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

Tuesday, November 11, 2003

1914 Ballot Discussion

We’re electing one this year . . . and there’s a slew of top new eligibles:

 WS  W3 Rookie Name-Pos (Died) 
305 85.9 1892  Joe Kelley-LF (1943)
290 81.0 1896  Fielder Jones-CF (1934)
274 83.0 1895  Jimmy Collins-3b (1943)
269 52.6 1899  Joe McGinnity-P (1929)
243 53.5 1895  Al Orth-P (1948)
233 50.8 1897  Jesse Tannehill-P (1956)
209 47.9 1895  John Anderson-LF/1b (1949)
183 43.1 1896  Dan McGann-1b (1910)
149 41.9 1901  Socks Seybold-RF (1921)
177 32.6 1899  Jimmy Slagle-CF (1956)
155 30.2 1899  Charlie Hickman-1b (1934)
126 22.4 1900  Sammy Strang-3b/2b (1932)
164 14.9 1896  Chick Fraser-P (1940)
119 22.8 1903  Jake Weimer-P (1928)
098 25.1 1899  Ossee Schreckengost-C (1914)

I’m assuming the WS and WARP numbers above are not adjusted for season length. Hopefully I’ll have the pennants added thread updated by tomorrow.

Joey Numbaz (Scruff) Posted: November 11, 2003 at 04:08 PM | 203 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

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

Page 1 of 3 pages  1 2 3 > 
   1. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 04:13 PM (#518951)
098 25.1 1899 Ossee Schreckengost-C (1914)

I know he has a crappy number of WS, but he probably was the best AL catcher at the beginning of the last century.

He still won't be on my ballot, though.
   2. Howie Menckel Posted: November 11, 2003 at 04:52 PM (#518952)
American Association players in HOM (minimum 10 G in that season)

1882 - McPhee CIN
   3. RobC Posted: November 11, 2003 at 04:55 PM (#518954)
Very early prelim ballot, not even time to make many comments:

1. Thompson
   4. Howie Menckel Posted: November 11, 2003 at 04:59 PM (#518955)
Career votes-points leaders
   5. MattB Posted: November 11, 2003 at 05:11 PM (#518956)
I have come to once again talk about pitching, and why everyone who claims we already have "too many pitchers" are wrong.

Today, I have looked at the Top 50 pitchers in terms of wins, under an initial assumption that, generally, the best pitchers will correlate highly with the winningest pitchers. The earliest of those pitchers to retire was Al Spalding, who recorded his last win in 1877. Beginning then, I divided the Top 50 pitchers into 5 categories, one for each of the five 25 year "eras" between 1877 and 2002.

Range: # in Top 50 in Wins (# in Top 25, # in Next 25)

1877-1901: 11 (6, 5)
   6. Howie Menckel Posted: November 11, 2003 at 05:12 PM (#518957)
Eastern League info helpful on McGinnity. In those days, you didn't always leave the majors because you stunk, so continuing to play has some merit (with big discounts, of course, but McGinnity only needs a little help if any at all....)
   7. DanG Posted: November 11, 2003 at 05:17 PM (#518958)
Ossee Schreckengost-C (1914)

I know he has a crappy number of WS, but he probably was the best AL catcher at the beginning of the last century.


Connie Mack would agree with you.

It's stories like Ossee's that make it fun to look at all the players on the newly eligibles lists. He sounds like the ideal person to be Rube Waddell's roomie.

He is apparently not in good health as we deliberate our ballots for 1914. He reportedly has kidney disease and seems unlikely to last through the summer.
   8. Chris Cobb Posted: November 11, 2003 at 05:29 PM (#518959)
MattB has mentioned that the shortness of Joe McGinnity's career is due to late arrival in the majors (age 28).

Howie has mentioned his play in the Eastern League after he left the majors.

I am certainly willing to give credit for non-ML play if it looks like the player was capable of competing at ML level but was for some reason prevented from doing so.

Everything I know about McGinnity suggests that this was not the case with him. His baseballlibrary.com bio indicates that he reached the majors late after developing a change-of-pace that greatly increased his effectiveness. His last major-league season indicates that his effectiveness as a ML pitcher was in serious decline.

It seems to me, therefore, that McGinnity is not a player who should receive credit for non-ML play.
   9. OCF Posted: November 11, 2003 at 05:50 PM (#518960)
Last year I said that I would be rethinking my votes for 3rd basemen this year. What follows is just some raw data, WARP 3 and OPS+: me thinking aloud, without much analysis yet.

WARP3, sorted by year, highest to lowest.
   10. MattB Posted: November 11, 2003 at 06:09 PM (#518961)
Eligible Pitchers ranked by WARP-1 (not adjusted for anything):

Tony Mullane: 100.4
   11. Paul Wendt Posted: November 11, 2003 at 06:18 PM (#518962)
AL 1900:
   12. DanG Posted: November 11, 2003 at 06:59 PM (#518963)
In support of MattB, here's a more simplistic way of looking at pitchers' representation.

In the HOM project, we're mainly looking at players born in 12 decades, the 1850's-1960's. After our final election (2012, at this pace) we will have 247 HoMers. 25% of that is 62. If we assume that many pitchers, that's obviously 5 born each decade.

Born in the 1850's we have:
   13. Marc Posted: November 11, 2003 at 07:26 PM (#518966)
Prelim ballot position players only. Major changes in my methods and therefore the results. 1) Have fully integrated WARP, not just using WS. 2) I now calculate a "prime" to go with "peak" and "career." Still a slight emphasis on peak, however, and career evaluation includes some rate numbers (e.g. EqA) as well as counting numbers (i.e. career WS, career WARP). 3) My peak calc used to be 3 consecutive years and 5 consecutive years. Now it's 3 consecutive and any 5.

1. Sam Thompson (13)--I had him as high as #2 and then as low as #15 but I was right the first time. Does not do well with WS or peak. Does very well with WARP and had a wonderful 9 year prime. Next to Bid McPhee, benefits more than anybody by differentiating the prime from the peak, and along with Charley Jones, benefits from the 5 year peak being non-consecutive rather than consecutive.

2. Charlie Bennett (1)--looks great from any angle.

3. Cal McVey (2)--benefits from consideration of a prime, but drops down a bit for peak using WARP.

4. Dickey Pearce (3)
   14. Marc Posted: November 11, 2003 at 07:38 PM (#518967)
If Fielder Jones were Minor League Player of the Year for 10 years he would still only be Jake Beckley, who I have as the #25 position player.

If you're looking for a player hurt by the management, take another look at Charley Jones. He had the career that the casual fan sometimes attributes to Pete Browning.

Jimmy Collins is a conundrum wrapped up in an enigma. We've had some famous players who don't have the numbers, but so far no player's numbers fall so far short of the reputation than Jimmy's. I see Ed Williamson as a much better player though I will add that I see Collins ahead of the 3B glut of McGraw, Lyons, Nash and Cross. Somebody please make Jimmy's case!
   15. Marc Posted: November 11, 2003 at 07:46 PM (#518968)
Finally, and then I'll shut up, for pure peak among position players, my excuse being that DanG asked for it. Remember it's a composite of WS and WARP, and of 3 consecutive and any 5. I straight rank order these 4 measures and then add up the rankings, low man wins.

1. Jennings
   16. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 08:18 PM (#518970)
I have Jimmy Collins at around #5. He's the best of the eligible third base guys. He doesn't fall too short of his rep, IMO.

Marc, how is Williamson better than Collins? I have Collins as better in hitting and fielding plus better competition.

As you can see from this chart, it was NOT easier for 19th Century pitchers to rack up lots of wins by pitching tons of innings. Rather 19th Century pitchers were about average in terms of racking up wins.

Matt, your comparing an era with one and two-man staffs to the current five-man staffs (plus many more teams) of today. As a percentage of the pitching pool during the 19th century, there were many more pitchers who had 250 wins or more than later eras.

