I can't see NOT ranking him #1. He's not quite as good as Keefe, but then again, Glavine wasn't quite as good as Maddux. That doesn't make Glavine's career any less impressive in my opinion.

2. Louis Santop (n/a) - The first Negro Leaguer to rank so high on my ballot. I have no doubt he belongs here.

3. Clark Griffith (2) - Big jump for Griffith. All of the talk on Welch has had me reevaluate Griffith. Turns out Griffith was much more valuable than I was giving him credit for.

4. Rube Waddell (3) - The top 5 in strikeouts for 10 consecutive years. He's #10 in the all-time ERA leaders.

5. Lip Pike (4) - I bump him ahead of a couple others this year as I am convinced he was a bigger stud than I was willing to let myself believe. I can see him finally getting in one of these days.

6. George Van Haltren (5) - Moves ahead of Beckley and Bresnahan.

7. Jose Mendez (n/a) - I thought I'd have him a lot higher than this, but I just don't feel like I have a firm grasp of his career yet. I do feel much more comfortable with him than I do with Rube Foster though.

8. Jake Beckley (6) - Big drop for a guy that would've been #2 on my ballot this year. I didn't find any reason to like him any less, I just found justification in moving several others higher than him.

9. Roger Bresnahan (7) - It's no secret that I love catchers. I would've ranked Roger higher had he caught more and played the OF less during his peak years.

10. Hughie Jennings (8) - Nothing new to add

11. Bobby Veach (9) - Not enough career for him to merit a higher ranking on my ballot, but enough peak to grab a lower spot.

12. Jimmy Ryan (10) - A watered down Van Haltren

13. Eddie Cicotte (11) - Underrated in my opinion. May not be HOM material, but underrated nonetheless.

14. Hugh Duffy (13) - Back onto my ballot. No new thoughts on him

15. Harry Hooper (14) - nothing overly impressive about his career. I originally thought he would rank much higher than this on my initial ballot, but he just doesn't meet the qualifications in my mind that everyone above him does.

Rube Foster would be coming in at #16 for me. I'm just not convinced that as a PLAYER he is HOM worthy. Should we take all of his other contributions into consideration, he'd be a slam dunk.
   224. John (You Can Call Me Grandma) Murphy Posted: August 07, 2004 at 10:02 PM (#783724)
I am going to be out of town all week, and would appreciate it if someone could cast this ballot for me when the balloting begins for the 1932 election.

No problem, Adam.
   225. Kelly in SD Posted: August 07, 2004 at 10:48 PM (#783810)
I posted a career sketch of Ed Walsh on the Pitchers Thread.
   226. Kelly in SD Posted: August 07, 2004 at 11:34 PM (#783941)
I posted a career sketch of Addie Joss on the Pitchers Thread.
   227. Brent Posted: August 08, 2004 at 12:45 AM (#784033)
A couple of days ago, Chris J. pointed us to his estimates of pitcher W/L records adjusted for run support and defense (using win shares above average). Thanks, Chris. He also added the following comment:

Also, offense, & defense aren't the only things that affect a pitcher's record. There's also that "third factor" that I intentionally don't define. Near the bottom of all those pitchers' notes section it says "Wilbur Cooper won 10 fewer games than would have been expected" or words to that effect.

The standard deviation of success in 394 trials is about 10, so I’d say the difference can most easily be attributed to chance. My understanding of Chris’s methodology is that he retains those 10 fewer wins in Cooper's adjusted record, but I don’t think I would.

Chris J’s adjusted W/L record for Cooper is 217-177. One approach (I'll call it the "statistical approach") would just remove the effects these random errors. Doing this would improve Cooper's adjusted record by 10 games to 227-167. By comparison, according to Chris, Clark Griffith won 7 more games than would have been expected. Adjusting his record for those 7 extra wins would bring his adjusted record down to 224-159. If one then adjusts for the fact that at least two of Griffith's seasons were in very weak leagues (1891 AA and 1901 AL), Cooper and Griffith start to look very comparable. I mention this because Griffith did very well in the last ballot (4th overall, 2nd among pitchers), while most of the comments I've seen here suggest that Cooper may receive very few votes.

One could take another approach to adjusting for the difference, which I’ll call the “accounting approach.” This approach says that those 10 fewer wins have to be attributed to someone. That's how I interpret what Bill James does in win shares – he avoids having any residuals floating around by assigning someone responsibility for every win. But if you take this approach, why should the blame for having fewer wins than expected go entirely to the pitcher? Doesn’t it seem just as likely that maybe the offense didn’t come through with runs at the right time, or the defense messed up with the game on the line? Suppose that we assign 1/3 of the responsibility to the pitcher, half to the offense, and 1/6 to the defense. Then Cooper would retain responsibility for 3 wins below expectations and we would add the other 7 wins to Cooper's adjusted record, giving him an adjusted record of 224-170. And we would subtract 5 wins from Griffith’s adjusted record, making it 226-157. Griffith looks better in this comparison, but the two are still very close.

Griffith was on my ballot last time, and looking over Cooper's complete record, he looks fairly comparable to Griffith except for the W-L records. My interpretation of Chris's analysis is that the difference in their W-L records results from a combination of run support and luck, so Cooper will probably make it onto my ballot.
   228. Howie Menckel Posted: August 08, 2004 at 01:12 AM (#784049)
A puzzle of sorts, if anyone's game:
I have all HOMers per year from 1871 on. Can we do that as well (less confidently of course) for earlier years?
My basic premise has been listing all those who appeared in a 'major league' game that year. If fewer than 10 G, they are listed only incidentally. If fewer than half the games, with their limited games listed (like 62/140 games, or whatever).
I'll show 1875 and 1876 as examples..


1875 - White BOS, Hines CHI, Barnes BOS, O'Rourke BOS, Wright BOS, Anson PHI, Spalding BOS, Sutton PHI, Start NY, McVey BOS, Pearce STL (Galvin STL 8-58)

1876 - White CHI, Hines CHI, Barnes CHI, O'Rourke BOS, Wright BOS, Anson CHI, Spalding CHI, Sutton PHI, Start NY, McVey CHI, Pearce STL (25/64)
   229. DavidFoss Posted: August 08, 2004 at 02:36 AM (#784139)
Good question Howie, I was used to splitting players up in terms of major/minor teams ... Sutton for Rochester Alert in 1869 and McVey for Indianapolis Active in 1868. I think those are more cases of being in the "minor leagues" than having partial playing time.
   230. DavidFoss Posted: August 08, 2004 at 02:40 AM (#784146)
1857: 1 (Pearce)
1858: 1 (Pearce)
1859: 1 (Pearce)
1860: 2 (Pearce, Start)
1861: 2 (Pearce, Start)
1862: 2 (Pearce, Start)
1863: 2 (Pearce, Start)
1864: 3 (Pearce, Start, GWright)
1865: 2 (Pearce, Start)
1866: 3 (Pearce, Start, GWright)
1867: 3 (Pearce, Start, GWright) -- minor: (Barnes, Spalding)
1868: 5 (Barnes, Pearce, Spalding, Start, GWright) - minor: (McVey, JWhite)
1869: 7 (Barnes, McVey, Pearce, Spalding, Start, JWhite, GWright) -- minor: (Sutton)
1870: 8 (Barnes, McVey, Pearce, Spalding, Start, Sutton, JWhite, GWright)


Notes:
-- As before, perhaps omitting the "minors" seasons is the best bet, they are listed here for completeness and if someone disputes a "minor" label they can be moved into the major list.
-- GWright team-hopped in 1866, spending partial seasons for both NYGotham and Union Morrisania. I'm giving him a full season for that.
-- Most sources list Pearce as playing for Brooklyn Atlantic in 1856. The Marshall Wright book only goes back to 1857 (beginning of the 9-inning game).
   231. DavidFoss Posted: August 08, 2004 at 02:44 AM (#784149)
1857-1870 HOF Teammates:

Start/Pearce -- 1862-1870 Brooklyn Atlantics
Barnes/Spalding -- 1868-1870 Rockford Forest City (1867 minor)
McVey/GWright -- 1869-1870 Cincinnati Red Stockings
Sutton/White -- 1870 Cleveland Forest City
   232. Michael Bass Posted: August 08, 2004 at 03:10 PM (#784363)
Complete veering off topic here, but an, I'm sure, amazingly controversial topic that has caught my eye after reading the Maddux 300 win thread, which had a sidebar about Pedro and Koufax.

What player(s), generally viewed as lock HOFer/HOMer by most baseball fans, do you see as not as worthy?

My current list of two:

- Lou Brock. By WARP3, slightly worse than Harry Hooper. By WARP1, taking out the league quality/timelining effects, a significantly worse player than Harry Hooper. Not the kind of decription likely to get one high on my ballot.

- Sandy Koufax. Is he even better than Rube Waddell? Even taking timelining into account, I like Rube a little better, both peak and career, closer on the latter, surprisingly enough.

That's not to say I won't vote for either of these guys at all; I have no idea what the ballot depth will look like by that time. But to me they are not the slam dunkers their images project.

So who are everyone else's contrarian dislikes?
   233. John (You Can Call Me Grandma) Murphy Posted: August 08, 2004 at 04:17 PM (#784443)
So who are everyone else's contrarian dislikes?

I don't know if he's a truly contrary pick, but Don Sutton would be high with Brock on my list.

Koufax has a great peak argument, so the writers were compelled to extrapolate the rest of his career for him so he would easily be elected. I would certainly not argue against his selection, but the idea that he was the greatest pitcher of all-time by some people is just plain silly.
   234. karlmagnus Posted: August 08, 2004 at 07:17 PM (#784842)
I've moved pitching up a bit in geeneral, since we fairly clearly don't have enough of it. waddell, Willis and Foster now make my ballot. So, between Waddell and Willis, does Sam Leever, who I think the pitching experts need to look at more closely -- he didn't make the majors till 27, presumably being blocked by the 90s league, but had a superb W/L and a very decent ERA+ once he got there. He also slots into a least the latter end of the 90s gap, and was a key part of that Pirates juggernaut.
   235. Kelly in SD Posted: August 08, 2004 at 09:10 PM (#785057)
I posted a career sketch of McGinnity on the Pitchers Thread.
   236. Kelly in SD Posted: August 08, 2004 at 09:55 PM (#785113)
Two factors to consider with Leever are his Run and Defensive Support as figured by Chris J.:
Run Support 112.92 which is 7th best of all pitchers whose career ended before 1930.
Defensive Support is +14 which is tied for 16th with McGinnity for all pitchers whose career ended before 1930.
The Pirates were one of the three dominant teams of the National League during the 1900s and their defense must have been comparable to the Cubs as Phillippe had Defensive Support of 13.8, Babe Adams was 12.8, and Willis was 8.8. With Wagner and Leach on the left side of the infield, Pirates' pitchers had a great advantage.

