From DanG:

***1912 (October 26)-- elect 2

I have Griffith somewhere between 20 and 25. Dont have an exact rank at that point. He is either 2nd or 3rd amongst currently eligible pitchers.

I don't have the new guys slotted in yet, but I think Clark Griffith has been underrated historically. I think there's been a tendency to underrate the so-called "pioneers" and "executives." We tend to think guys like Connie Mack, Charlie Comiskey, Clark Griffith et al must be OVERrated as players and so therefore we discount (underrate) them as compensation. Well, every case is different and Mack wasn't much of a player, Comiskey is still overrated. John McGraw has been underrated in this way (possibly even by me), though I would not say the same of Wilbert Robinson. But, in any event, Griffith could pitch. He may not make my ballot but he deserves to get considered and evaluated.

Ignore that last line from post #1.

McGuire is the best choice from this crop, IMO, but not a slam dunk guy by any means. He'll probably be somewhere on the second half of my ballot. Burkett and Start should be the top choices this "year."

I'm looking at Tiernan and Thompson more closely and may find prominent positions on my ballot. I have been going by the logic that right field, left field and center field are separate positions and have been comparing each outfielder to starting players from those respective positions. However, when it comes to a positional career adjustment, I have been lumping them all together. I think this is probably wrong. I'll go into more detail later on this or next week.

Unlike many who have an idea that pitchers should theoretically fill about 20% of the HoM, I was thinking a more appropriate percentage would be about a third. So, also unlike many who are concerned about an overrepresentation of pitchers, I am concerned about an underrepresentation.

If there was some sort of concensus about which of Welch, McCormick, Mullane, Griffith, and Caruthers was the best pitcher, that pitcher would move up into the top 10 (instead of all hovering down below 15). I'd like to extend the pitcher discussion from just focussing on Griffith to which of the five is most worthy and why.

<i>Games played as leadoff batter, NL 1901

My early prelim...

(I don't think I'm giving away any surprise about who wins for 1911...)

1. Jesse Burkett (3,-,-)

I like to stay a little conservative on new guys, so I dropped him in below Start last week. I'm convinced that he is actually better.

2. Joe Start (2,2,3)

Eventually.

3. Bid McPhee (4,3,5)

I'm not a huge supporter of Bid, but below Start there really isn't anyone I am excited about.

4. Cal McVey (5,4,6)

Below here, I don't know if I really support anyone for HoM right now. That could change (as it has a few times already.) :)

5. George Van Haltren (6,6,4)

Good point, MattB. Griffith deserves consideration but he probably is not the best eligible pitcher. I still like McCormick until I am convinced otherwise.

We elect TWO this year! I assume last year's results, not yet published as I write this, had Nichols and Burkett on top. It is less obvious (not having counted the votes myself) whether Start held the next spot vs. McPhee and McVey or whether there were some changes. But assuming no changes, then I agree with John, Burkett and Start would be the faves. But keeping the Pud Galvin story in mind, last year's results won't necessarily predict the other winner this year.

The upcoming dearth of top-notch newly-eligibles (1924-32 if I recall?) makes these elections important not just for who wins and loses, but for establishing a pecking order. Despite the occasional leapfrog, the order has been fairly stable. So if Start, McPhee, McVey and Bennett, say, are going to go in over the next decade or so, the question becomes what about the OF glut. Can anybody shed any light that will reshuffle that pecking order? How about Frank Grant? Is anybody going to discover some new information that adds power to his '90s record? And how about those pitchers, and Hughie Jennings? 100 years from now we won't remember exactly who got elected when and so the top 5-6-7 positions maybe won't matter so much. But the 8-9-10 to 15 positions will matter a lot because a couple of them will get elected and the others never will.

2 this year? Great, that means one of the hold overs goes in for sure. That is good news, because ballot stagnation is never good... :)

Prelim. Not to tough this week since none of the new candidates are particularly inspiring:

Jim,

His 1898 season (1.88 ERA, 190 ERA+, 325 IP) is a great peak year. For his career, his ERA is 21% better than league, park-adjusted. 3386 IP is not a short career, and he was pitching from the modern distance. He's 91 games over .500 for his career. How can he be off the ballot?