As for the 20% cutoff, I have said it should be increased as the pitching staff expands. When we get to the four and five-man staffs, four or five pitchers (on average) should be selected (more or less).
   17. jimd Posted: November 11, 2003 at 08:22 PM (#518971)
Plainly, the quality falls somewhere between the worthless UA and the 1880s AA, but where?

The numbers that I derived from (BP) Davenport's DERA offset would place the Federal League at about a 13% discount to the AL/NL. This is about the same level as the AA (when averaged over its 10 year span). However, the AA ranged wildly in quality depending on the year, according to the DERA offset. In 1882, it's pretty comparable to the UA, though much better organized, and around AAA quality in 1883 (relative to the NL). 1884 and 1890 are similar to 1883, but the NL took a good hit in those years from the UA and the PL, narrowing the gap considerably. From 1885-89, the AA was definitely major-league, in the same sense that the AL was major-league from 1956-1965. So the Feds appear to stack up like the AA of 1884, 1890, 1891, ie, AAAA between AAA and major-league.

NL dominance over AA (1882-1891)
   18. Marc Posted: November 11, 2003 at 08:46 PM (#518972)
John, I will take a closer look at Big Ed and Little Jimmy. But my first question would be which WARP are you using? 1-2-3? Adjusted how? If you take WARP 3 as is, the timeline is huge. Reduce the timeline--i.e. adjust WARP1 (and with it the various runs measures) to season length--and Williamson starts looking very, very good very, very fast.

adjWS (for season length) 3 consecutive year peak--Ed 109 Jimmy 89
   19. MattB Posted: November 11, 2003 at 08:51 PM (#518973)
John wrote:

"Matt, your comparing an era with one and two-man staffs to the current five-man staffs (plus many more teams) of today. As a percentage of the pitching pool during the 19th century, there were many more pitchers who had 250 wins or more than later eras.

As for the 20% cutoff, I have said it should be increased as the pitching staff expands. When we get to the four and five-man staffs, four or five pitchers (on average) should be selected (more or less)."

I strongly disagree. This would make sense if there were HoM spots for "staff ace," for "#2 starter," for "#5 starter/long reliever", but realistically, when considering who is HoM-worthy, we are only considering staff aces. In a rare situation (e.g., Yanks or D-Backs), a team will have two pitchers who could be considered aces, but more often not. The best are still the best.

Imagine, in comparison, that in a major rule change, permitted rosters are cut dramatically next year (say, to 20 players). Assume further than the majority of the teams determine that the most effective way to adjust is to only carry one catcher. The backup will be some catch-and-throw guy in AAA. Suddenly, durability becomes a major issue. Teams give up a little offense on the margins for the ability to catch 150+ games in a year. A catcher who goes 162 games really helps out his team a lot now.

Under your logic, though, when we cut back to one catcher, even though the quality of the team's catcher becomes even more important, we're going to cut back the number of catchers we elect to the HoM.

19th Century pitchers were selected in large part for their durability. A Pedro Martinez-type would be all well and good, but he just wouldn't cut it on a 19th century pitching staff. Pedro might be just as good in the 19th century, but his skill sets would be about as welcome as a pole vaulters.

Pitching was a different position then. Not better or worse, but different. If you could only go 200 innings, you might as well just stay home. You might not pitch as long, but you got all your value in by showing up every day.

In any event, Griffith and McGinnity played primarily in 3 man rotations and McCormick and Welch played primarily in 2 man rotations. I don't see how that fact alone should increase Griffith and McGinnity's chances of election with respect to the others.
   20. Chris Cobb Posted: November 11, 2003 at 09:08 PM (#518974)
1914 Preliminary Ballot

This is a very preliminary ballot. I definitely want to see Pennants Added calculations on these newly eligibles before I settle on placements for them.

A few notes on the numbers. Win shares are adjusted for fielding, season-length, and league quality. Pitchers' win shares are derived from WARP ratings, not official BJWS. In ranking infielders against outfielders and pitchers, I give them a 10% bonus to career value. In ranking pre-1893 pitchers, I apply a 25-33% deduction to total peak, which is the sum of the win shares a player is above average in each season of his career. These last two modifications are not reflected in the printed numbers.

1) Cal McVey (3) (4) (3) (1) 354 CWS. Total peak = 76. Peak rate, 71-79 = 34.86 Among top 5 position players 5 times, at/above avg. WS in 9 seasons. After a near miss last year, McVey remains at the top of my ballot. His peak is the third-best on the ballot, after Jennings and McCormick, though Kelley's is a close fourth. McVey was more dominant relative to his league than Kelley was. Among the very best in the game for a decade, and possibly longer. Wasn't a great defensive player, but his ability to play key defensive positions and his very high peak make him the best player eligible.
   21. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 09:33 PM (#518975)
19th Century pitchers were selected in large part for their durability. A Pedro Martinez-type would be all well and good, but he just wouldn't cut it on a 19th century pitching staff. Pedro might be just as good in the 19th century, but his skill sets would be about as welcome as a pole vaulters.

Are you saying that the 19th century guys were more durable? They threw less pitches (and less harder) than contemporary hurlers so it was easier to achieve their season totals.

The 19th century pitchers were not iron men compared to today. There is no way that they could duplicate what they did in today's game. You're overrating them in the same way as someone stating that the power hitters of the nineties were so much better than the seventies.

Pedro would have vastly more wins per season if he had pitched during the 1880s, except he might only have had an eight year career instead. He would still be much better than McCormick, Griffith and the rest, however.

I have no idea where McGinnity goes yet, BTW. He has a good peak argument.

I strongly disagree. This would make sense if there were HoM spots for "staff ace," for "#2 starter," for "#5 starter/long reliever", but realistically, when considering who is HoM-worthy, we are only considering staff aces. In a rare situation (e.g., Yanks or D-Backs), a team will have two pitchers who could be considered aces, but more often not. The best are still the best.

I actually agree with you here, but the problem here is that we have a vast number of HoM slots to fill. If I have a choice to honor the second best pitcher or second best catcher in later elections, I'll honor the second best pitcher (on average) instead. If we were honoring just the cream of the crop from each position, we would have to decrease the number of HoMers for the later elections.
   22. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 09:48 PM (#518976)
Marc:

As you know, I couldn't care less about the timeline. The early stars of baseball can compete with the contemporary guys on my ballot anytime. However, you still have to adjust for competition or else the top stars of the 1880s will stand out to a greater degree than they deserve compared with the later guys. That's why Collins is comfortably ahead of Williamson in my book.

As for WARP1,2 or 3, I've been vaccinated for it. :-)
   23. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 09:50 PM (#518977)
Actually, I like the pitching/bad fielding team aspect of WARP.
   24. MattB Posted: November 11, 2003 at 10:26 PM (#518978)
John wrote:

"Are you saying that the 19th century guys were more durable? They threw less pitches (and less harder) than contemporary hurlers so it was easier to achieve their season totals.

The 19th century pitchers were not iron men compared to today. There is no way that they could duplicate what they did in today's game."

I am not trying to put the 19th century players on a pedestal. They didn't pitch as hard each game, but they pitched many more games per season, but they burnt out earlier.

What I am saying is that I think it all evens out in terms of final results.

It was no easier to win 300 or 250 or 200 games then than it is now, and we should be equally impressed by those early win totals.

When we get to Jim Palmer and his 268 wins and 3948 innings of 125 ERA+, there will be no way that he's not near the top of every ballot. What I saying is that what he did was no harder, and no rarer, than what Jim McCormick did with 265 wins and 4275 innings of 118 ERA+.

And McCormick was both approximately the 8th best pitcher of his quarter-century, while Palmer was about the 12th best pitcher of his.
   25. Paul Wendt Posted: November 11, 2003 at 10:51 PM (#518979)
In #58, I wrote:
   26. Chris Cobb Posted: November 11, 2003 at 10:53 PM (#518980)
Marc, if I recall correctly, you give a percentage increase to fielding WS up to 1892, and drop the increase completely from 1893 on. If you're still doing that, you should note that pre-1892 position players will enjoy a built-in peak advantage (in addition to the competition advantage that John Murphy mentioned).