Leever was a good pitcher and I don't want to tear him down. The Pirates were a great team for a decade and Leever was an integral part of it.

I will try to get a Leever profile up soon along with Waddell and Brown. That should allow some comparing amongst the "short" career HoM pitchers with other "short" career candidates.
   237. Yoenis Cespedes, Baseball Savant Posted: August 09, 2004 at 12:23 AM (#785260)
A surprisingly strong rookie class of players throws my ballot up in the air. The best class of Negro Leaguers we've had, plus a couple of pitchers who surprised me. Five players from the Class of 1932 make my ballot. Five Negro Leaguers make my ballot. The positional breakdown is as follows: 2 C, 1 1B, 0 2B, 2 SS, 2 3B, 0 LF, 3 CF, 0 RF, 5 P.

1. Louis Santop. A catcher who could mash. The closest thing to a no-brainer on this ballot.
2. Rube Foster. Big tips of the hat to Chris Cobb and Chris Jaffe for their work on Foster. We don't know how dominant he was, but given why we don't know any better, I'm inclined to place him here. Gets a boost from his hitting ability.
3. Rube Waddell. An exemplar of the dominant frontline starter who could push his teams to a pennant. Seems to be a forgotten man in the HOM discussions, which is a real shame.
4. Roger Bresnahan. Played catcher for about 1,000 games and was an on-base machine. To quote the 2001 Baseball Prospectus, "two great tastes that taste great together." Gets a boost after I reworked my catcher rankings.
5. Frank Chance. I was surprised when I saw how highly Chance ranked in my system. Sure, he was injured a lot, but he was a dominant player in all facets of the game while managing and acquiring the personnel that steamrolled the National League in the middle of the 1900s. Would rank higher, but is placed here because of a completely arbitrary rule about player-managers placed into effect on the basis of an ill-defined "consensus."

6. Dobie Moore. The comparisons in the research I've read place Moore is similar to Jennings: historically high peak for his position, short career. I'm giving him the benefit of the doubt over Hughie, so here he is.
7. Hughie Jennings. Much like the climb up Alpe d'Huez, Jennings' peak is hors category. Stood tall as the best player in the only game in town.
8. Jose Mendez. I see Mendez as a high-peak David Cone. My interpretation of his career compares very well to Adams, but there is more ink written about the greatness of Mendez, so he gets the edge.
9. Pete Browning. The guy could rake -- he's absolutely the best hitter eligible. I ignore the AA discount given that he was playing with an inner-ear ailment that drove him to alcoholism, insanity and an early grave.
10. Babe Adams. Ranks higher than this in my system, but I can't quite believe that he is better than some of the players listed above him. The 117 ERA+ in about 3000 innings doesn't excite at first glance, but he pitched about 400 horrible innings, distorting his rate stats. In reality, he had an excellent peak, with several high-impact seasons.

11. Spotswood Poles. Fights his way onto the ballot because I give him full credit for the 1918 season, when he was fighting in World War I. Compares well with Hugh Duffy, though I see Poles as slightly better at the plate and on the basepaths while Duffy is superior in the field.
12. Lip Pike. Great player. Hit for a high average, fast, had light-tower power. But I just can't seem to pull the trigger and put him higher. Slips on a ballot surprisingly crowded with new eligibles.
13. Wilbur Cooper. The transition we are making from evaluating 19th-century pitchers to looking at this century is very interesting, indeed. The only pitchers we've elected with "modern-looking" careers are Brown and Plank. The other 20th-century guys, while effective, made their names by being workhorses. Pitchers with 3500 innings and a 116 ERA+ don't exactly grow on trees.
14. Tommy Leach. Underrated, but hitting slightly above league average and playing Gold Glove-caliber defense for about 15 years has a way of sneaking up on you.
15. John McGraw. Speaking of on-base machines, he's the biggest beneficiary of my ranking system. I want someone who I know was great for a relatively short period of time over someone who was pretty good for a while and filled a space the rest of the time.

Players falling off of ballot (1931 ballot rank)
(12) Larry Doyle. The ballplaying ancestor of Jeff Kent. The best eligible second baseman in National League history to this point.

(13) Clark Griffith. Not much to say about him. Great peak from 1894-1901, spanning most of the one-league era.

(14) Hugh Duffy. Spotswood Poles' white twin; Duffy's career stats are similar to my intrepretation of Poles' i9s stats. The AA season, Poles' 1918 season and my policy of giving Negro Leaguers the benefit of the doubt in close cases is what places them apart on a very tight ballot.

(15) Gavy Cravath. My favorable interpretation of his minor-league career would have him fighting for one of the top five to seven spots on my ballot, but I can't trust that projection with enough confidence to do so. That puts him here.
   238. Yoenis Cespedes, Baseball Savant Posted: August 09, 2004 at 12:25 AM (#785262)
Other consensus Top 10 players left off of my ballot
Jake Beckley. I feel very strongly that the electorate will be making a grave mistake if we elect Beckley, who was the Harold Baines of his time. He is credited as being the best first basemen outside of the Anson-Brouthers-Connor troika, but what kind of accomplishment is that, really? He was often the second or third-best position player on mostly mediocre teams:

Season W/L records for teams where Beckley was one of the three best position players, based on Win Shares:

66-68
61-71
60-68
55-80
65-65
71-61
76-56
83-67
62-77
52-87
70-70
75-79
58-96

854-945---.475

Season W/L records for teams where Beckley was not one of the three best position players:

80-73
81-48
65-65* (average of NYG and PIT records in 1896; Beckley had 5 WS for each team)
92-60
74-65

392-311---.558

Only three of Beckley's teams enjoyed winning records when he was one of their three best position players. Two more finished at .500. Both times that Beckley was the best position player on his team, the team finished with a losing record.

On the other hand, four of the five teams where Beckley was not among the top three position players had winning records -- the other team finished at .500 (see note above).

There were 12 seasons in which Beckley was one of the five best players on his team, pitchers included. Only two of those teams finished with winning records, the 1895 Pittsburgh Pirates, who finished 7th with a 71-61 record in the 12-team National League and the 1899 Cincinnati Reds, who finished 6th with an 83-67 record in the 12-team NL.

(Incidentially, both seasons ended up with exceptionally bad doormats in the National League. The 1895 season had the Washington Senators, St. Louis Browns and Louisville Colonels far down at the bottom of the standings:

WSH---43-85---.336
STL---39-92---.298
LOU---35-96---.267

Those three teams alone were bad enough to allow the other nine to finish with a winning record, something which almost happened again in 1899. This time, it was the historically bad Cleveland Spiders that pushed most of the teams above .500 with a truly dreadful 20-134 (!) record, good for a .130 "winning" percentage. The Spiders finished 84 games out of first place and 35 games out of 11th place. Eight teams finished above .500 and the Washington Senators fell just two games short of that break-even mark.)

I would argue that every one of of HOM selections thus far, with the possible exceptions of Bid McPhee and Pud Galvin, could be the best player for a pennant-winning team. Jake Beckley doesn't come close to meeting that standard.

George Van Haltren. Very good player for a long time, but gets pipped at the post by players who had higher peaks.

Jimmy Ryan. His counting stats are buoyed because he was mostly durable and mediocre for the last decade of his career, with the exception of 1898.

Cupid Childs. Good player, but I don't see him making my ballot. I've decided he ranks below Doyle, and the best Doyle can hope for in a good year is to be in the bottom third of my ballot.
   239. John (You Can Call Me Grandma) Murphy Posted: August 09, 2004 at 12:46 AM (#785286)
Would rank higher, but is placed here because of a completely arbitrary rule about player-managers placed into effect on the basis of an ill-defined "consensus."

Actually, this was a majority rule thing before the project was created. Joe (admittedly on his part) just didn't clarify it enough in the Constitution. If this is defined as arbitrary, then the whole Constitution is arbitrary.
   240. OCF Posted: August 09, 2004 at 01:11 AM (#785339)
while managing and acquiring the personnel that steamrolled the National League in the middle of the 1900s.

It was Frank Selee's team. Chance deserves tremendous credit for winning with the team he was handed, and for finding the spare parts needed to keep it working at his peak, but it was Selee who put together the core of that time and established its style, including the distinctive spread-the-load handling of the pitching staff. After this team grew old together, Chance never put together another one like it (a key distinction between him and McGraw.)

Don't despair: John Murphy and I both have been putting Chance on our ballots - but for his playing value only.

By the way, my preliminary ballot (#220) has a flaw: I overlooked the eligibility of Jose Mendez. I'll fix that before I post it as a ballot, and Mendez will be on it somewhere.
   241. karlmagnus Posted: August 09, 2004 at 01:22 AM (#785359)
If we do a managers' wing I'd put Selee very high on my ballot for it, because he did the Beaneaters and then put together Chance's Cubs. Very few people put together 2 dynasties at that level.
   242. Yoenis Cespedes, Baseball Savant Posted: August 09, 2004 at 01:22 AM (#785360)
John,

I apologize for the harsh tone of language in the Chance comment.

I do think that there is some limited value in being a successful player-manager. My best guess is that Chance's managing was worth 1-3 wins per year to his team at its peak, a small but significant impact. To say that there is no usable value or positive impact from player-managing because we can't measure it clearly promotes ignorance on the topic.
   243. DavidFoss Posted: August 09, 2004 at 01:49 AM (#785416)
(a key distinction between him and McGraw.)


An aside here... but McGraw's first job was as the lesser end of a syndicate team. Hanlon's Brooklyn team got most of the best players in the syndicate and he still managed the team to a fine 86-62 record. Very impressive, in my opinion.
   244. Chris Cobb Posted: August 09, 2004 at 03:46 AM (#785545)
1932 Preliminary Ballot

With so many new eligibles of interest, I've commented on my placement of them; the returning candidates are listed for context. I've made some small re-arrangements among returning candidates.