Well, his best season ('98) wasn't anything like Amos Rusie's in '94. Nicols was clearly the best pitcher that year. Fine, fine season, but not nearly being one for the ages.

3386 IP for that era wasn't that impressive. As for the modern distance, that means zip to me. Pitching under improved competitive conditions compared to pre-1893 does, however.

To compensate for a fine, but not overwhelming peak, you need Eddie Plank-like career numbers. The Old Fox didn't do that. Being the fourth best pitcher of your time is not too shabby, though.

Here's info I alluded to last "year" on the top three pitchers according to the WARP system, 1871-1912. I've listed everyone with two or more placements in the top 3 according to WARP1, which includes hitting, and according to pitching runs above replacement level. Esp. prior to 1890 there is significant variance between the lists.

Number of Seasons Among Top 3 Pitchers, 1871-1912

According to WARP1

10 Cy Young

Last place Cincinnati might have been better off with Heinie Peitz batting leadoff instead of Dobbs

I said that in post #6, but I just went back and looked at the SB column. Mostly, things blur together into "everyone has 20-25 SB," but it isn't quite everyone. Peitz had so few SB that I've got to conclude that he ran like a catcher - and that's another reason not to bat him leadoff.

In this post I'll compare Griffith to Jim McCormick, the best of the 1876-1890 pitchers still eligible. Later I'll look at Griffith in comparison to George Van Haltren, a lower-down member of the outfield glut.

McCormick and Griffith are very close, by both traditional and advanced metrics.

Traditional

McCormick

I convert WARP into WS-scale numbers for pitchers...

Chris, did you mean this only for NA seasons? That makes sense to me, since WS are not available for NA seasons.

If not -- if you are also taking non-NA seasons and converting WARP to WS for pitchers -- then my question would be "Why not just use WS with some adjustments"? I think it is potentially confusing to other readers to see McCormick has 340 Career Win Shares and Griffith 313 Career Win Shares, or McCormick with 25.57 WS/360 IP and Griffith with 29.03 WS/360 IP, if in fact those aren't really Win Shares at all. I.e., they are WARP on a different scale. Maybe they should have a different name, like adjusted WARP or something. Just a suggestion, because I think we've seen quite a bit of variation between how a player will be ranked under WARP and how he will be ranked by WS.

Since I prefer WS (with pitching/fielding split adjustments) to WARP, I want to make sure I'm not seeing WARP in WS clothing.

Jeff M,

Who will be first to 10,000? Not Start, I suspect.

Career votes-points leaders

I suspect Stovey will get that honor. He'll continue to get support, but it doesn't appear to be enough to get him elected.

Is there an electronic database with unadjusted WS or WARP numbers? I don't have the time to enter everything manually, and all I have found in the past are databases with raw numbers. That would be extremely helpful for comparisons.

BTW, I just got the Win Shares book, and was surprised to learn that according to this system, Jesse Burkett is Tim Raines (389 WS to 390 WS total, 334.2 bat WS to 336.1 bat WS, 52.9 field WS to 52.7 field WS) with a slight bit of pitching thrown in.

Jeff,

I don't use pitching WS with adjustments for two reasons. First, it appears to me that win shares for pitchers are too much affected by team performance, even as late as 1910, for me to trust them fully. Any kind of adjustment that _I_ could make to WS I would have to make globally, and I don't think a global adjustment can fix the team-record problem. Second, I'm convinced that the WARP system does a better job than WS of capturing the gradual shift of defensive value from fielders to pitchers. For _fielding_ value, I adjust fielding WS rather than using WARP's fielding runs because I think WARP's fielding evaluations are seriously flawed, but since WARP's pitching numbers are more reliable, I prefer to their methodical accounting for the change to any patch-work adjustment I could apply.