I handle matters differently in two ways. First, I am gradually phasing out my fielding bonus after 1900. It drops to 1.25 1901-1910, 1.2 1911-1920, 1.1 1921-1930, 1.05 1931-1940. With a fielding adjustment, Collins's numbers look a good deal better. Second, when I calculate peak, I calculate it using WS above average, not WS above zero, so as the fielding bonus drops, average also drops, so peak values do not make a significant slip downwards.

I like Williamson -- he continues to hold down a spot on my ballot -- but Collins looks to me like a much better player. Even with the adjustments I've mentioned, his peak is a little lower than Williamson's, but both his prime and his career are substantially better. Hence his high ranking on my preliminary ballot.
   27. Paul Wendt Posted: November 11, 2003 at 10:57 PM (#518981)
(sigh)
   28. Marc Posted: November 11, 2003 at 11:04 PM (#518982)
Chris, you are right, but that is just WS. I lack the patience to repeat those calcs for WARP, and the WARP seasonal adj is just *2/3, and Ed still comes out ahead.
   29. Jeff M Posted: November 11, 2003 at 11:14 PM (#518983)
I think we are underrating Ed Williamson here.

...and I was blissfully thinking Williamson was gone for good. :)
   30. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 11:39 PM (#518985)
It was no easier to win 300 or 250 or 200 games then than it is now, and we should be equally impressed by those early win totals.

Depends on how you define "now" is. Youe 1977-2002 grouping has a disproportionately high number of pitchers who started their careers during the mid-sixties when conditions were at their most favorable. They have nothing to do with the 1980-2002 guys. How many pitchers who started their careers from 1977 are on your list?
   31. John (You Can Call Me Grandma) Murphy Posted: November 11, 2003 at 11:41 PM (#518986)
Let's try this again: :-)

It was no easier to win 300 or 250 or 200 games then than it is now, and we should be equally impressed by those early win totals.

Depends on how you define "now." Your 1977-2002 grouping has a disproportionately high number of pitchers who started their careers during the mid-sixties when conditions were at their most favorable. They have nothing to do with the 1980-2002 guys. How many pitchers who started their careers from 1977 are on your list?
   32. OCF Posted: November 11, 2003 at 11:48 PM (#518987)
<i> In #58, I wrote:
   33. jimd Posted: November 12, 2003 at 12:12 AM (#518988)
when considering who is HoM-worthy, we are only considering staff aces

I don't disagree with this, but I would like to point out that it may be more difficult to become the best "staff ace" than the best SS given that today there are 30 starting SS jobs and 150 starting pitcher slots. So maybe, a priori, there should be more pitchers than shortstops in the modern HOM.

Modern rosters are nearly one half pitchers. 50% pitchers in a modern HOM seems too high but 11% (one ninth) seems too low. Is there some method to work out the proper number?
   34. Marc Posted: November 12, 2003 at 12:28 AM (#518989)
How about 52% x 67.5% = 35.1%?

Or better yet, how about if 35.1% of the total value of players in our HoM was contributed by our pitchers?

The question would of course be how to determine value. adjWS? adjWARP1? WARP3?
   35. Chris Cobb Posted: November 12, 2003 at 01:39 AM (#518991)
Chris, you are right, but that is just WS. I lack the patience to repeat those calcs for WARP, and the WARP seasonal adj is just *2/3, and Ed still comes out ahead.

WARP already has higher fielding values for the earlier game built into its system. My firm belief is that its fielding values, esp. for the best defensive players, are too high, so that players like Williamson, McPhee, and Bennett are substantially overrated by WARP. In using fielding-adjusted WS as you do and comparing it to WARP, it appears to me that you are using systems that, rather than providing checks on each other's excesses, are likely to be overbalanced in exactly the same way. That's my take on the consequences of the metrics you are using.
   36. Jeff M Posted: November 12, 2003 at 02:22 AM (#518992)
1. Stovey, Harry -- I'm keeping him at #1. I would have elected him way back in 1905. Not quite as good a pure hitter as McVey, and a smaller peak, but a much longer documented career and I give the nod to documentation vs. non-documentation. He was also a run scoring machine with some pop in his bat. I've discounted his years in 83-85 and 89, and he has consistently held his spot on my ballot.

2. McGinnity, Joe -- Solid WS numbers. Fantastic winning percentage and excellent Wins Above Team. Has some nice counting stats and good grey ink scores. Would probably have won 2 Cy Young Awards. Easily the best pitcher available for the ballot.

3. McVey, Cal -- My park-adjusted normalized OPS for him is .892, which is damn good for a career. Essentially an all-star every documented year, and he was consistently 50% better than the league at the plate. Outstanding peak. I have given some intangible credit for pre-NA and post-NL play, but not as much as I gave Start for his pre-NA play (because of the age factor).

4. Kelley, Joe -- Another pure hitter that I've got at about 55% better than the league from a RC/27 perspective. Excellent WS peak and good WS career total. Would not have won MVP awards, but would have been in contention for a few. Was regularly an important player on championship teams.

5. Bennett, Charlie -- Gets a boost for being a catcher because my rating system seems to undervalue catchers a bit. I've got him about 20-25% better than the league as a hitter, which is pretty good when you consider what an outstanding defender he was. He also has a nice peak compared to other catchers.

6. Browning, Pete -- I've been on the Browning bandwagon for a while. He's even a better pure hitter than McVey, but probably not quite as good as Kelley. His suspect defense drops him behind the other four guys. A bit one-dimensional. I have discounted his 82-85 and 89 seasons but he proved in the PL that he was no fluke. I think he's a HoMer.

7. Collins, Jimmy -- Fantastic on defense at a key position. I would have him ranked ahead of McPhee, because he played a tougher position and was a better hitter. Among the best 3b in history (though I admittedly see 3b as a fairly weak position over the course of MLB history).

8. Griffith, Clark -- I believe he is the second best eligible pitcher. An excellent win pct on some bad teams. I boost his win totals and win pct by approximately 1/2 of his Wins Above Team, which are outstanding. Has a nice career Linear Weights total also. I'm not convinced he's a HoMer, but I'm comfortable with his placement here.

9. Thompson, Sam -- Another pure hitter with questionable outfield defense. I don't think he was as good a hitter as browning. He didn't have an incredible peak or career, from a WS perspective, as outfielders go.

10. Jones, Charley -- I give no additional credit for blacklisted seasons. If I did, he would probably be Top 5. He hit about as well as McVey, with power, but with a smaller WS peak and fewer WS per 162 games.

11. Grant, Frank -- Since our initial discussions, no new evidence has come to light so I haven't really moved him. I don't see clear and convincing evidence that he is a HoMer, but I see evidence he would have been a very good major leaguer. I give him the benefit of the doubt.

12. Long, Herman -- Not quite the hitter that McPhee was and not as historically dominant at his position as McPhee, but he did play a tougher defensive position very well for a long time. He has a nice WS peak as shortstops go.

13. Welch, Mickey -- Even though it was easier to earn wins then than now, he also had a significantly better win pct than you would expect, given the teams he played on. I don't see him in the HoM, but I think he continues to deserve a spot on the ballot.

14. Duffy, Hugh -- Frankly not a much better hitter than Bennett and at an easier defensive position. He has a decent career length and some nice Run and RBI totals which help get him on the ballot, but not in the HoM. I give him a nod over some of the other outfielders because he contributed to several championship teams from a key position.

15. Jones, Fielder -- This one totally surprised me. He bumps Beckley to #16. Tremendous defense and pretty good peak and career WS numbers. At times hit almost as well as Kelley and Browning. Wish he played a bit longer and had a little more grey ink.