1. Louis Santop – Best catcher we’ve seen since Buck Ewing. Slightly better offensively and defensively than Roger Bresnahan, and a full-time starter, at outfield and at catcher for significantly longer. With no competition adjustment for any post-1875 play, I have him as the #31 player so far, right between Buck Ewing and Frank Baker.
2. George Van Haltren
3. Clark Griffith
4. Mickey Welch
5. Rube Foster
6. Hughie Jennings
7. Spotswood Poles
8. Lip Pike
9. Jose Mendez – Excellent seven-year prime and occasional fine work thereafter place him in the middle of the ballot. Trails Walter Johnson, Pete Alexander, and Smokey Joe Williams by a mile, but possibly the next best pitcher of the teens. Not certain he’s a HoMer, but not certain he isn’t, either.
10. Tommy Leach
11. Larry Doyle
12. Hugh Duffy
13. Gavvy Cravath
14. Roger Bresnahan
15. Wilbur Cooper – Peak not as high as that of Mendez, but a very fine pitcher for eight years running, 1916-1923. As the first pitcher whose peak extended significantly into the lively ball era, he’s a bit hard to place, since we haven’t really examined exactly what the impact of the lively ball on pitching value is. May move substantially up or down as we study contemporaries like Coveleski, Mays, Shocker, and Luque, and the significantly longer-careered Faber and Rixey. His career shape is rather like Griffith’s, and he may take a while to gain traction with the electorate for the same reason that Griffith did – lack of a dominant peak. But if other pitcher candidates keep falling short of him (I don’t know whether they will or no), he’ll rise as Griffith has.
16. Jimmy Ryan
17. George J. Burns
18. Cupid Childs
19. Bill Monroe
20. Dobie Moore – Like Cooper, a hard player to rank. Since most of his true contemporaries won’t be eligible for a while, it’s hard to say exactly how good he was relative to his peers. I see Frisch, Cronin, and Sisler at their peaks as good comps for Moore as a hitter, though he was slightly better than Frisch and Cronin, not quite as good as Sisler. But his career, even with credit for his army play, was short. Frisch and Cronin will be easy choices when they become eligible because they have both peak and career. Sisler will be closer to the borderline because he tailed off after his eye problem; Moore has less career than Sisler. If we elect a Negro-League shortstop from the 1920s, it should probably be Dick Lundy.
21. Harry Hooper
22. Rube Waddell
23. Herman Long
24. Tommy Bond
25. Bobby Veach
26. Charley Jones
27. Bruce Petway
28. Fielder Jones
29. Babe Adams – A Career full of ups and downs. Seldom pitched two outstanding seasons back to back, so his durability seems to have been a bit lacking. But he was a highly effective pitcher when he was healthy. Low walk rate is well-known, but also appears to have been very good at suppressing hits on balls in play. Trails Mendez, whose career has a somewhat similar shape, by a substantial amount because a) he pitched a bit less b) his per-inning peak rate was a little lower, and c) most importantly, he put together fewer great seasons. I’m not giving Adams credit for minor-league pitching. I can see giving him credit for 1908 because it does look like he was good enough to pitch in the majors. I’d like to know more about the relations between the Pittsburgh club and the American Association Louisville club. Did Adams have a real chance to have his contract purchased by any major-league team, or did Louisville have an understanding with Pittsburgh? Did the fact that Pittsburgh had bought his contract at one point mean that they retained major-league rights to him? If he was seriously blocked by the strong Pittsburgh staff, I might give him that year.
30. Jake Beckley
31. Frank Chance
32. Tony Mullane
33. Dick McBride
34. Ed Konetchy
35. Lave Cross
36. John McGraw
37. Joe Tinker
38. Johnny Evers
39. Ed Williamson
40. Addie Joss

Other New Eligibles

appx 82. Stuffy McInnis – all-star first basemen during the Athletics glory years, nothing special after that. A larger percentage of his value was in his batting average than for any player we’ve seen recently. Well behind both Konetchy and Daubert among 1910s first-baseman.

appx 100. Hooks Dauss – An amazingly average pitcher! Never varied more than 2.5 wins above or below average for his career. Was 1.1 wins above average for his whole career. A pitcher you’d love to have on your team, but unlikely to be the top starter on a pennant-winner. Did suffer from poor defensive support – his Tigers were below average in terms of defensive efficiency 10 out of his 15 seasons (a strong argument against enshrining Donie Bush) but these were just below average defensive teams, not awful ones, so they cost him only about 3 wins.

Dropping out of top 40

Tom York (37)
Eddie Cicotte (38)
Jim McCormick (39)
Mike Tiernan (40)

These players might reappear near the ballot some time down the road, but probably not.

Other Players Deserving of Comment

Sam Leever was mentioned. Won't make my top 40 -- no way he was as good as his contemporary Addie Joss, and Joss is at 40 now.

Hippo Vaughn has been mentioned. I've realized I haven't given him a serious look, and he deserves one. Without having studied him, I'm guessing he'll make the top 40 but not the ballot. But I didn't think Wilbur Cooper would make my ballot and he did, so I'm not ruling out the possibility that he could turn out to be better than Jose Mendez. . . It's more likely that he could be better than Babe Adams, and land in the mid-twenties.
   245. John (You Can Call Me Grandma) Murphy Posted: August 09, 2004 at 02:22 PM (#785779)
Don't despair: John Murphy and I both have been putting Chance on our ballots - but for his playing value only.

Unfortunately for the "Peerless Leader," he won't be on my ballot in a couple of years (and probably for good).

I apologize for the harsh tone of language in the Chance comment.

Don't be. I wasn't offended by it.

I do think that there is some limited value in being a successful player-manager. My best guess is that Chance's managing was worth 1-3 wins per year to his team at its peak, a small but significant impact. To say that there is no usable value or positive impact from player-managing because we can't measure it clearly promotes ignorance on the topic.

I don't think anyone is saying that Chance's managing wasn't worth anything, only that it shouldn't be included with his playing for this branch of the HoM. When the manager's wing is created someday, I assume he'll be a much discussed player there.

I honestly don't think there is some type of extra value created by player/managing (as opposed to full-time managing). The idea that a player's on-field play may suffer from player-managing may have some value, though.
   246. PhillyBooster Posted: August 09, 2004 at 02:34 PM (#785788)
DanG wrote:

Another thought takes me back to revisiting some very early ideas in this project. The old question of When did pitching become valuable?

This issue does seem to be a recurring theme, but I'm not exactly sure where the leap comes that because pitching was "different" before 1893 (to various degrees) it was less valuable. Part of the problem for very early years (1870s, say) is that one pitcher pitched most of the innings for a team, so it is difficult to compare directly to teammates. In the 1880s, however, this is not a problem, as most teams have at least 2 major pitchers. For Welch, we are not much helped by a comparison to Keefe, but we can look at other teams in his era to determine whether pitching mattered.

Picking 1885 as representative of the "heart" of Welch's career, here is the ratios of the best and worst pitchers' RA (min. 100 IP) for each team (from most divergent to closest):

Detroit: 43% (Lady Baldwin over Charlie Getzein)
Buffalo: 25% (Pud Galvin over Pete Wood)
St. Louis: 25% (John Kirby over Charlie Sweeney)
Boston: 18% (Charlie Buffinton over Jim Whitney)
Chicago: 17% (John Clarkson over Jim McCormick)
Providence: 12% (Charlie Radbourn over Dupree Shaw)
New York: 11% (Mickey Welch over Tim Keefe)
Philadelphia: 1% (Ed Daily over Charlie Ferguson)

The bottom five teams on the list used two regular pitchers, and the top three used 3 or 4 regular pitchers. The top 3 also finished in sixth, seventh, and eighth place.

I don't see how one can look at these spreads of RA between teammates and not see that pitching quality was a crucial variable.
   247. John (You Can Call Me Grandma) Murphy Posted: August 09, 2004 at 02:40 PM (#785796)
This issue does seem to be a recurring theme, but I'm not exactly sure where the leap comes that because pitching was "different" before 1893 (to various degrees) it was less valuable.

I don't think it was less valuable, Matt, but I also don't think it was more valuable either.
   248. PhillyBooster Posted: August 09, 2004 at 02:50 PM (#785810)
I agree that pitching itself wasn't more valuable. But I think that there were more individual pitchers who were more valuable.

There is general agreeement, I believe, that Welch was the 5th or 6th best pitcher of the 1880s. The issue becomes whether there should be a sixth 1880s pitcher inducted.

If pitching was as important then as it is now, then I think it's very easy to come to the conclusion that the 1880s were a decade of great pitchers, and six should be inducted. I think to get from Welch's statistics to a conclusion that he is not ballot-worthy requires a conclusion about the importance of pitching overall.
   249. John (You Can Call Me Grandma) Murphy Posted: August 09, 2004 at 03:04 PM (#785820)
I agree that pitching itself wasn't more valuable. But I think that there were more individual pitchers who were more valuable.

Well, I do think there were more better pitchers during the '80s than the '90s, for example, so I probably lean toward your side of the argument. I'm still not convinced that Welch is a HoMer though (but he will be on my ballot one last time before he disappears forever when all the great newbies start popping up shortly).
   250. DanG Posted: August 09, 2004 at 03:15 PM (#785827)
This issue does seem to be a recurring theme, but I'm not exactly sure where the leap comes that because pitching was "different" before 1893 (to various degrees) it was less valuable.

I think it mainly stems from analyzing Wins Shares. In his formulas, James did not adjust for any major difference in pre-1893 pitching, so all of the greatest seasons in history were by pitchers from that era, if you take WS literally. Also, adjusting WS for season length makes many more pitchers no-brainers for the Hall. Not only Welch, McCormick, Mullane; also Buffinton, White and Whitney.

James himself says those numbers should be discounted. The $64,000 question is by how much? Attempts made here to adjust pre-1893 pitcher's WS has the era producing about as many great pitchers as any other. The question is, was this outcome a pre-ordained goal of the adjustments or were they the results of careful, defensible modifications?

I must leave it to those who crunched the numbers to answer that last question.

As to your study, how does 1885 NL compare to the AA or the surrounding seasons?
   251. karlmagnus Posted: August 09, 2004 at 03:16 PM (#785828)
He won't disappear forever, he's like Beckley. In 50 "years" when there's a weak ballot, and Welch/Beckley have hugely more wins or hits than anyone else on it, he'll have a good chance for induction. The same wouldn't have been true for Pearce, Caruthers or Thompson, and won't be true for Van Haltren, Pike or Browning.
   252. John (You Can Call Me Grandma) Murphy Posted: August 09, 2004 at 03:22 PM (#785836)
He won't disappear forever, he's like Beckley. In 50 "years" when there's a weak ballot, and Welch/Beckley have hugely more wins or hits than anyone else on it, he'll have a good chance for induction. The same wouldn't have been true for Pearce, Caruthers or Thompson, and won't be true for Van Haltren, Pike or Browning.