I make the conversion of pitching value to WS, however, because I use WS for batting and fielding value (for both I find WS more reliable), and I want to have all of my ratings in a single metric. Since I explain where the WS I use for pitchers come from, I don't think I am being misleading in presenting my numbers. All of the WS I use, whether they are adjusted for original WARP or raw but official Bill James Win Shares, have undergone considerable processing in any case, and I try to make clear what the process has been whenever I present them.

Chris,

For some reason, it has always seemed a little clearer to me to look at a pitcher's RA rather than ERA. Mostly it tells the same story, but in a few cases, it puts a slightly different spin on things. For instance, Jim Whitney's RA is less impressive than his ERA in 1883 and 1887. That may be one reason I have Whitney lower than some other people do. Looking it over again for this week, I think I've been a little too harsh on both Mickey Welch and Tony Mullane. I still have Jim McCormick ahead of both of them, but not by much. My order within pitchers:

McCormick

<i> Chris,

Matt B wrote: Unlike many who have an idea that pitchers should theoretically fill about 20% of the HoM, I was thinking a more appropriate percentage would be about a third. So, also unlike many who are concerned about an overrepresentation of pitchers, I am concerned about an underrepresentation.

If there was some sort of concensus about which of Welch, McCormick, Mullane, Griffith, and Caruthers was the best pitcher, that pitcher would move up into the top 10 (instead of all hovering down below 15). I'd like to extend the pitcher discussion from just focussing on Griffith to which of the five is most worthy and why.

Such a consensus would advance the top pitcher candidate, but it probably would take the pitcher only into the top 15, because much of the support that pitchers are receiving is concentrated in a few ballots. Of the 42 ballots cast in 1911, 19 had no pitcher aside from Nichols on the ballot, but 10 of the 23 containing another pitcher contained at least 2 pitchers besides Nichols. Even if all 23 voters supporting a pitching candidate agreed on the same one for their top pitching spot, that'd only bring the pitcher up with Childs and Jennings. The case for these pitchers is going to have to convince the other half of the electorate to start giving pitchers other than 1st-ballot HoMers some support if any of the pitchers currently eligible is going to move towards the top 10.

I think a consensus among those who are giving pitchers credit would help convince others as well, of course, but serious comparative work of pitchers to hitters is also going to have to take place if the electorate's overall evaluation of pitchers is going to change.

OCF(#6) replied concerning NL1901 leadoff batters.

Paul, I put that comment in this thread because it prominently featured Burkett (now under consideration) and included comments on several players that we'll have to deal with before too long, like Kelley, Keeler, Thomas, et al. Unless you have some specific plans, I don't see any reason not to continue this conversation on ballot discussion threads. We're not just trying to pick the best candidates for the HoM, we're also educating ourselves about the history of the game - so the more excuses any of us have to seek out information, the better.

Since I explain where the WS I use for pitchers come from, I don't think I am being misleading in presenting my numbers.

Chris: Wasn't accusing you of being misleading (hope you know that). Just wanted to make sure everyone understood your methodology. Someone who only scans your post and sees WS might not realize that it is really WARP (though someone reading the entire post ought to realize that).

<i>Jeff M,

serious comparative work of pitchers to hitters is also going to have to take place

Agreed. There are at least two different frameworks under which I can see this happening. One is comparing "value" using any or all of the integrated evaluation systems (Win Shares, WARP, TPR, others I don't know about) and electing those with the "best numbers", however weighted. The other is a "quota" approach, determining an appropriate ratio between pitchers and position players and then appealing to "imbalance" when arguing for or against.

The number 20% has been mentioned for an "appropriate ratio", that is 2 pitchers for each 8 position players. But this implies that each team has only two regular pitchers and that the rest of the pitching staff are subs, bench players. We're already into the era where a team needs 4 starting pitchers to go with the 8 regular position players. 33% or a 4 to 8 ratio would seem to be more appropriate for the period from 1904 to 1975. Why would this number be too high?

jimd:

>33% or a 4 to 8 ratio would seem to be more appropriate for the period from 1904

jimd, and to elaborate on this bias or preference:

Make a list of the top 10 position players and the top 10 pitchers for any given period of time. Now, say you've elected 8 position players and 3 pitchers, and now you're going to elect one more player from those lists (the #4 pitcher or the #9 position player). I think the problem for the pitcher is that it is harder to be sure you've got the right pitcher at #4 (McCormick? Griffith? Caruthers?) than that you've got the right position player at #9. The consensus is easier for the position player and with 40+ ballots that's a wrap.