The consensus Top-10ers who I do not include on the ballot are (1) Jimmy Ryan. I see Ryan as a good, but not great hitter. I don't see much of a peak as outfielders go, his career WS numbers don't floor me compared to other outfielders and he was nothing special defensively. He's just one of the outfield glut. (2) Lip Pike. My ranking of Pike hasn't changed. He's usually lingering at the bottom of my ballot. For some reason, he suddenly shot up on the consensus ballot, though I haven't seen any new evidence that would merit such a change. I would have Pike above Ryan, for sure, and Pike's ranking is probably around #18 in my system.
   37. Marc Posted: November 12, 2003 at 03:03 AM (#518993)
Our HoM contains 27 players, 7.5 of them (27.8%) pitchers. I separated John Ward's value as a pitcher vs. hitting/fielding. I did not separate any other hitting by pitchers or pitching by hitters (Burkett, et al), which means that pitching is probably underrepresented in these numbers. i.e. Of the 734.4 WARP1 earned by pitchers, some of that was for their hitting.

On raw WARP1, the pitchers account for 26.1% of the value.
   38. Howie Menckel Posted: November 12, 2003 at 03:09 AM (#518994)
"I am certainly willing to give credit for non-ML play if it looks like the player was capable of competing at ML level but was for some reason prevented from doing so.

"Everything I know about McGinnity suggests that this was not the case with him. His baseballlibrary.com bio indicates that he reached the majors late after developing a change-of-pace that greatly increased his effectiveness. His last major-league season indicates that his effectiveness as a ML pitcher was in serious decline.

It seems to me, therefore, that McGinnity is not a player who should receive credit for non-ML play."

this was Chris Cobb's comment from above. I'll duly look into this before I vote. Gracias.....
   39. jimd Posted: November 12, 2003 at 04:30 AM (#518995)
Or better yet, how about if 35.1% of the total value of players in our HoM was contributed by our pitchers?

I know that number is derived from Win Shares. It also matches up fairly well with the current level of pitchers in the HOF (by that standard, the HOF is a little tough on pitchers but not by a lot). That value matches up very well with a 4-man pitching staff model; 8 everyday players, each contributing 8.1% of the total value, plus 4 starters doing 8.7% of the total value.

The problem comes when that "total value" is divided up amongst more players at one position than at another. The example we are seeing in the early years is Catcher. 19th century teams (after the schedule hit 100 games or so) routinely started dividing up the catching duties amongst two or three players (barehanded catching was a b**ch), just like the pitching duties. If the total catching value is equal to the total SS value (for example), then catchers just can't accumulate enough value over a career or a season (for peak). If a person's evaluation system doesn't adjust for that somehow, then the personal HOM has no catchers (which may or may not be OK with that person).

We may run into similar issues with pitching, particularly in the era of the 5 man rotation with bullpen specialists; the innings get spread too thin for individuals to accumulate career value to compete with the everyday players. OTOH, the era we're currently evaluating may have the pitching evaluated too highly with 3 starters splitting that value; it doesn't seem to be an issue though due to the shorter-career/wear-and-tear issues.

Again, I'm bringing this up because I'm interested in exploring how different people are approaching the resolution of this problem in their various ranking systems.
   40. MattB Posted: November 12, 2003 at 04:37 AM (#518996)
"How many pitchers who started their careers from 1977 are on your list?"

Well, obviously very few who started their careers after 1977 are on the list, as the vast majority of top quality pitchers who began in the last 25 years have started very recently and are still throwing today. The 1980s were an historic lull for top notch pitchers, but a decade-long blip hardly seems sufficient to conclude a trend. I chose "year of final win" because we track full careers with it. Half of the pitchers in the "began after 1977 list" began after 1990, so don't have enough of a track record to predict.

The 16 in the Top 50 include active players Clemens, Maddux, and Glavine. Jack Morris also doesn't become a regular until 1979. So that's four. Randy Johnson should be in the Top 50 soon. And I suspect that Mike Mussina, Kevin Brown, Pedro Martinez, and perhaps even Andy Pettitte are the most likely among those who are active to join them.

So that's 8 or 9, perhaps, among those who BEGAN 1977 to 2002, recognizing that those who began in the last ten years of that era are still plugging along and who knows if Russ Ortiz, Mark Prior, Mike Muler or Roy Halladay isn't about to string together ten consecutive 15+ win seasons and make 1977-2002 one of the BEST quarter-centuries to start throwing.
   41. Chris Cobb Posted: November 12, 2003 at 05:06 AM (#518997)
This responds to jimd's excellent comments above.

I think the distribution of pitching win shares among many more players _will_ have something of the effect that you describe, Jim, but there are two factors that can/will diminish the effect.

First, among starting pitchers, the pitchers who are throwing the best tend to throw the most innings, so that even as larger rotations spread out the value, the top pitchers will still accumulate value in amounts that will measure up nicely against position players, I expect. This concentration of value in the best pitchers seems to me to be one reason not to expect pitchers to make up 35% of the Hall. I don't expect value to be distributed among pitchers in anything like a linear fashion.

Second, if one takes peak into account in one's rankings, one can always adjust average value to reflect actual usage rather than a theoretical average deduced from WS. I'm starting to do that with pitchers now -- I'm adding up all of the innings thrown by the top two starters on each team to find an average number of innings thrown by pitchers working a full load (that'll expand to three or four starters later, I expect). That percentage of total innings can be used to determine a percentage of the average win shares earned by an average starter on an average staff on an average team, with, at this point, a little bit of batting value thrown in to reach a nearby round number. As pitching time is split among more pitchers, average value drops, so peak will remain higher.

But that means voters will need to take pitchers' peak value seriously. I had been looking at pitchers' peaks as so consistently high, among top pitchers, that I felt I should discount it in some way. As I see more, I'm realizing that their career value does not tend to be proportional to the peak, so if pitchers are going to get a fair shake, I shouldn't write their peaks off, just because they tend to be the highest on the board.

Relief pitchers will probably need to be handled somewhat like catchers are being handled now, but I'm going to worry about that when we start having some real relief pitchers show up . . .
   42. Howie Menckel Posted: November 12, 2003 at 02:09 PM (#518998)
hope this helps...

HOMers

CATCHER (1): Buck Ewing (C-1B/0F), see also White, Kelly
   43. Rusty Priske Posted: November 12, 2003 at 02:54 PM (#518999)
For the first time since I started (1900), I don't really want to put anyone in the number one spot. Since leaving it blank and starting at #2 is not an option, here is my prelim ballot:

1. George Van Haltren (2,5,6)
   44. Marc Posted: November 12, 2003 at 03:53 PM (#519000)
Rusty, you raised a question in my mind with your ballot. The question is more of a general theoretical one than directed specifically at you or your ballot. But I notice that other than Frank Grant, who is a subjective choice you might say, your top 7 choices are all corner/hitters (with a half an exception for Cal McVey), and you're uncomfortable with any of them being #1. Is there anything in your method that overrates corner/hitters? Maybe there's a pitcher or middle infielder who should be #1 but your methods suppress his value?

The reason I ask is that I struggle with this myself. My own personal HoF (right up to today) is heavy with LFers and RFers. And intuitively, you know what? I think the heavy hitters of history SHOULD be highly represented. Players don't hit as LFers and RFers and 2Bs. They hit as hitters, so maybe the distribution question is the wrong question. But maybe not.
   45. Rusty Priske Posted: November 12, 2003 at 04:22 PM (#519001)
That is a good question, but the one I was thinking of was somewhat different:

There is a point bonus for the players that are in the #1 spot (or #2 on a 2 inductee year, and so on) because we are implicitly lobbying to get those players into the HoM.

But what if there are no players that you wish to lobby for?

I guess that comes back to how many candidates you want to see in the Hall. Some people like bigger halls and some people like smaller.

OTOH, I won't be disappointed if Van Haltren gets in (which he probably won't) or McVey gets in (which he probably will), so I might be overstating my position somewhat.

As for a positional bias, I have always thought of myself as biased towards pitchers, but we ran out of pitchers that deserve to get it, imo.

Oh, in my personal HoM, Van Haltren, McVey, and Ryan have already been inducted.
   46. MattB Posted: November 12, 2003 at 04:55 PM (#519002)
Preliminary Ballot:

1. Harry Stovey -- AA may have been weaker overall, but that doesn't mean it didn't have any stars. McPhee is a start.
   47. Howie Menckel Posted: November 12, 2003 at 04:56 PM (#519003)
I think the positional issue is playing out beautifully.
   48. Paul Wendt Posted: November 12, 2003 at 07:43 PM (#519004)
OCF (#37) and others. Regarding #29, let me repeat #31:

(sigh) That article [#29] is related to the "gap" in prime candidates newly eligible for the HOM in 1924-1932. I composed it for "New Eligibles Year by Year" [not "1914 Ballot Discussion"] and it refers to #58 on that board. I will copy it there.