I keep thinking that there will be an abundance of new quality candidates out there, but there probably will be some slow periods ahead. I haven't done any forecasting models myself, so you may be right, karlmagnus.
   253. PhillyBooster Posted: August 09, 2004 at 03:50 PM (#785880)
I think it's not so much that there will be more "dead periods" like 1924-1932, as there will be more stretches with only one great new candidate per year during periods where we are elected 2 or 3 people.

A list of new eligibles that reads:

WS W3 Rookie Name-Pos (Died)
314 73.5 1914 Edd Roush-CF (1988)
245 68.3 1913 Wally Schang-C (1965)
177 45.3 1918 Bill Sherdel-P (1968)

takes on a very different character depending on whether it is an Elect One (add one to the backlog), and Elect Two (maybe both go in now), or an Elect Three year (take one from the backlog).
   254. karlmagnus Posted: August 09, 2004 at 04:02 PM (#785894)
Quite so, Phillybooster. The problem for backers of golden oldies, 50 years dead, will be to keep them afloat waiting for the weak year in the vast sea of new plausible candidates that will flood over us each year. Because of Beckley and Welch's hard numbers, that problem will be easier in their case.
   255. DavidFoss Posted: August 09, 2004 at 04:10 PM (#785898)
Because of Beckley and Welch's hard numbers, that problem will be easier in their case.

Perhaps true. Though I believe it is uniqueness and not hard numbers that will hold a player high in the backlog for "generations". Pearce and Caruthers had this unique character to their career, but I agree that Thompson did not and Pike, Van Haltren and probably Browning will not.

Being part of a large group of similar candidates will cause a player to be part of some sort of "glut", players in these gluts will have a hard time gathering support... especially as the glut gets bigger. Players in the gluts seem to split each others votes.
   256. karlmagnus Posted: August 09, 2004 at 04:17 PM (#785908)
David, I agree entirely. But there can't be a glut of candidates with 300 wins or almost 3000 hits. It's the next statistical level down where players will get lost in the tidal wave.
   257. Chris Cobb Posted: August 09, 2004 at 04:26 PM (#785918)
In other words, this project isn't going to get any "easier" as we move towards the present :-/ .

One factor that will help early candidates is that the electorate will have developed a fair degree of confidence in those players' value, and a certain commitment to seeing it honored eventually.

I can't predict who will be elected, of course, but I'm sure that the group of Rube Foster, George Van Haltren, Clark Griffith, Lip Pike, Jake Beckley, Jimmy Ryan, Hughie Jennings, Roger Bresnahan, Cupid Childs, and Mickey Welch will not disappear from serious consideration, regardless of who from that gropu achieves election in 1932 or 1933. They're going to drop precipitously in 1934, but some of them will be back in the top 5 by 1938.

I think Phillybooster's analysis of how these players might be elected down the road is plausible. I do think that we're going to have the sense that the ballot is very crowded and very strong for a long time to come, probably until we reach 1960. My sketchy looks ahead show the 1930s as indeed something of a golden age. Lots of great players, and no way to explain away their greatness due to declining quality of competition. We're going to have some tough choices to make down the road.
   258. PhillyBooster Posted: August 09, 2004 at 04:29 PM (#785923)
As to your study, how does 1885 NL compare to the AA or the surrounding seasons?

Here's 1886 NL. The differences appears more pronounced between best and worst, but here only Detroit and New York are using primarily 2 pitchers instead of 3 or 4:

Washington: 68% (Shaw over Madigan)
Philadelphia: 54% (Ferguson over Daily)
Boston: 45% (Radbourn over Buffinton)
Kansas City: 38% (Whitney over Conway)
St. Louis: 35% (Boyle over Kirby)
Detroit: 32% (Baldwin over Geitzen)
New York: 19% (Keefe over Welch)
Chicago: 12% (McCormick over Clarkson)

As to Win Shares's treatment of 19th century pitchers, I am agnostic. It does not seem intuitively implausible to me that the game was entirely dominated by pitchers, or that all pitchers were interchangeable. The numbers I have looked at seem to militate against the "interchangeable" thesis, but doesn't necessarily take us all the way to the other extreme.

If he's still on the ballot in 1995, a side-by-side with Tommy John would be very interesting, as the two have very career similar stats. Tommy John would make my ballot, but Welch, I think would be ahead.
   259. DavidFoss Posted: August 09, 2004 at 04:41 PM (#785950)
David, I agree entirely. But there can't be a glut of candidates with 300 wins or almost 3000 hits. It's the next statistical level down where players will get lost in the tidal wave.


Very true, Beckley and Welch also have "unique" candidacies that will most likely hold them high in the backlog if they manage to escape election.

Sam Rice will be an interesting candidate at 2987 hits, but I would rate Beckley ahead of Rice. Getting a few years ahead of myself here.
   260. DanG Posted: August 09, 2004 at 05:04 PM (#785985)
I'm wondering if the data shown by PhillyBooster might support the theory that 1880's pitchers were not dominant, but instead were very dependent on their team's defense.

Let's start by asking, if the difference between the good and bad teams was mainly due to defense, what would we expect to see?

Well, we would expect to see smaller difference between the peformance of the "ace" pitcher and other pitchers, right? The team has a good record because, regardless of who is pitching, they make outs. Most of their success is attributable to the good defense and not their pitching. PB's study may show this.

The ace of a bad defensive team would be more on his own to get guys out and the bad pitcher is more dependent on his defense to make outs. The differences between the good pitcher and bad would be magnified by the weak defense. PB's study may show this.

Another way to interpret the data is to say that the best teams are the ones that could find two similar quality pitchers. The bad teams tended to have only one starter while struggling to find another. OK, it says not just anyone could pitch in the 1880's. But this points away from the value of dominance and toward competence.
   261. Chris Cobb Posted: August 09, 2004 at 05:41 PM (#786084)
Following up on DanG's inference, let's see how Phillybooster's data correlates with WS ratings of team fielding (WS may give too much credit to pitchers, but I think it is fairly reliable at distinguishing good-fielding teams from bad-fielding ones).

1885

Team -- field ws -- Phil Ratio
Chicago -- 36.6 -- 17%
New York -- 36.6 -- 11%
Philadelphia -- 34.2 --1%
Providence -- 28.3 -- 12%
St. Louis -- 23.5 -- 25%
Boston -- 21.0 -- 18%
Detroit -- 19.1 -- 43%
Buffalo -- 14.8 -- 25%

Avg. of top 4 -- 10.25%
Avg. of bottom 4 -- 27.75%

1886

Chicago -- 40.8 -- 12%
Detroit -- 40.8 -- 32%
Philadelphia -- 38.5 -- 54%
New York -- 32.1 -- 19%
St. Louis -- 24.9 -- 35%
Boston -- 24.1 -- 45%
Kansas City -- 20.6 -- 38%
Washington -- 18.5 -- 68%

Avg. of Top 4 -- 29.25%
Avg. of Bottom 4 -- 46.5%

The data aren't conclusive, but there does seem to be some correlation. However, there's not enough correlation to say that the difference is mainly due to defense, only that the difference between the ratio of best RA+ to worst RA+ is influenced by the quality of fielding.

Another factor that would need to be accounted for in analyzing this data is pitcher usage patterns. We know from other studies that at this time pitcher usage was not rotation-based but was heavily influenced by team choices about match-ups. Middling or poor teams have a strategic choice to make in usage. They can run out their best pitcher against their best opponents in hopes of winning as many of those games as possible while hoping their other pitchers can be adequate against lesser competition, or they can "concede" games against their best opponents and save their best pitcher for games that the team might actually win. The former pattern would compress RA+ ratios, the latter would expand them. I think we'd have to look at who the pitchers pitched against before we could reach firm conclusions about the meaning of this data.
   262. Dr. Chaleeko Posted: August 09, 2004 at 06:01 PM (#786139)
I think _[pick your favorite 1932 backlogger]_'s best chances will either be 75-76 or else in the late 80s or 90s when we elect three/four candidates at a time. Of course they will then be battling other second-tier career-value guys like Pinson, Dawson, Parker, Lance Parrish, Tommy John, as well as the less careerish Dale Murphys, Jim Rices, and ace relievers like Rollie Fingers and Goose Gossage.

Even though there are soft-spot years before the mid 70s (37, 38, 46, 60, 67, 68), it seems like the backlogs will be pretty tough to break through. The 1934 class is something like 8 deep in serious candidates, and it's not going to dissipate too too much in four let alone ten years because Alex and Heilmann hit the ballot in 36, serious Negro League candidates like DeMoss, Marcelle, and Redding come on from 1935-1938, and post-1932 backloggers or well-known players like Carey, Roush, Wheat, Groh, Schang should all be lurking as well.

It's a rough time to be a HOM backlog candidate!
   263. OCF Posted: August 09, 2004 at 06:33 PM (#786217)
My gut response to the discussion about backloggers is this: The people we have to make sure to give a fair shake to are marginal candidates, arguably as good as some that have been on the ballot forever, who enter the ballot together with several shoo-ins. All of us are going to fill the top end of the '34 ballot with new players, shoving everyone else down. Those players from the top of the '34 ballot will all get elected in a few years, but that doesn't mean that the bottoms of our ballots should necessarily come back to just where they were - there might be some players first eligible in flush times who should be spliced into those lists. Stan Coveleski, for instance, or Wally Schang.
   264. Chris Cobb Posted: August 09, 2004 at 06:44 PM (#786252)
My gut response to the discussion about backloggers is this: The people we have to make sure to give a fair shake to are marginal candidates, arguably as good as some that have been on the ballot forever, who enter the ballot together with several shoo-ins.