4th starter is not a position. Nobody selects an all-star team and picks the best 4th starter in the league over the 2nd-best first starter (unless all four selections are from one pitching-loaded super-team).

If some every-day positions receive "bonuses" due to degree-of-difficulty in a given rating system, shouldn't pitchers receive the same, based on an expected ratio of 4 pitchers to 8 position players?

Is the 4th pitcher as good as the 9th player?

How does one make that determination absent an IES (integrated evaluation system such as Win Shares, WARP, TPR, others I don't know about)? The traditional metrics for each side don't compare well with each other, and developing conversion factors is similar to creating one's own IES.

Bias is a problem. That doesn't mean it's justified, though. If one argues that catchers/3b-men/whoever do not have enough representation, then I think it's hard to argue against 33% pitchers (from the period 1905-75).

Preliminary ballot for 1912. Significant changes since the 1911 ballot are

EricC, you may be new at this, but this is a nice ballot. I'd be curious, did you identify any specific comps for Wright's and Pearce's decline? Who?

Correction to my previous post- I retract my ranking of Bobby Mathews as the 2nd best available pitcher. This doesn't make sense given his raw numbers. For top 20 players, I've tried to carefully double-check my ratings, but not for those lower down, and I don't want to lose credibility here. :-)

Eric, since there's been a good deal of talk about pitchers vs. hitters, how do you weigh their relative value? I'm esp. curious since you find your principles and your practice not in perfect alignment.

Also, your ballot favors players from before 1880 quite heavily, but it doesn't include any pitchers from an era in which we have so far elected only one pitcher. Not that I'm personally advocating for those pitchers, but if you're going to bring Levi Meyerle onto your ballot, why not Mathews or Bond?

re: pitchers

If the day came that each team would be utilizing a 10-man rotation, would we have to start electing that many more pitchers?

I believe there are more good players as the decades roll on by, so it makes sense to have a few more pitchers per era later on in the project. However, I don't see the logic of having at least 4 per decade (on average). If there are four equally good candidates, then fine. But the holdovers (and Griffiths) were all fine pitchers, but they weren't the big men on campus.

I'll stick with 20% pitchers for now.

I hear you, John Murphy! :-)

I'm always glad when someone does. :-)

<i> The post-1870 record for H. Wright looks a little better

1. (3) Jesse Burkett - He vaults ahead of McPhee based on hitting dominance during the 1890s.

2. (2) Bid McPhee - He was a solid, solid player. Aside from the 2nd base pre 1920 is 3rd base post 1920, he may be Willie Randolph - perennial All-Star, but never quite the best second baseman in the game.

3. (4) Joe Start - I predict that Joe and Jesse will make the HOM this year. I guess Joe fits my ideal long career profile since we have no idea just how good his actual peak was.

4. (5) Cal McVey - I really don't see any bad years out of this guy. He may eventually move ahead of McPhee when I take a closer look at whether I should be giving more credit to higher peak values.

5. (6) Charlie Bennett - If we don't elect him, there will be only two catchers for a long time to come. Fortunately, I think he will someday have a plaque in the HOM.

6. (13) Hugh Duffy - The biggest beneficiary of a recent career evaluation. WS sees him as a fine fielder, and his WARP numbers are outstanding all around. I hereby shed my anti-peak bias, and promise to look carefully at peak values.

7. (7) Harry Stovey - Consistency is key, and he had it for a long-enough career for eventual election.

8. (9) Frank Grant - He seems to be moving down as I closely examine the remaining available holdovers, but that is probably because of the lack of information.

9. (10) Jimmy Ryan - My outfield glut is moving over the pitcher "glut" this preliminary ballot. I am just not sure of either group's eventual election, although the '24-'31 ballots will be interesting.