Discussion of that generation which entered the major leagues early in the 20th century belongs there, not here.
   49. John (You Can Call Me Grandma) Murphy Posted: November 12, 2003 at 08:37 PM (#519005)
When McVey and Bennett get in, we're certainly in reasonable shape at C. Collins and Cross might deserve a shot, and here comes Home Run Baker down the road!

Heinie Groh is going to have a great case a little farther down the road.
   50. Chris Cobb Posted: November 12, 2003 at 08:50 PM (#519006)
The set of prelim ballots that we've seen may not be representative, but I'm a little surprised at the lack of support for Jimmy Collins so far. He was a truly outstanding defensive player, a very good hitter, with a career of decent length and featuring a high peak (1 MVP-calibre year, others close). What's not to like?

Here are a couple of statistics to ponder on Collins: a list of all players to win 4 or more WS "gold hand/glove" awards at second base, third base or shortstop through 1914, and a list of the career OPS+ for each of these players.

Infield Defensive GG

7 McPhee (2B)
   51. OCF Posted: November 12, 2003 at 09:31 PM (#519008)
(PLEASE tell me it's a nickname...)

Here it is, at least the ones who spelled it "Heinie."

Henry Kappel 1887-1889
   52. John (You Can Call Me Grandma) Murphy Posted: November 12, 2003 at 09:34 PM (#519009)
How about Kaiser Wilhelm?
   53. John (You Can Call Me Grandma) Murphy Posted: November 12, 2003 at 09:41 PM (#519010)
Matt:

Just so you know where I am, I'm not working on a quota for pitchers (or any positons for that matter). Welch and McCormick are not that far off my ballot. Of almost all the pitchers available, I like those two the best. I would take them over Griffith and McGinnity fairly easily.

The only exception that might be better (and make my ballot) is Bobby Mathews. I should make a final determination on him this week.
   54. OCF Posted: November 12, 2003 at 09:48 PM (#519011)
Gehrig missed the chance to be "Heinie" by going by his other given name.
   55. Marc Posted: November 12, 2003 at 10:10 PM (#519012)
To all you Jimmy Collins fans.

> 5 Ward (4 SS 1 2B), Latham (3B), Pfeffer (2B), J. Collins (3B)
   56. DanG Posted: November 12, 2003 at 10:16 PM (#519013)
Strange that Collins has such a large edge in Gray Ink over Williamson, 128-73.
   57. Marc Posted: November 12, 2003 at 10:36 PM (#519014)
Well, Ed is aheEd in Black Ink, howsomever.
   58. John (You Can Call Me Grandma) Murphy Posted: November 12, 2003 at 10:46 PM (#519015)
<i>How is it that Big Ed is overrated again?

J. Collins 113
   59. Marc Posted: November 12, 2003 at 10:57 PM (#519016)
John, you always say you don't care about the timeline, but you do appeal to it from time to time, like now.
   60. Chris Cobb Posted: November 12, 2003 at 11:04 PM (#519017)
Strange that Collins has such a large edge in Gray Ink over Williamson, 128-73.

That's largely a quirk of the grey ink system, which reflects old-timey baseball values rather than new-agey sabermetric values. Collins wasn't much for a walk, but he hit for good average and power; Williamson would take a walk, but hit for less avg. and less power. Collins gets 12 points of gray ink for being thrice among the top 10 in BA (though he never was among the top 10 in OBP), Williamson gets 0 points of gray ink for being among five times among the top 10 in OBP.) Here's a full breakdown of how their placements in the top 10 compare, with gray ink points in parentheses.

Cat. -- Collins -- Williamson
   61. Chris Cobb Posted: November 12, 2003 at 11:16 PM (#519018)
Sorry about the long double post -- forgot I hadn't closed the window after I posted, and then re-loaded the page. . . and bad things happened.
   62. John (You Can Call Me Grandma) Murphy Posted: November 12, 2003 at 11:21 PM (#519019)
John, you always say you don't care about the timeline, but you do appeal to it from time to time, like now.

That's definitely not a timeline argument. You're confusing that argument with compensating for inferior competition.

Bill James' timeline argument states that the 19th century guys are really not comparable to contemorary players (which I agree for the most part, BTW). My argument is, for HoM purposes, the top guys from the 19th century can compete with contemporary players as long as you take into account the different competition levels using standard deviation. By doing this, all players are placed on a level playing area.
   63. Marc Posted: November 13, 2003 at 12:15 AM (#519020)
John, you've definitely lost me on that one.

I mean what is the timeline, if it is not a mechanism to create a numerical measure of the inferiority of earlier players? How are these two different things?

It appears to me that you want to level the playing field long enough to derive the numbers that show that Williamson and Collins are roughly comparable. THEN you inject the inferiority (timeline) argument in Collins' favor. How is this different than James' timeline?
   64. ronw Posted: November 13, 2003 at 12:29 AM (#519021)
On raw numbers, without regard to value, we have elected:

P - 23.9%
   65. Marc Posted: November 13, 2003 at 12:44 AM (#519022)
I am inclined to think the lack C and 3B has to do with the difficulties of playing the positions, which will continue to be characteristic pretty much forever. RF is simply a matter of being manned by the weakest link until about 1885 or so. It will catch up.

I've got Charlie Bennett near the top of my ballot, but no 3B. I am still trying to figure out how to move Jimmy Collins up without a blatant appeal to the bulls**t dump.
   66. John (You Can Call Me Grandma) Murphy Posted: November 13, 2003 at 12:55 AM (#519023)
I mean what is the timeline, if it is not a mechanism to create a numerical measure of the inferiority of earlier players? How are these two different things?

Marc, using a timeline will eventually knock all of the players from the 19th century from contention after a certain amount of time. My argument will not allow that.

It appears to me that you want to level the playing field long enough to derive the numbers that show that Williamson and Collins are roughly comparable. THEN you inject the inferiority (timeline) argument in Collins' favor. How is this different than James' timeline?

Marc, during Cobb's and Wagner's time, batters would have league leading BAs 140 points higher than the league average. Today's hitters will lead the league with half that amount because there are many more good players at the bottom end of the batting pool. But if you take a list of normalized BAs, there would still be more players from the beginnings of baseball than later on the list. Does this make sense to you? It doesn't to me.

OPS+ is not as problematic as normalized BAs, but it is still not perfect and needs to be corrected or the earlier players will be overrated.

BTW, I'm not related to Jimmy Collins. I couldn't care less where he ranks other than that he be ranked correctly as possible. I have no bias for or against him as your second-to-last sentence infers.
   67. Marc Posted: November 13, 2003 at 02:02 AM (#519025)
>Marc, using a timeline will eventually knock all of the players from the 19th century from contention after a
   68. OCF Posted: November 13, 2003 at 02:09 AM (#519026)
For me the race this year will be run in several heats. I have a fairly fixed idea of what to do with Stovey, Bennett, and McVey. The heats are as follows:

Heat 1: Kelley and F. Jones versus Ryan, Duffy, Thompson, et al.

My verdict is that Kelley goes right to the front of that line (but only by a little bit) and that Jones falls behind most of them. Not that there's that much difference - the whole pack is tightly bunched.

Heat 2: Collins versus Williamson and Cross (and McGraw, and Nash, with Childs hovering off to the side). On looking at this, I see the order as Collins > Williamson > Cross. I like career (I rated Beckley highly, after all), but Cross has too many years that are just dead weight. Collins > Williamson is partly career and partly that the batting statistics of Collins's time give a more reliable message than those of Williamson's time. But the whole exercise has the effect of making Williamson look good all over again. Before last year, I had been putting Williamson up around 10th or 12th when he'd fallen off most others' ballots. Maybe I was right to have him there - Williamson goes back on my ballot to about the position I had him before.