That's good; both the marginal candidates and the backloggers will need looking after. The special challenge with the marginal candidates will be to keep up with them and not let anybody slip through the cracks, as Wilbur Cooper might and Hippo Vaughn might have, for a while at least. If we look after both groups assiduously, we should end up about right.
   265. Kelly in SD Posted: August 09, 2004 at 08:28 PM (#786518)
Rube Waddell:
Teams: Lou 1897, 1899, Pit 1900-01, ChiN 1901, PhiA 1902-07, StLA 1908-10.
Record: 193-143 2.16 era/3.23 runsallowed, K/W 2.88, WH9IP=9.92
Win Shares: Career 240; 3 yrs cons 94; 7 best yrs 186; per 40 starts 25. Seasons with 20+: 6. Seasons with 30+: 3.
AllStars: STATS 3, WS 4
Fibonacci WinPoints: 161
ERA+: 134
Black Ink/Grey Ink: 46/158
Bill James Rank: 53
Top 10s: Ks and K/9IP 10 times with 6 first in Ks and 8 in K/9IP. SHo 9 times. H/9IP 8 times with 2 firsts. aERA+ 7 times with 3 firsts. ERA and BBH/9IP 6 times each with ERA 2 firsts. Wins 5 times.
Other/Unique: Usage Pattern almost every year, usually b/c he jumped the team or was suspended: 1901 last start on 8/28.
1902 - played for LA in PCL and didn't start for Phi until 6/26 and had his last start on 9/22, he had 27 starts and 6 relief appearances in 81 games. No starts after game 131.
1903 - 38 starts in game 106, none after.
1905 - didn't start until game 15 and between games 120 and 148.
1906 - only one start in a month between 5/21 - 6/19.

Career Breakdowns:
Against HoMers:
Young      8-5
McGinnity  0-0 with a tie
Mathewson  1-1
Plank      1-3
Walsh      3-6
Johnson    2-0
total:    15-15
Also: 
Griffith   4-4 
Joss       7-2
Cicotte    0-1  

Record depending on opponent's finish:
finish   Waddell
1st      12-25
         .324
2nd      12-20
         .375
3rd      34-21
         .618
4th      23-18
         .561
5th      18-25
         .419
6th      24-13
         .649
7th      30-14
         .682
8th      35-10
         .778
9-12      1-0
        1.000  

Records against over/under .500 teams
Record  Waddell
.500+    88-93
         .486
.500-   101-54
         .649   

% of career dec'n vs. .500+teams
.537
   266. jimd Posted: August 09, 2004 at 08:52 PM (#786571)
I'm not sure what is meant in this discussion by pitching "dominance". But if what we're talking about has any correlation with strikeout rates, then pitchers in the early 1880's were as dominant as any we've ever seen (from a 1930 vantage point) and pitchers in the late 1890's were the least dominant we've seen (excluding the NA years). The mound change dropped strikeout rates back to the underhand days of the late 1870's and they didn't come back until foul balls were made into strikes in 1902.

% of outs by Strikeout, by 5 year intervals:
1871-75  2.1
1876-80  8.9
1881-85 14.6
1886-90 13.5
1891-95 10.3
1896-00  8.6
1901-05 13.1
1906-10 13.9
1911-15 14.6
1916-20 12.0
1921-25 10.3
1926-30 10.8

I don't where the idea came from that pitchers before 1893 just initiated the action, but it is absolutely and totally false. You don't strike out 19 major-league batters in a 9 inning game (Charlie Sweeney against defending champion Boston in 1884) while just waiting for them to hit the ball to the fielders.
   267. KJOK Posted: August 09, 2004 at 10:47 PM (#786739)
I checked a little further, and I found an intriguing discrepancy. I've been using the park factors printed in the STATS, inc. All-Time Major League Handbook, first edition (copyright 2000). For 1900, this book shows Boston having the rather startling park factor of 121, sandwiched between a 103 in 1899 and a 109 in 1901. It's not just a misprint in Willis's entry - it shows the same way for Nichols. The Baseball-Reference park factors for all three years cluster around 110. Nichols did not take a big hit in his ERA/RA in 1900 but Willis did. Is that 121 believable? And why are the two sources so very different?

Baseball-Reference uses the "Total Baseball" park factors, which are based on the current year PLUS the prior year PLUS the following year (3 year park factor)

STATS All-time Major League Handbook apparently uses the current year only.

If you're after value, you should use the Stats 1-year Park Factor.

If you're after ability, you should use the Total Baseball Park 3-year Factor.
   268. jimd Posted: August 10, 2004 at 12:53 AM (#787003)
STATS, inc. All-Time Major League Handbook, first edition (copyright 2000).

Be careful using STATS "park factors" and read the fine print. They're not always ready-to-use.

Nobody does park factors the same, and unfortunately, they use the same name "park factor" to refer to different processes. Maybe someday it'll all be standardized.

James' (STATS?) park factors usually don't bother with a couple of extra steps that Palmer (and BR.com) uses.

In Win Shares, James appears to assume that the road sample is equivalent to a neutral sample from all of the parks, which is not true unless you happen to be playing in a neutral park. If you're playing in Baker Bowl when it's amplifying offense by 35%, your road sample is from 95% of average, and simply doing the home/road division 135/95 and using that as the base estimate for Baker Bowl overstates its effect. (James' estimate 142; compare to "true" 135; it gets much worse for Home Run factors. I don't have the Handbook, but I have the All-Time Major League Sourcebook, and the "factors" given in there are these home/away ratios which are the raw material from which BR park factors are calculated).

OTOH, Palmer (and BR.com) includes another factor, independent of the park, which is a competition quality factor. It adjusts for not playing against your own defense (or pitching against your offense). Which is why there are separate Offense and Defense "Park" Factors. Ginger Beaumont get extra credit in 1909 for playing against the Pittsburgh (110-42) pitching staff while Fred Clarke is penalized for playing against the Braves (45-108) instead. It can make a difference in extreme cases like this. (This is why I wish that those who calculate composite records of a pitcher's opponents would also include the "expected" record. OPS+/ERA+ from BR.com already factors in the expected quality of opposition and if a pitcher's actual strength-of-opponents matches the expected strength-of-opponents then no further correction needs to be done, as was true for Bob Caruthers whose ERA+ and WARP already factor in his playing mostly for championship teams. Win Shares does not adjust for this).
   269. OCF Posted: August 10, 2004 at 03:38 AM (#787370)
jimd - there is a paragraph in the introduction of the STATS Handbook that includes the following sentence:

The effect of a park which increases runs scored by 10 percent is different in an eight-team league than it is in a 10-team league, because the Colorado Rockies are the only team in the National League that doesn't get to play any road games in Coors Field, and the impact of this is more diluted when there are more teams in the league.

I take it from that that they did do the first of the two corrections you mention above. The same paragraph also talks about adjusting for the actual number of home and road games, and correcting for the fact that bad teams play fewer defensive innings on the road than good teams.

They clearly don't make the second correction you mention, and they have a single park factor, not separate offense/defense factors.

The single-season park factors are just for years before 1909; after that, they use a weighted 5-year average except for documented changes in park.
   270. Kelly in SD Posted: August 10, 2004 at 06:55 AM (#787473)
I posted a career sketch of Three Finger Brown to the Pitcher Thread.
   271. Kelly in SD Posted: August 10, 2004 at 07:09 AM (#787478)
I have posted career sketches of Three Finger Brown, Joe McGinnity, and Arm Fall Off Boy Walsh because they all finished their careers with fewer than 3500 IP. Other than Welch, the Negro Leaguers, and the automatics: Johnson and Alexander, the current and upcoming candidates are pitchers with under 3500 career IP. Are these current candidates comparable in merit/value/achievement/etc. with the enshrinees who did not pitch a huge amount of innings? Yes, ERA and some rate numbers won't be matched, but how do they compare using numbers that are not so era-specific.
Hope the numbers help. I hope to do Leever, Cooper, Shawkey, Shocker, and S Coveleski (maybe Phillippe and Adams depending on their support)before Sunday, but my wife and I are trying to plan a relocation to Seattle and these sketches take a good deal of time.
Does anyone have any suggestions about information they would like included?
   272. Thane of Bagarth Posted: August 10, 2004 at 03:47 PM (#787762)
This my first time voting. There are a lot of players I see as closely bunched that just miss the ballot. I'm going to try to stick to this, but things could change before I submit it officially.

Right now I'm evaluating primarily with the BP cards and Baseball-Reference, plus Riley's book for Negro Leaguers. For hitters I focus on WARP3, EQA, OPS+, and defensive Rate (before timeline adjustments) and FRAA (timeline adjusted). On the pitching side of things I look at mostly at Translated IP, ERA+, WHIP, PRAA (season & all-time), and DERA (season & all-time). I try to use the BP timeline adjustments to proxy for era and league strength, but I pay more attention to how players did vs. their own league.

I give preference who have been able to dominate. I take longevity into account, but the HoM seems to me to be meant for players who were considered the "best of their time," not "pretty good for a long time."

First I'll post the 15 on my ballot, then number 16-40, which should cover most of the other notables who missed.

1) LUIS SANTOP: Every source I have seen says he is one of the all-time greats. Sometimes referred to as "the black Babe Ruth." He might not be Josh Gibson, but he's good enough for the top spot.
2) LIP PIKE: I'm wary of comparing players from the dawn of pro ball to the more modern guys, but he did seem to stand out in his time, however brief. Impressive 155 OPS+ and was even better than average in CF according to BP (Rate = 103). I don't think he is too far behind his contemporaries that are already in the HoM.
3) RUBE WADDELL: Definition of a dominating pitcher. Strikeout king of the '00s. Slight edge in career length puts him higher than A. Joss for me. Tied with Vic Willis for most shutouts for HoM eligible players that have yet to be elected.
4) ADDIE JOSS: 8 excellent seasons in a row. Clearly one of the top pitchers of the first decade of the 1900s. Not as many Ks a Waddell, but fewer walks and HRs made him just as effective. Third lowest BB/IP of his era (better than Cy Young and C. Matthewson). 2nd lowest opp. OBP and all-time leader in WHIP--impressive in any offensive context.
5) PETE BROWNING: One of the premier hitters of the American Association. Even if it wasn't always the strongest league, his success in the NL toward the end of his career leads me to believe that he would have been a star anywhere.
6) ROGER BRESNAHAN: His offensive contributions reletive to his position (even including CF) and consistancy when he was in the lineup make him a worthy candidate. His worst OPS+ in his prime (1903 to 1912) was 104 and every other year was between 129 and 162. Plus, catcher has seemed to be an underrepresented position.
7) RUBE FOSTER: His reputation by word of mouth puts him in, or close to, the top 5 Negro League pitchers ever. Projections are interesting and helpful, but the lack of thorough stats to back up his legend has me ranking him conservatively.
8) JOSE MENDEZ: May have been better than R. Foster, just not as much hype to spread the repution. Regardless, I'll put him behind Foster and remain confident that they were both great.
9) HUGHIE JENNINGS: Fantastic peak at the plate and in the field from '94-'98 are hard to ignore. He earned the superstar status he had in the late 19th century.
10) CHARLEY JONES: Conservatively ranking him down here. Has numbers that are comparable to Pike's (CJ: 149 OPS+/.321 EQA vs. LP: 155 OPS+/.326 EQA), though he played a little later and lasted a bit longer. Not sure how to treat his missed time.
11) BABE ADAMS: Surprisingly good. Missed some time with a sore arm, but came back to have 3 fantastic years from 1919-1921(ERA+: 150,153,144; led league in WHIP each year) to catapult his career from decent to very good. In 5 out of his 11 years with >150 IP he posted the lowest opp. OBP in the league.
12) ED CICOTTE: Started to dominate the AL when he got to Chicago in 1913. Two off years which weren't *that* bad ('15 &'18) can't cloud the fact that he was one of the elite pitchers of the 'teens. I see him as the 2nd or 3rd best pitcher in the AL from 1913-1920 behind Big Train, and maybe Stanislaus Kowalewski (boy, that sure has a ring to it).
13) FRED DUNLAP: Pretty good hitter for a second baseman (.307 EQA/ 133 OPS+) and BP card is high on his defense (Rate/Rate2 of 113/107 at 2B). I am considering moving him down for lack of games played, though the BP Translations have him at only 500 less ABs than Childs.
14) DOBIE MOORE: Lots of stories and some statistics indicate that this guy was fantastic in the 6 or 7 Negro League seasons before he got shot (ouch). I imagine he was as good or better than Jennings, it's just too hard to prove. His playing time on a top notch military team helps, too. I'm giving serious thought to moving him up.
15) JOHN McGRAW: OBP machine! An offensive force to be reckoned with and BP cards give him decent ratings for defense at 3B (Rate: 104). If he played a few more games I would have him much higher.
   273. Thane of Bagarth Posted: August 10, 2004 at 03:54 PM (#787771)
By the way, that should say "this IS my first time voting." Me know how to write good--honest.