10. (11) George Van Haltren - See Jimmy Ryan.

11. (14) Sam Thompson - I keep thinking we are underrating him, but the WS numbers are so strongly against him that he can't seem to rise above the fray. I'm looking forward to the recalculation of the BP WARP numbers to see if Thompson stays high.

12. (8) Jim McCormick - The biggest drop. In reexamining his career, I saw seven solid years, and about three All-Star type seasons. When compared to the ten + solid years of some of the outfielders, and the 5+ All-Star seasons of Duffy, I don't think that the pitchers jump ahead of the outfield glut.

13. (12) Mickey Welch - Again, maybe seven solid years, here two possible All-Star type seasons. The Pud Galvin factor makes me want to elect Welch, but I feel that I shouldn't compare available players to those already in, but rather to their actual competitors.

14. (--) Clark Griffith - He had maybe six solid years, one or possibly two All-Star seasons. I don't see him getting elected. Very similar relative to Mickey Welch.

15. (--) Mike Tiernan - He replaces Mike Griffin this year. I don't think either belong.

1912 Preliminary Ballot

Mostly moving players up this week, since the top new arrival places 15th. Some changes in the second half of the ballot pursuant to 1890s infielders and pitchers. That's where the comments are.

1) Jesse Burkett (n/e) (2) With Nichols elected Burkett rises to the top.

If the day came that each team would be utilizing a 10-man rotation, would we have to start electing that many more pitchers?

That depends. Read further.

I'll stick with 20% pitchers for now.

I'm trying to examine the rationale behind assertions such as 20%.

The usage pattern of 4 starters was present for a long period of time (the lifetime of the 154 game schedule, and on into the 162). 33% pitching also is one of the underlying tenets of Win Shares; IIRC, James doesn't explain why, so it might just follow from 4 starters, or OTOH, it may result from experiment or some unshared insight.

Except that James, for his top 100 list in the NBHA, carries two pitchers for every eight position players (which is 20% coincidentally).

But say we abolished the four-man rotation and could only have a one-man tomorrow. Would we still have to elect four pitchers per decade? At the most, the 1880s had up to a 2-man rotation, yet we have already elected Galvin, Radbourn, Clarkson, and Keefe (part of Ward, too). Yet, we still have Welch, McCormick, Whitney and whoever else from that era on some ballots still. They couldn't all be great, could they? I'm asking: I do not have definite views on this.

<i>I'm trying to examine the rationale behind assertions such as 20%.

Does anyone know, offhannd, the % of pitchers in the HoF?

Since he's probably going in this year, here's the Jesse Burkett bio as written by the West Virginia Hall of Fame for his induction in 1982. Couple of interesting tidbits I didn't know.

<I>As a boy, he loved to swim in the Ohio River and he learned the game of baseball on Wheeling Island, the neighborhood of his formative years. As a man, he stood only 5 ft. 8 but he became one of the true giants of his sport. The first West Virginia native to be elected to the National Baseball Hall of Fame at Cooperstown, N. Y., was Jesse Cail Burkett, a native of Wheeling. He was one of the game's great players in the pioneer days of professional baseball.

Only three men in history ever batted over .400 on three different occasions in the major leagues. They were Ty Cobb, Rogers Hornsby and Jesse Burkett. Batting champion of the National League three times, Burkett compiled a lifetime average of .342. A left-handed, line-drive hitter who was also recognized as an outstanding bunter, he enjoyed a spectacular season with the Cleveland Spiders in 1895 when he his .423 to lead the league. He also led in base hits, with 235, while scoring 149 runs and stealing 47 bases, and paced Cleveland to the Temple Cup, symbolic then of the championship of the sport.

The next year, Burkett, nicknamed "The Crab", batted .410. After being traded to St. Louis, he hit .402 in 1899. In all, Burkett spent 16 seasons in the majors. He fell just 128 hits shy of the 3,000-hit total, which has be attained by only 11 men.