Heat 3: McGinnity vs. Griffith vs. the 1880's backlog (McCormick, Welch, et al.) I haven't made up my mind yet, so what do the sample ballots above say?
   69. Marc Posted: November 13, 2003 at 02:36 AM (#519027)
OCF, that was a prelim ballot ranking the position players only. There will be pitchers on my ballot.
   70. Chris Cobb Posted: November 13, 2003 at 03:34 AM (#519028)
On the timeline issue:

Marc, I suspect that you and John (if I am interpreting his concerns correctly) are differing in your usage of time-line in this manner. For John (and for myself) a timeline adjustment is that kind of adjustment that Bill James uses in his NBJHBA. On the strength of evidence that the quality of competition has generally improved across baseball history, he creates an element in his ranking system that awards bonuses to players based on their birth year -- the later, the higher. This is a timeline adjustment, because it is calibrated only to the passage of time.

John is arguing, and rightly, I think, that the adjustments he is advocating are not timeline adjustments, because they are not calibrated according to the passage of time, nor are they based on any pre-judgement that later players are better than earlier ones. They are calibrated according to changes in the standard deviations (SD) of player performance. It is true for the most part that the SD in player values has declined over time, suggesting that level of competition has improved. It is this general trend that provides some justification for using a timeline of the Jamesian sort. But if John looks at the SD scores, and makes adjustments purely following them, then he is adjusting for level of competition, and not according to a timeline. If the SD gets larger at some points later in baseball history, John would not award bonuses to the later players (as a timeline adjustment would), simply because they were later in time. He would adjust the values of the later players downwards because the quality of competition has dropped. Therefore, he is not using a timeline adjustment.

But when we're using WS or WARP...value is value, wins are wins, right? How is that analogous to BA and OPS and normalized BA and OPS+?

WS and WARP are ultimately derived from raw stats like batting average, so they are subject, potentially, to all the same vagaries as the raw stats are. The process of calculating WS, like the process of generating normalized BA and OPS+, includes steps that eliminate the effects of inequalities in playing conditions that make raw stats somewhat unreliable guides to quality of performance.

In converting player statistics into wins, WS is further normalizing values so that inequalities in playing conditions that obtain between different seasons, based on things like the relative balance between hitting and pitching, are smoothed out. We can (fairly) reliably compare Fielder Jones's best seasons during the heart of the dead ball era to Hugh Duffy's best seasons in the mid-1890s.

There is yet one further step in a process of normalization that WS does not take, but that the WARP system attempts in WARP2: correcting for imbalances in the quality of competition between seasons. It says that the "a win is a win is a win" position is not entirely true. The merit, though not the pennant value, of a win changes if it becomes harder to obtain a win, and that increased value translates back into all the individual aspects of player performance that contribute to creating a win.
   71. Jeff M Posted: November 13, 2003 at 03:36 AM (#519029)
How does Win Shares see Joe Kelley's defense?

Very good. Gets an "A-". However, it only shows him winning two gold gloves.
   72. Chris Cobb Posted: November 13, 2003 at 03:58 AM (#519030)
TomH wrote:

How does Win Shares see Joe Kelley's defense?

As an outfielder, it sees his defense as excellent. He earned 3.57 WS/1000 innings, which gets him an A- letter grade. He's better than Van Haltren, who earns 2.90 WS/1000 (B) and Ryan, 3.15WS/1000 (B+), but behind Hugh Duffy (3.99 A+) and Mike Griffin (3.74 A). However, he also put in significant time at first base later in his career, which probably brings his overall defensive value down a bit.

But his WS per year (24.3 for whole career) are surprisingly low.

How do you derive this number? The NBJHBA puts his career WS/162 games at 26.82. Even that is not a fabulous number, but Kelley put in a lot of unspectacular seasons in the second half of his career. He was a truly great player from 1893 to 1900, but after that he was, overall, average at best, as WS sees the matter. He wasn't playing at a spectacular level, and he wasn't playing full seasons, missing an average of 36 games a year from 1901 to 1906.
   73. MattB Posted: November 13, 2003 at 04:01 AM (#519031)
Numbers Here! Get your 19th Century Negro League Numbers While They're Hot!

This is from the appendix of the most recent edition of Sol White's History of Colored Base Ball. I don't know how these numbers were derived, or where they came from exactly year-wise or what they completely represent. But for those who wanted "just a little more" about Frank Grant, this might do it.

I apologize in advance for the formatting.

The chart is identified as "Compiled by Bob Davids with the assistance of members of the Society for American Baseball Research (SABR), this originally appeared in the Bud Fowler Memorial Project."

I am only including offensive numbers for players for whom more than 2 years of data were compiled. George Williams is listed first, with a .368 average over 2 years of data. Arthur Thomas is listed third, with a .344 over 2 year. Neither, I believe, had very long careers.

Sol White -- Pos. 2/3 -- 5 Yrs of Data -- 152G -- 644AB --169R -- 231H --40(2B) --12(3B) -- 7HR -- 41SB -- .359BA

Frank Grant -- Pos. 2/3 -- 6 Yrs. -- 458G -- 1879AB -- 410R -- 634H -- 123(2B) -- 34(3B) -- 31HR -- 149SB -- .337BA

Bud Fowler -- Pos. 2/P -- 10 Yrs. -- 465G -- 2039 AB -- 455R -- 628H -- 112(2B) -- 38(3B) -- 7HR -- 190SB -- .308BA

Clarence Williams -- C/OF -- 3 Yrs. -- 102G -- 420AB -- 98R -- 126H -- 17(2B)-- 3(3B) -- 1 HR -- 41 SB -- .300BA

Richard Johnson -- C/OF -- 4 Yrs. -- 337G -- 1471AB -- 297R -- 398H -- 63(2B) -- 29(3B) -- 14 HR -- 130 SB -- .271BA

George Stovey -- P/OF -- 6 Yrs. -- 122G -- 464AB -- 68R -- 121H -- 16(2B) -- 3(3B) -- 1 HR -- 28SB -- .261 (also 60-40 record as pitcher)

Jack Frye -- 1B/C -- 5 Yrs. -- 124 G -- 446AB -- 87R -- 113H -- 23(2B) -- 7(3B) -- 3 HR -- 27 SB -- .253BA

Moses Walker -- C/OF -- 5 Yrs. -- 354G -- 1295AB -- 215R -- 293 H -- 15 (2B) -- 4 (3B) -- 100 SB -- .226BA

For those who really want it, I have two years of data for George Williams, Arthur Thomas, Abe Harrison, William Selden (hitting & pitching), William Whyte (pitching) and Robert Higgins (pitching).
   74. Chris Cobb Posted: November 13, 2003 at 04:06 AM (#519032)
TomH. wrote:

<i>On another subject, I cannot see how Beckley finished ahead of Van Haltren. Per AB, Van Haltren is clearly a bit ahead on offense, no? Beckley's career is slightly longer, but not a huge thing. VanH played CF, Beckley 1B.
   75. OCF Posted: November 13, 2003 at 04:08 AM (#519033)
Marc - I'm sorry. I was just doing a quick scan of this thread to look at things that looked like sample ballots, and I didn't stop to read your intro to #16, which plainly stated what it was.
   76. MattB Posted: November 13, 2003 at 04:10 AM (#519034)
I have only glanced through the book so far, but what is most striking to me are the numerous illustrations (dozens) provided by Sol White. The captions are written by White, and are very matter of fact.

"Andrew (Rube) Foster, pitcher, Philadelphia Giants, 1904 to 1907" is below the picture of the burly Foster. "Grant (Home Run) Johnson, captain and shortstop, Royal Giants" under a picture of Johnson.

On page 103 is a picture of Frank Grant. The caption reads: "Frank Grant, second baseman, for Buffalo, International League Team, 1888. The greatest base ball player of his age."