THE REST:
16) Levi Meyerle: Another excellent early professional. Short career and weak defense have him just off my ballot. Could easily be flip-flopped with McGraw, in my mind.
17) Gavy Cravath: Very impressive slugger, if he took advantage of the his home park I think he still deserves the credit. Not so hot on defense. More consideration for his minor league performance would probably put him on my ballot. As it is, I think he's an appropriate player to keep Meyerle company.
18) Vic Willis: Peak seasons spread out between some mediocre valleys. Very close to Griffith in quality (118 vs. 121 ERA+) and edge in IP lands Vic one notch above him.
19) Clark Griffith: Numbers seem pretty close to Wilbur Cooper's, who I've ranked much lower. Providing his team with a little offense puts him ahead of Cooper and Leever.
20) Cupid Childs: I don't think his numbers (esp. on defense) on BPro make him look as good as some people have argued here. I am swayed some by those arguments, but he still just misses my ballot.
21) Bill Monroe: I think he was probably as good as Childs, or better, but not significantly better, so, I'll rank him conservatively due to the lack of stats. Like Dobie Moore, I'm thinking of moving him up.
22) Mike Griffin: Of the glut of OFs who all have about the same career length and OPS+ in the low 120s, Griffin gets the best defensive ratings, so he's on top of my list but most of these guys seem interchangeable.
23) Roy Thomas: For some reason Thomas' EQA is better than Griffin's (.309 vs. .302) but they have the same OPS+, so Thomas adds a little more value than Griffin with the bat, but gives up just a smidge more with the glove. Like I said, interchangeable.
24) Frank Chance: Quality first baseman, hitting and fielding, I just wish he played a little more. I might be moving him down.
25) Ross Youngs: Very good player, hard to rank because of his untimely death. Could see him justifiably ranked 5-10 spots higher or lower.
26) Mike Tiernan: BP card stats show him to be pretty similar to Cravath--both have .315 EQA, defensive rate of 98. Lower OPS+ (138 vs. 155) has him down here. I may move those two closer together upon further review, but it still won't affect this year's ballot.
27) Fielder Jones: Very close to Duffy. Jones was slightly worse with the bat and a little better with the glove. Almost the same number of games.
28) Spotswood Poles: I feel like putting Spot between Jones and Duffy does justice to an approximation of his career.
29) Hugh Duffy: Performance ranks slightly below Griffin and Thomas. Duffy played longer and is close to them in defensive Rate, but is offense not quite as good.
30) Jimmy Williams: Good combo of offense and defense for 2B. Did not play very long. Poor man's Cupid Childs?
31) Sam Leever: Excellent pitcher--along the lines of Adams, Griffith, and Cooper. He just didn't pitch very long. Incredible winning percentage.
32) Wilbur Cooper: Consistently good for about a dozen years. Peak not as high as Willis or Adams.
33) John Donaldson: Sounds like he was a good pitcher but not in the Negro League elite.
34) Jim McCormick & 35) Mickey Welch: I know there has been a lot of analysis of these two guys, but I just can't get excited about their stats. BP Translations are not terribly kind to either. I think their values are very close, but I like McCormick's numbers more.
36) Jimmy Ryan, 37) George Van Haltren, 38) Jake Beckley, 39) Bobby Veach & 40) Harry Hooper: The last five are all moderate-peak, good career guys that I don't think stand out. Their order could be rearranged, but I don't think they should make the HoM. However, all could easily be ranked up with Griffin or Thomas.
   274. TomH Posted: August 10, 2004 at 03:56 PM (#787776)
Van Haltren vs Ryan

This should be easy – two guys who played in the same years, at the same position, both with long careers and no huge peaks.

A.The WARPed view

I throw out years at the beginning and end of careers with less than 2 WARP as superfluous fluff.
GVH 1888-1901 14 yrs 83 WARP including –5 for pitching ….88 w/out pitching
JR ….1886-1903 17 yrs 90 WARP including –2 for pitching ….92 w/out pitching

I’m not going to penalize someone for their time on the mound that resulted in supposedly below replacement level efforts.

I don’t know about you, but an additional 4 ‘wins above replacement’ in 3 full years is not an advantage. 1 win for George

B. Win Shares

GVH 344 total, 28.1 per year
JR ...316 total, 25.4 per year

Ryan ranks ahead in the NBJHA, solely on the basis that his top 5 consecutive seasons (late 80s / early 90s) were better. Van Haltren had a good long string of seasons in the one-league 90s when Jimmy wasn’t doing much. To me, this is another win for George.

C. Offense/Defense/Pitching stats

GVH .620 OWP .287 EqA
JR ….596 OWP .286 EqA

I trust OWP more, since B James put so much time into making things work for the deadball era.
GVH has (about 45) more RCAA by the Sinins encyclopedia, but the RCAP are nearly the same – which makes little sense to me, since Van Haltren spent more time in CF than Ryan did?!?! Even if I gave Ryan an edge on defense, that would only bring them even. And then you have GVH’s near-league-avg pitching.

No, they aren’t far apart. Yes, it’s pretty close. But at least unlike attempting to rank Rube Foster vs Roger Bresnahan vs Pete Browning vs Clark Griffith, it’s pretty clear, ain’t it? And with the ballot as squished as it is, a “pretty small” difference can result in many ballot spots. At least it does on my ballot. I Vote for George.
   275. John (You Can Call Me Grandma) Murphy Posted: August 10, 2004 at 04:05 PM (#787790)
Welcome, Thane of Bagarth (Charlton Comics fan, huh? :-)

Ballot looks good, so you can add it to the ballot thread any time that you want.

The one thing you need to do is add comments for any players that you left off your ballot, but are top ten returnees from last "year." This would include Jimmy Ryan, George Van Haltren, Jake Beckley, Clark Griffith and Cupid Childs.
   276. DavidFoss Posted: August 10, 2004 at 04:06 PM (#787792)
Welcome Thane!
   277. Thane of Bagarth Posted: August 10, 2004 at 05:31 PM (#787918)
Thanks for the warm welcome, John and David.

I guess Thane of Bagarth is not as obscure as I thought. I had never heard of it 'til a friend gave me an issue she found at a yard sale 3 or 4 years ago. I'm not really much of a fan of the comic, but my 1st name is Thane, so what's not to like?
   278. PhillyBooster Posted: August 11, 2004 at 05:05 PM (#789948)
Still thinking about pitchers -- quality versus quantity. Would I vote for a pitcher with fewer than 3000 innings without a Caruthers-esque bat?

I don't have Babe Adams on my ballot, at 2995. I do plan to have Eppa Rixey high on my ballot, although his only advantage appears to be about 1500 league-average innings. Should I be giving more credit to short-career guys?

I find myself considering extreme cases of players who I won't have to consider for several more years.

Anyway, this is the question that I am currently finding myself unable to answer, and that has ramifications for every pitcher who ever comes onto the ballot:

If you were casting the deciding vote for the final member of the HoM, would you choose John Hiller or Jim Kaat?

Any thoughts?
   279. John (You Can Call Me Grandma) Murphy Posted: August 11, 2004 at 05:13 PM (#789962)
If you were casting the deciding vote for the final member of the HoM, would you choose John Hiller or Jim Kaat?

Hiller is also on my radar, Matt. I finally have my system so that it gives proper credit for relief pitchers (without overdoing it) and Hiller appears to be helped by my adjustments. Mike Marshall may have been more popular, but I think Hiller was better in his prime.

With that said, I don't know if either one will make my ballot. Too early to tell.
   280. OCF Posted: August 11, 2004 at 05:22 PM (#789982)
There are two more or less contemporary 3000-inning pitchers whom I'd take ahead of Rixey: Vance and Coveleski. I like Adams, but not as much as I like those other two.

For me, if pitcher A had about 1500 league average innings on pitcher B (processed through my system, that would come out as about an additional 80-80 record), then I would put A ahead of B. If the additional record was more like 65-95, then I'd rank them about even.
   281. PhillyBooster Posted: August 11, 2004 at 05:52 PM (#790043)
Rixey: 4500 IP, 115 ERA+
Vance: 3000 IP, 125 ERA+
Covaleski: 3000 IP, 127 ERA+

Vance + (1500 IP at 95 ERA+) = Rixey
Covaleski + (1500 IP at 91 ERA+) = Rixey

That additional value needed to bridge the gap from Vance/Covaleski to Rixey is essentially Len Barker's career.

So, substituting in our Covaleski formula above:

Covaleski + Barker = Rixey, or
Covaleski = Rixey - Barker

If "Barker" is a positive value (and I believe that it is, even leaving aside the perfect game), then I am finding it hard to say that I wouldn't pick Rixey over (Rixey - Barker), pretty much as a mathematical truism.