Now, I had heard that quote about Grant before, but I had always assumed that, to a certain degree, White was acting as a booster for Negro League players in general. What strikes you looking at the actual book is that Grant is the ONLY player who receives praise in his caption. Or rather, it isn't praise at all, but simply a fact.
   77. OCF Posted: November 13, 2003 at 04:20 AM (#519035)
MattB -
   78. Marc Posted: November 13, 2003 at 04:46 AM (#519036)
Chris, thanks that was great. John, I'm sorry I doubted you!

But...next question. How does a mortal human being such as myself have a clue what to do with WS or WARP or any other measure to adjust for changes in the competitive environment. If Williamson and Collins are roughly comparable in WARP and WS and OPS and etc. etc... well, I can draw an inference from Bill James that, well, of course, Collins would be better, right? But how do I demonstrate to myself anything more than that?

And moving from the specific to the general...is another way of saying this: Competition is improving. The average player is better, the replacement player is better. (We infer this from the population size, but as you say "we" can also demonstrate it from the SDs among players in a given time.) But it does NOT follow that a star player of 1920 is any better than a star player of 1890, since they are both outliers in an statistical universe anyway. Is that the theory?
   79. MattB Posted: November 13, 2003 at 05:09 AM (#519037)
Well, I would assume that about half of the statistics on Grant came from his International League days, which are before 1894. On the other hand, I have no idea whether black ball tracked white ball in run scoring, so that may or may not be relevant.
   80. Chris Cobb Posted: November 13, 2003 at 05:17 AM (#519038)
Chris, thanks that was great. John, I'm sorry I doubted you!

You're welcome! I hope I have represented John fairly -- he'll tell us if I haven't, I'm sure!

But...next question. How does a mortal human being such as myself have a clue what to do with WS or WARP or any other measure to adjust for changes in the competitive environment. If Williamson and Collins are roughly comparable in WARP and WS and OPS and etc. etc... well, I can draw an inference from Bill James that, well, of course, Collins would be better, right? But how do I demonstrate to myself anything more than that?

This is a question for the real statisticians among us, and those with comprehensive league data in spreadsheets, and I would be as interested in their answers to this question as you. I might be able to explain the theory, but I can't run the studies or even explain how to set them up competently. I'm a professional when it comes to semantics, but an amateur when it comes to statistics . . .

And moving from the specific to the general...is another way of saying this: Competition is improving. The average player is better, the replacement player is better. (We infer this from the population size, but as you say "we" can also demonstrate it from the SDs among players in a given time.) But it does NOT follow that a star player of 1920 is any better than a star player of 1890, since they are both outliers in an statistical universe anyway. Is that the theory?

Yes. As I see it, the fact that the average player is better and the replacement player is better in 1920 than in 1890 does not mean that a star player in 1920 is better than a star player in 1890, because they are both outliers. There are probably HoM voters who disagree with this statement, but many (more?) would agree with it.

I would add only that the refinement of playing tactics and strategy is a third element from which we can infer improvement in competition. As the rules stabilize and players and managers learn from experience, sub-optimal strategies and styles of play are gradually discarded and better ones adopted. Once we stop hearing stories about players inventing the bunt, the curveball, the change-up, or deciding to play off the bag at first, we are entering a more difficult competitive environment. Innovation will always provide an edge, but there's less room for innovation, and fewer players who are way behind the curve in their approach to the game. This sort of improvement in competition isn't something that's directly quantifiable in comprehensive metrics (that's the advantage of looking at SDs), but it shows up in stuff like fielding stats: fewer errors, more double plays, etc.
   81. MattB Posted: November 13, 2003 at 05:53 AM (#519040)
yest:

Some food for thought. Bob Caruthers v. Joe McGinnity.

Record
   82. Chris Cobb Posted: November 13, 2003 at 06:47 AM (#519041)
This comparison was prompted by yest's prelim ballot. I was puzzled by Sam Thompson appearing at #1 and Joe Kelley not making the top 15. Now, I don't know what yest's ranking criteria are, but it sure looks to me like Joe Kelley basically had Sam Thompson's career, and then had some more years of average play in addition. This can be seen if one compares their 10-year primes. For Thompson, that's 86-95, and he has only two other seasons in which he plays any amt. (85 and 96). For Kelley, that's 93-02, and he has two partial seasons preceding that and a decline phase that lasts from 03-07. He's still an above average hitter in most of these latter seasons by OPS+, but he's nothing special.

For these seasons, I calculated their average OPS+ by multiplying their OPS+ for each season by their plate appearances, summing the totals, then dividing by total plate appearances (serious stats folks please point out any flaws in this approach).

Kelley 5680 PA, 143.6 OPS+
   83. John (You Can Call Me Grandma) Murphy Posted: November 13, 2003 at 06:48 AM (#519042)
Chris, thanks that was great. John, I'm sorry I doubted you!

No problem. I just wish Chris had typed my other earlier posts because he seems to be able to articulate my points better than I do! :-)

Believe me, if there is an enemy of the timeline here for this project, it's me.
   84. Brian H Posted: November 13, 2003 at 07:28 AM (#519044)
Now it seems to me that Bob Carruthers was not merely a lucky freeloader on tnese good teams. He was one of the primary reasons that they were good. In considering the contributions that great players make to Pennants James' recently revised PB version of the Historical Abstract introduced the statistic of Pennant Impact. According to James calculation Caruthers was second all-time behind only Kid Nichols.

With this in mind Carruthers was not just fortuitously "on some great teams", he made some good teams great.

I would add that STATS places Caruthers on 5 all-star teams with 2 Cy Youngs and 2 MVPs. The Ironman makes 4 all-star teams and gets 2 Cy Youngs (no MVPs).
   85. Jeff M Posted: November 13, 2003 at 09:16 AM (#519046)
Now, I have McGinnity on my ballot, too, so I'm not criticizing anyone for having him on. But if you do, please look at Caruthers again and put him at least a little bit higher than McGinnity!

I'm not denouncing the above argument. I do, however, want to point out that while Caruthers has better strikeouts per nine IP and better walks per nine IP, those have to viewed in league context. In those categories, Caruthers was less frequently in the top ten than McGinnity.

Also, if you look at the NRA/DERA comparison in the BP numbers, it shows that Caruthers benefited significantly from pitching in front of an above-average defense.
   86. MattB Posted: November 13, 2003 at 02:53 PM (#519047)
Caruthers was in the Top 10 in strikeouts twice (6th and 8th), which McGinnity was in the Top 10 four times (2nd, 5th, 6th, and 9th). Neither made the Top 10 in K/9. McGinnity was a workhorse, and several of those Top 10 finishes put him over players with significantly better K rates who simply did not pitch as many innings. McGinnity may have an overall strikeout advantage, but it is negligible at best, and I would put it entirely in the longevity category. Neither makes my HoM due to their strikeout abilities!

In terms of walks, Caruthers was in the Top 10 in BB/9 six times (a first, two seconds, two thirds, and a sixth). McGinnity has seven appearances, but they were much lower in the Top 10 (a fourth, two fifth, two sixth, an eight, and 1902 -- where he got ninth in one league and tenth in the other. I only count 1902 as one Top 10 finish!)

Caruthers was the better player at preventing walks, even taking into account their eras.

"Also, if you look at the NRA/DERA comparison in the BP numbers, it shows that Caruthers benefited significantly from pitching in front of an above-average defense."

I agree with this, but after adjusting for defense (the "DERA"), the career numbers are nearly identical (3.78 for McGinnity and 3.83 for Caruthers). This adjustment shows that properly adjusted for defense, Caruthers and McGinnity were about equal as pitchers -- Caruthers doesn't blow him out of the water.

Now add in hitting (the best ever hitting pitcher versus the worst ever!) and the decision is clear to me. They are about the same on the mound. Do you prefer McGinnity's longevity or Caruther's hitting? I think Caruther's hitting was much more valuable to his teams.
   87. DanG Posted: November 13, 2003 at 03:26 PM (#519048)
The major differences between Caruthers' and McGinnity's pitching value is longevity and durbility. Joe pitched two seasons more than Bob and led the league in IP four times. Caruthers never came close to leading the league in IP.

This may be enough to make up for Caruthers' edge in hitting.