To vote for Covaleski over Rixey (or Hiller over Kaat), I need to find a way to bridge the "Barker gap". The other alternative is to conclude (and it may not be an unreasonable conclusion) is that the greatest thing a pitcher can do is pitch a lot at a minimum level of competency, and "relievers", which are really usage patterns, not distinct positions, simply need not apply.

In other words, in comparing 1000 innings of GREAT to 4500 innings of average, it's entirely possible to that 4500 innings will win every time. I don't know. I'm not writing off relievers entirely at this point, but my current prefence of Rixey over Covaleski seems destined to lead to huge preference for Kaat over Hiller, and I'm not quite sure (A) if that is the right answer, or (B) if it 's the wrong answer, what's wrong with it.
   282. OCF Posted: August 11, 2004 at 06:13 PM (#790086)
I have Rixie as having the equivalent record that would go with a RA+ of 111, with Coveleski and Vance both at 125. In other words, looking at RA+ rather than ERA+ hurt Rixie more than the other two. That leaves in my system the difference between Rixie and Covaleski at 66-90 and between Rixie and Vance at 74-95. That difference is less than Len Barker, but, I admit that they're all really about even, along with Faber (whose IP total is intermediate). Part of my attraction to Vance is that I'm a sucker for all those strikeouts.

Rixie is a worthy candidate. Note that the arguments used here for Rixie can also be used for Vic Willis, although IP mean something a little different in the 10's/20's than they do in the 90's/00's.
   283. PhillyBooster Posted: August 11, 2004 at 06:25 PM (#790110)
Hm. Maybe the perspective of watching Paul "3-11" Abbott trotted out every fifth day to lose to a broad variety of National League foes has left me salivating at the thought 66-90 quality pitcher.
   284. andrew siegel Posted: August 11, 2004 at 07:32 PM (#790253)
Covaleski/Vance/Rixey/Faber/Mays above Willis/Waddell/Joss/Griffith, below them, or mixed together?

Right now, I'm thinking mostly above but that may just be because I haven't looked at them closely enough to find their flaws.
   285. John (You Can Call Me Grandma) Murphy Posted: August 11, 2004 at 07:44 PM (#790277)
Rixie is a worthy candidate. Note that the arguments used here for Rixie can also be used for Vic Willis, although IP mean something a little different in the 10's/20's than they do in the 90's/00's.

True. They both need to be compared to their respective pitching environments.
   286. OCF Posted: August 11, 2004 at 08:24 PM (#790346)
Coveleski/Vance/Rixey/Faber/Mays above Willis/Waddell/Joss/Griffith, below them, or mixed together?

I'm thinking mostly above. Question for the Cuban experts: does Luque have enough outside of the major leagues to make him a member of that group? Luque does have probably the best single pitcher's year of the 20's in 1923.
   287. PhillyBooster Posted: August 11, 2004 at 08:38 PM (#790366)
Well, Griffith has not been chosen as one of the top 3 pitchers of the 1890s (Young, Nichols, and Rusie are already in), and Willis, Joss, and Waddell have not been chosen as among the top 5 pitchers in the 1900s (Mathewson, Plank, McGinnity, Brown, and Walsh are already in). Meanwhile, it's starting to look like the 1910s will have no challengers after Johnson and Alexander go in (unless there's a big surge in Cicotte support).

The top 3 1920s pitchers (Rixey, Covaleski, Vance) all go in before the rest. The next 4 (Luque, Hoyt, Mays, Faber) get mixed in with Griffith, Joss, Willis, Waddell, and Cicotte, with Luque edging the pack, and Griffith and Cicotte close behind.
   288. jimd Posted: August 11, 2004 at 08:55 PM (#790394)
I see Vance/Covaleski/Luque/Faber/Shocker all having peak arguments for being the best pitcher from the period between the decline of Walter Johnson and the rise of Lefty Grove. All else being similar, I'm more likely to vote for the pitcher who can claim to be the best (even if it was weak competition) over the one who can only claim to be third behind Mathewson and Young. Right now I'm leaning towards Vance, but I have to examine this more closely.
   289. Chris Cobb Posted: August 11, 2004 at 09:12 PM (#790421)
Meanwhile, it's starting to look like the 1910s will have no challengers after Johnson and Alexander go in (unless there's a big surge in Cicotte support).

Well, the Negro Leagues will supply some. Smokey Joe Williams will be elected, no question. We'll see how Jose Mendez does this year, and Dick Redding may get support a few years down the road (though the i9s projections do him no favors whatsoever).
   290. John (You Can Call Me Grandma) Murphy Posted: August 11, 2004 at 09:20 PM (#790433)
and Dick Redding may get support a few years down the road (though the i9s projections do him no favors whatsoever).

Looking at those projections, he won't be anywhere near my ballot.
   291. karlmagnus Posted: August 11, 2004 at 10:16 PM (#790503)
88 Given the 1924 WS, Johnson's 20 wins in 1925 and the fact that Grove got going properly in '26, there wasn't really a period between the decline of Johnson and the rise of Grove. The best pitcher in baseball in 1917-20, on the other hand, was surely Cicotte.
   292. Kelly in SD Posted: August 11, 2004 at 11:04 PM (#790552)
This is a long post with lots of comment and just a few numbers about pitchers in the 20s. This took a long time to write so if someone already posted similar ...

Re: Dominant pitchers between Johnson and Grove:

I am not sure everyone realizes that Johnson was the best pitcher in the AL in 1924 (29) and 1925 (26) by WS and the second/third best pitcher each year overall (behind Vance's 36, then Donahue's 28 and Luque's 27). Only two seasons would pass before Grove was the best in the AL - 1928 with 27. The gap is even less because Grove was 2nd in the AL in 1926, and 3rd in 1927. There really isn't much room between Johnson's decline and Grove's rise.

However, if we look back to 1920 when Johnson failed to appear in the AL's top 4 pitchers for 4 consecutive years and consider his "decline" to have started then, there is a different conclusion. The argument can definitely be made that he was not the dominant pitcher in the league from 1920-23.
Even in his decline phase, 1920-26 (end of career), he still had the 8th most WinShares of all pitchers and the 5th most in the AL with 141. The only pitcher comfortably ahead of Johnson was Urban Shocker with 166. The other pitchers were only ahead of Johnson by 5 win shares or less. (see chart below)
              WS 20-26 
Shocker, AL:    166    
Alexander, NL:  153    
Luque, NL:      151    
Rixey, NL:      150    
Coveleski, AL:  146           
Faber, AL:      146    
Rommel, AL:     145    
Johnson, AL:    141    


Looking at the best pitchers AL or NL based on Win Shares from the decline of Johnson, say 1920, until the rise of Grove, 1928. WS are the win shares in the period 1920-28. 20+ are years with at least 20 win shares - includes those years over 30. 30+ are years with 30 or more. Lg#1 are years with most pitching win shares. LgAS are years ranking in top 4.
         WS 20-28 20+ 30+ LgAS Lg#1 
Alexander:   200    5   1   3    2
Rixey:       187    6   0   5    0
Grimes:      187    5   2   4    1
Shocker:     182    5   1   4    0 (regular til 27)
Luque:       182    3   1   3    1
Pennock:     172    4   0   2    0
Rommel:      170    4   0   2    0
Faber:       164    3   2   2    2
Hoyt:        162    5   0   2    0
Vance:       158    4   2   3    2 (reg starting in 22)
Uhle:        149    3   1   2    2
Coveleski:   148    4   1   2    0 (done in 26)
Johnson:     146    4   0   2    2 (done in 26)
Cooper, Wil: 146    5   1   4    1 (done in 25)
Jones, Sad:  144    4   0   1    0
Lyons:       100    3   1   3    1 (regular 4 yrs)
Grove:        85    3   0   3    1 (regular 3 yrs)

Pitchers with #1 years who did not above list include: 1920 AL Bagby 34, 1925 NL Donahue 28, 1926 NL Kremer 25, 1927 NL Haines 28 (tied with Alexander).

If you choose 1920 and 1928 to mark the period between Johnson and Grove, for the Major Leagues, I would say Alexander based on peak and contributions to pennants/world series. Maybe Grimes over Rixey based on better peak years. Luque is also behind Alexander for peak and length, though Luque's single year peak of 39 beats Alexander's 36. Luque's career has the bonus, if you want to include it, of his time in Cuba. However, as Luque was starting every year in the 20s, I don't know how much credit you could give him for this time period.
For AL only, I guess Shocker would be the best candidate for best pitcher between Johnson and Grove.

However, you could make the argument there was no gap in the AL as Johnson led the AL in 1924 and 1925 and Grove was top 3 in both 1926 and 1927. Also, as Alexander is still one of the dominant pitchers of the decade, the 20s centered pitchers are playing for 3rd at best for the decade behind Alex, Johnson at the start, Grove at the end(IMHO).

Personally, I don't think most of the above pitchers will be making my ballot, and if they do it will be at the bottom.
Looking ahead to voting in 1933, 34, 35, 36, I "know" who is playing right "now" and can compare them to the above pitchers. Alexander, Johnson, and Grove are the best and I am not sure if anyone else stands out to be the best of the second tier. Vance may because of the condensed nature of his career, Luque could based on the addition of his Cuban years.
   293. OCF Posted: August 11, 2004 at 11:21 PM (#790577)
Thanks, Kelly. But I would warn agains concentrating only on the 20's. For that matter, I don't see "between Johnson/Alexander and Grove" as being a criterion to care about. Outlier cases do not define the boundary of the HoM.

Coveleski's three best years were 1917, 1918 and 1920.
Rixey was already a good pitcher in 1912. About 30% of his total value is 1920 and earlier.
Faber has a concentration of value in 1920-21-22, but he had good years both before and after that.
Mays was better in the teens than in the 20's.
Adams belongs to the teens.
Shawkey's value is nearly evenly divided between the teens and the 20's.
There are quite a few pitchers who were good both before and after 1920.
   294. Kelly in SD Posted: August 11, 2004 at 11:47 PM (#790652)
OCF,
I agree with you on your points. I was responding to posts in the 80s that talked about comparing these pitchers and seeing which one was the best between the decline of Johnson and the rise of Grove.
I thought the debate was forgetting the way that Johnson overlapped into the middle of the 20s at a high level, that Alexander was arguably the best pitcher of the 20s, and that Grove started to dominate as early as 1926.
   295. jimd Posted: August 12, 2004 at 12:05 AM (#790710)
Everybody looks at these things differently. I try and look at the player's context because the same height of peak when measured by WARP or WS will be perceived differently depending on what a player's peers are doing. If Johnson and Alexander (or Mathewson and Young) are performing at levels 15-25% higher, it isn't quite the same quality of peak as when the player is "the best" during that period. If you just look at the absolute height of the peak, well, there are more 1880's pitchers besides Welch still to be considered then.

One way I measure these things by rolling 5-yr totals:
(The year is centered: 1918 means 1916-1920.)
Year  WARP-1    WARP-3     ("Old" WARP; not yet updated)
1911: Johnson   Johnson
1912:
1913:
1914:
1915:
1916:
1917:
1918: (Ruth)    (Ruth)     Ruth is P/OF mix 1916-20
1919: Johnson   Johnson    Ruth is OF/P mix 1917-21
1920:
1921: Alexander Coveleski
1922: Faber     Faber
1923: Luque     Shocker
1924: 
1925: Vance     Vance
1926: 
1927:           Lyons
1928: Grove     Grove

There is a gap 1921-27 during which Walter is arguably no longer the best pitcher in baseball (though he is still the most prestigious, like Clemens today), Alexander is also in decline, and Grove is not yet established. It's not that they are done, it's just that they don't dominate anymore like they did the 1910's, and the other pitchers put up enough competitive seasons to factor into the mix of "who is the best", before Grove gets established.

It's just one way of looking at it.

Also "old" WARP-1 had Walter second best in 1924 to Ehmke and behind Lyons and Pennock in 1925.
   296. John (You Can Call Me Grandma) Murphy Posted: August 12, 2004 at 01:19 AM (#791146)
There is a gap 1921-27 during which Walter is arguably no longer the best pitcher in baseball

The new "Lively Ball" era must have been doing a number on the youngsters' arms during that time.
   297. Chris Cobb Posted: August 12, 2004 at 02:40 AM (#791491)
I wrote:

and Dick Redding may get support a few years down the road (though the i9s projections do him no favors whatsoever).

John Murphy wrote:

Looking at those projections, he won't be anywhere near my ballot.

It's too early to get into serious discussion of Dick Redding, but this is a case where expert opinion should make us look hard at the i9s projections -- he was selected as a HOFer by 80% of the CPPD experts, which puts him in very solid company. Here are the top expert vote-getters in that poll:

Biz Mackey 100%
Turkey Stearnes 96%
Christobal Torriente 96%
Dick Lundy 96%
Mule Suttles 88%
Hilton Smith 84%
Dick Redding 80%
Ben Taylor 80%
Jud Wilson 80%
Oliver Marcelle 80%
Louis Santop 76%
Jose Mendez 72%
Willard Brown 71%
Pete Hill 68%
Ray Brown 64%
Chet Brewer 64%
Frank Grant 64%

I'm not saying we'll elect all these guys, but we ought to consider the evidence carefully before summarily dismissing the merits of any one of this group.

I've argued in the past against simply accepting expert opinion on very early players, so here I'll argue against simply accepting the i9s projections for Redding. We ought to look at the real data first.
   298. Brent Posted: August 12, 2004 at 02:45 AM (#791510)
If I can change the subject from pitching to hitting :-)

As a new voter, I’ve been trying to work through evaluations of position players from the 1870s and 80s. I know you've been discussing Lip Pike and 1870s batting statistics for 35 “years” now, so I hope you will excuse me. My concern is how well the statistics that we use to evaluate offensive contributions – OPS+, RC, and EQA – work for 1870s baseball.

Perhaps the biggest statistical difference between modern baseball and baseball in the 1870s is all those errors. While in modern times, the events included in on-base and slugging percentages represent 85 to 90 percent of the actions going into run scoring, in the 1870s it was more like 55 to 65 percent. Most of the unmeasured 35 to 45 percent represent batters reaching base, advancing, or scoring on opposition errors. Since we know little about this hidden part of 1870s offense, I am concerned that we may be over-interpreting the parts that we do know.

There is no denying that Pike’s measured offensive statistics are impressive relative to the norms of the time: his career OPS+ of 155 is the same as Henry Aaron, Joe DiMaggio, and Mel Ott. If you prefer BP statistics, Pike’s career EQA “adjusted for season” is .326, again comparable to the same three.

My concern, though, is that these statistics are summarizing only the part of 1870s offense that is observable. My understanding of EQA is that there isn't anything special done to adjust the formula to account for all the errors in 19th century baseball. (Please correct me if I’m wrong.) That means that the folks at BP are essentially assuming that a player’s success on opposition errors is proportional to his success recorded by hits, total bases, and so forth. The RC formulas do make some adjustments to account for errors in 19th century baseball, but those adjustments appear to me to be seriously messed up; I’ll write about that another time.

It seems to me that if you want to make the case that Pike is a great hitter, you have to make some assumption about the “hidden” part of his offense. His unadjusted statistics – for example, a career .803 OPS when teams scored an average of 7 runs per game - are simply not strong enough by themselves to constitute “historic” offensive value. However, if his recorded hitting and baserunning skills were matched by similar proficiency at reaching on errors, then we can easily agree that he was one of the top offensive forces of his time.

I can think of two bits of admittedly indirect evidence that seem relevant, and both, at least for me, cast doubt on the idea that Pike was especially proficient at ROE. First, there is research on the types of batters who tend to ROE in the modern game. Craig Wright looked at this question in The Diamond Appraised, and more recently Keith Woolner analyzed it at the BP Website (here and here ). Both authors found strong evidence that individual batters and their characteristics do affect the probability of ROE. The batters who are most likely to ROE are right-handed ground-ball hitters – that is, the opposite of Lip Pike. Since in the 1870s, as today, the largest number of errors are committed at 3B and SS, I suspect that this characteristic prevailed even then. Surprisingly, Woolner found that speed had little effect on success at ROE.

The other evidence is that Pike, despite reaching scoring position under his power more often than any player of his time other than Barnes, was seldom one of the league leaders in runs scored – his highest rank was 10th in 1871. Several players of the time scored about as often as they reached base on hits and walks (for example, see Dave Birdsall and Ned Cuthbert for 1871, George Wright and Ezra Sutton for 1872). Pike was never one them.

Although baseball analysts are now trained to avoid using runs scored because of their biases, for analyzing the 1870s this statistic may be particularly important because it is the only one available that includes the effects of a player’s ability to ROE. (It’s interesting to note that the earliest widely accepted offensive statistic was runs per game, until it was supplanted in the late 1860s by hits per game and somewhat later by batting average.)

I believe that if one could collect them, there might be quite a few other pieces of data that could shed light on the role of errors in 1870s and 1880s offenses. For example, on p. 18 of The Hidden Game of Baseball, Thorn and Palmer mention that “reached first base” (including ROE) was recorded as an official statistic in 1879. Is anyone aware if these data are available? IIRC, I believe there were also attempts in the early statistics to split offensive runs into earned and unearned varieties. Finally, a study of box scores might help determine which teams received the most benefit from opposition errors.

I’m old enough to remember quite a few “discoveries” of sabermetric research that we later learned were simply misinterpretations of the numbers – for example, Roy Smalley as a defensive wizard, Ashburn as a better center fielder than Mays, and Thorn and Palmer’s nominations of Fred Pfeffer and Harlond Clift to their version of the HOM. I fear that with additional research we may come to consider the movement to enshrine Lip Pike in the HOM in this category. To me it doesn’t seem implausible that once we better understand the role of errors in 1870s baseball, Pike’s statistics could be deflated and some other unexpected player could take his place as the 2nd best offensive player of the NA era. Can any of Pike’s supporters reassure me that reaching base on errors wasn’t possibly a gap in his offensive credentials?
   299. DavidFoss Posted: August 12, 2004 at 03:19 AM (#791608)
Huh... interesting post Brent!

Unfortunately, Lee Sinin's encylopedia doesn't do the NA, but here are the totals for 1876-77, the last two years that Pike played full time


1876-1877
LEFT HANDED HITTERS
HITS displayed only--not a sorting criteria
WALKS displayed only--not a sorting criteria
AT BATS displayed only--not a sorting criteria

RUNS                             R        H       BB       AB     
1    Deacon White                117      207       15      569   
2    George Hall                 104      185       20      537   
3    Lip Pike                    100      169       17      544   
4    Joe Start                    95      163        7      535   
5    Tom York                     90      135       13      500   
6    Tim Murnane                  83      126       14      448   
7    Jimmy Hallinan               80      119        7      402   
8    Bob Addy                     63      108       11      387   
9    Dick Higham                  59      102        2      312   
10   Orator Shaffer               38       74        9      260   




RIGHT HANDED HITTERS
HITS displayed only--not a sorting criteria
WALKS displayed only--not a sorting criteria
AT BATS displayed only--not a sorting criteria

RUNS                             R        H       BB       AB     
1    Ross Barnes                 142      163       27      414   
2    George Wright               130      180       17      625   
3    Jim O'Rourke                129      198       35      577   
4    Cal McVey                   120      205       10      574   
T5   Cap Anson                   115      196       21      564   
T5   John Peters                 115      195        4      581   
7    John Clapp                  107      172       16      553   
8    Paul Hines                  106      174        2      566   
9    Jack Burdock                101      152       15      586   
T10  Andy Leonard                 99      163        9      575   
T10  Jack Manning                 99      156       12      540   

   300. OCF Posted: August 12, 2004 at 03:19 AM (#791609)
I'm the one around here who tried to use actual runs scored, or R*, as I called runs scored scaled to a number which was a "standard of excellence" for each league in each year. I used this - successfully, I think - to argue for the election of Harry Stovey. At first glance, Pike should be similar to Stovey: power hitter, great reputation for speed. There are two problems with attempting to apply it to Pike. The first is that I only extended it back to 1876, through the NL years. The irregular, unbalanced schedules of the NA make me reluctant to set it up for those years. The second is something that Brent alludes to: Pike's runs scored just weren't that impressive. In 1876-77-78, the standard of excellence (what it took for 100 R*) was 65, 56, 52. The league leaders were 126 (Barnes), 68 (O'Rourke), 60 (Higham). Pike's runs for these three years? 55, 45, 32. That doesn't tell us much, since Pike was in his 30's and declining - but Brent wasn't that impressed with the runs from Pike's NA days. Pike was clearly not being used as a leadoff hitter, which has a large impact on runs.

It's not an argument against him, but he doesn't get that special boost that Stovey got.
Page 3 of 4 pages  < 1 2 3 4 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
robinred
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.4665 seconds
49 querie(s) executed