At this point, small differences in quality can make a huge difference in where players rank.
   88. MattB Posted: November 13, 2003 at 05:10 PM (#519049)
I don't really think it is that close.

Consider:

Pitching Runs Above Replacement

Caruthers: 595
   89. Marc Posted: November 13, 2003 at 05:18 PM (#519050)
Back to the timeline discussion and I use the term "timeline" broadly to refer to a formal method like James' but also anything that "considers" that earlier players faced weaker competition.

The other difficulty here, as I see it, is: How much consideration can you give to the level of competition before you are basically abandoning the measurement of value in favor of "ability"? It seems that a whole series of adjustments, after while, after while will add up to something (a numerical score, whatever) that is not just different in degree but in kind.
   90. MattB Posted: November 13, 2003 at 05:35 PM (#519051)
"How much consideration can you give to the level of competition before you are basically abandoning the measurement of value in favor of "ability"?"

I think if you do it right, you ARE accurately measuring value instead of ability.

Consider it like this:

Assuming OPS+ is the best way to determine offensive value, how valuable is a 140 OPS+?

In 1880 in the NL, 9 players had an OPS+ over 140 (Gore, Connor, Dalrymple, Anson, C. Jones, both O'Rourkes, Hines, and Dunlap.)

In 1980 in the NL, only 5 players had an OPS+ over 140. (Mike Schmidt, Jack Clark, Keith Hernandez, Cesar Cedeno, and Ted Simmons).

Where the 1880s stars better? Probably not. More likely, the average player was worse, so it was easier to be 40% better than that.

Was Dunlap (9th best OPS+ of 141) as "valuable" as Ted Simmons (5th best OPS+ of 141)? No.

Was George Gore (1st best OPS+ of 183) as "valuable" as Mike Schmidt (1st best OPS+ of 170)? Probably they were.

A "timeline" would say that Gore was worse by whatever the timeline adjustment is. A proper adjustment shows that a 140 OPS+ just isn't as valuable if everyone else is doing it.

That's why I like to look at where a player ranked alongside his peers before looking at intergenerational rankings.
   91. ronw Posted: November 13, 2003 at 05:55 PM (#519052)
Current HOM: If McVey Elected

P - 23.9% P: - 23.2%
   92. Howie Menckel Posted: November 13, 2003 at 06:05 PM (#519053)
I've been in and out on Caruthers and the HOM, and the McGinnity comparison is interesting.
   93. DanG Posted: November 13, 2003 at 06:31 PM (#519054)
Maybe someone could do a head-to-head comparison of Duffy and Kelley.

In the past four elections, Duffy has finished just behind the duo of Stovey & Bennett. These two are on track for eventual election, behind them is The Glut, which Duffy is at the head of.

The comparisons of newbies so far has shown Collins to be very close to Williamson (who only one voter among us sees as a HoMer) and McGinnity close to Caruthers (who has about seven friends). Kelley seems like the only newbie who may rise above The Glut.
   94. Chris Cobb Posted: November 13, 2003 at 06:54 PM (#519055)
I wouldn't be so quick to write off Collins as someone who should rise above the glut. The fact that he's close to Williamson, but a bit better at the plate, in the field, and in career length, does not necessarily mean that he's not any better than the outfielders that most voters have been rating ahead of Williamson. Given how closely packed the candidates for election are, it could mean that Collins is above the glut as well. Or it may be that Williamson has not been looked at seriously in a while by a number of voters, and deserves to be ranked higher than he has been in comparison to the current slate of eligibles.

Some head-to-head comparisons of the top infielders to the outfield glut ought to be made, and there will surely be more debate about issues of career length and infield play.
   95. DanG Posted: November 13, 2003 at 07:07 PM (#519056)
Chris, I totally agree. I was just giving the weather report, seeing how the winds were blowing.

Williamson has finished 27th in our past two elections. I think we may see more support for him this election, but it's an awfully long way from 27th place to the Hall of Merit. As I wrote, Collins needs to show he's not Ed Williamson.

On another topic...

With McVey?s election imminent, that will bring the total of HoMers on our first ballot in 1898 to twelve. That?s the backlog we started with, more than I thought there was at the time. It says to me we ought to have started earlier.

Back in 1905 I posted this: ?If we had decided to start the HOM eight years earlier, in 1890, and elect one player each year, this is whom we would have elected: 1890-Barnes, 1891-Wright, 1892-Start (1st-ballot), 1893-Spalding, 1894-Sutton (1st-ballot), 1895-McVey, 1896-White (1st-ballot), 1897-Hines (1st-ballot), 1898-Gore (1st-ballot), 1899-O?Rourke (1st-ballot), 1900-Clarkson (1st-ballot), 1901-Kelly, 1902-Brouthers (1st-ballot), 1903-Connor (1st-ballot), 1904-Anson plus Ewing bring us to our current count of 16 HoMers.?

Bringing this list up to date: 1905-Ward, 1906-Keefe, 1907-Hamilton (1st-ballot), 1908-Glasscock and Rusie, 1909-Delahanty (1st-ballot), 1910-Radbourn, 1911-Nichols (1st-ballot), 1912-Burkett and Richardson, 1913-Galvin. It?s nearly identical to our HOM so far, except we have elected McPhee when McVey should be in.

OK, so it?s no big deal, we?ll end up electing the same players, just in a different order. Wright would?ve been the 2nd HoMer instead of the 10th, Start the 3rd, not the 26th, Spalding the 4th, not the 19th, Sutton the 5th, not the 21st, McVey the 6th, not the 28th.

Where we would?ve really benefited is by having fewer first ballot electees. Kelly, Ward, Ewing and Rusie would?ve been subjected to greater scrutiny, always a good thing.
   96. Marc Posted: November 13, 2003 at 07:26 PM (#519057)
I've emerged as a de facto FOEW and apparently EOJC, and yet Big Ed has never ranked among my top 10. But to Dan's point, a survey of the available numbers--both raw counting stats, and WS and WARP--has Ed and Jimmy very close. I would have to say that this is one of the more pleasant aspects of this project--that is, to suddenly wake up and say, hey, wait a minute, are you sure we've given X a fair shake? A new perspective is a good thing to get.

I have always been an admirer of Jimmy Collins. I knew that his raw career WS totals aren't all that imposing, yet before looking at the numbers I assumed that there would be something in there that would jump off the page, followed by the head smack and the "Doh!" and then slotting him in somewhere around 3rd to 5th on my ballot. I am shocked that it didn't happen.

And so in the absence of any compelling evidence that Jimmy is NOT basically a younger version of Ed, the answer has been that, well, Jimmy faced better competition. But I have not seen an adj WS or WARP number that somehow factors in the SD and pops out a number. So there is still a quasi-objective (but semi-subjective) operation needed to separate the two.

Ed and Jimmy aside, THAT really qualifies as a new perspective!

And as a further aside, as others have noted, comparing a pitcher from a two-man rotation (Jim McCormick, best of the holdovers) to a "modern" pitcher from a 3 or 4-man rotation (Iron Man Joe McSomebody) is really tough! I agree, though, that how he compared to his peers (in SDs???) is the most common element to use. Comparing WS and WARP is almost as unenlightening as comparing wins and ERA.
   97. ronw Posted: November 13, 2003 at 07:30 PM (#519058)
Interesting because the posts are in proximity, with McVey in this year, 12 of your top 13 in 1898 will have been elected. The only missing member: Ned Williamson #10 on the 1898 ballot.
   98. Paul Wendt Posted: November 13, 2003 at 07:33 PM (#519059)
tidbits

1. Heinie
   99. ronw Posted: November 13, 2003 at 08:01 PM (#519060)
5. HOM-cumulative games played at each fielding position (by Ron Wargo)
   100. Chris Cobb Posted: November 13, 2003 at 08:24 PM (#519061)
Quick answers to tidbits pertaining to my analysis:

?<i>tidbits

2. Gray Ink
Page 1 of 3 pages  1 2 3 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Phil Birnbaum
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.6518 seconds
49 querie(s) executed