Born in Wheeling on Dec. 4, 1868, Burkett was influenced by the fact that the city was baseball-conscious. He developed rapidly as a player and turned professional while still in his teens. In 1888, he played for Scranton, Pa., in the Central League. He was a pitcher then and won 27 games in his rookie year. The next season he advanced to Worcester, Mass., of the Atlantic Association and compiled a brilliant 39-6 record while batting .280 as a second baseman -- all for $125 a month.

Worcester became the "second city" of his life and he adopted it as his home. He lived there until his death on May 27, 1953. Burkett broke into the majors with the New York Giants in 1890. He went on to play with Cleveland, the St. Louis Cardinals, St. Louis Browns, and Boston Red Sox. Later, he became owner, manager and outfielder of the Worcester team of the New England League and won the league batting title. His team won four successive pennants from 1906 to 1909.

But say we abolished the four-man rotation and could only have a one-man tomorrow.

There are different approaches to this problem, and I'm trying to sort through what seems to be best.

If we went to football-style baseball with one game a week (saturday say), each team would carry one ace starter, and a couple of guys who could start/setup/close. The league MVP candidates would be guys like Randy, Pedro, Maddux, etc. Bonds couldn't accumulate enough at-bats in 26 games to approach their value under Win Shares; but then again, pitching replacement level might be much higher, so maybe Bonds would catch up under WARP. I don't know.

One starter resembles the practice of the 1870's. Roster construction/usage patterns from then would indicate an 8:1 ratio. But this assumes that the other players are as important as the pitcher, which may not be true then either. (And I have previously argued that the pitcher was more important.) And if the pitcher was more important, then should relatively more pitchers be elected to reflect that?

Common practice from about 1885-1920 would be for a team to carry two (or more) starting catchers. Should we elect two catchers for each 1b-man? Or is it the case that the catching position is accumulating the same value as the 1b position, and the two catchers are just splitting that value? Does this then justify electing no catchers because they couldn't accumulate the same value as the players at other positions?

Modern day teams carry about 20 regular players: 8 every-day players plus 3 regular "backups" (catcher, infield, of/pinch/designated hitter), and 9 pitchers (5 starters, 4 relievers). (And a host of backups and tryouts rotate through the other 5 roster spots.) Is this indicative of the ratios for how we should be electing players from the 1990's? Or is the value distributed unevenly amongst those regulars so that pitchers will have an even tougher time getting elected?

I don't have the answers to these questions. I'm just trying to stimulate discussion on this topic, because it may influence the shape of the HOM going forward.

John, I've read many of your arguments in favor of players who have not accumulated the same counting stats but play at positions which appear to be disadvantaged during a particular era with respect to playing time. The roster argument in favor of 33% pitchers from 1905-1970 seems to me to be in synch with this.

Does anyone know, offhannd, the % of pitchers in the HoF?

Not that it really matters to a theoretical debate, but the BBWAA has elected about 33% pitchers; the VC about 28% from MLB to the HOF. A 20% argument implies that they've both elected way too many pitchers. Maybe they have, but at the moment, I don't think so.

Al,

We're not developing a 90's pitchers backlog, so the backlog we have - McCormick, Welch, Caruthers, Whitney, Mullane - is held over from the 80's. There does seem to be a shortage of good careers among pitchers who were rookies between about 1890 and about 1899 - Griffith seems to be about the best we have. Probably that's because the high-offense times ground up pitchers and made it hard for them to establish themselves. The 90's are repesented in the Hall anyway with Rusie, Nichols, and (eventually) Young. Looking ahead, there won't be much debate over Mathewson or Johnson. The trickier question will be about how to compare the other pitchers from the Oughts (Willis, Walsh, Waddell, Joss, Plank, Brown) to our backlog pitchers from the 80's. How does a short, flashy career like Whitney or Caruthers compare to a short, flashy career like Waddell or Walsh? And what is career bulk, anyway? Is it IP (advantage, 80's guys) or is it number of years as a full-time player (advantage often to the Oughts guys)? Some examples: Vic Willis pitched as many years as Mickey Welch (with a slightly higher ERA+), but 800 fewer innings. Eddie Plank had about 2 more years in the majors than Pud Galvin (with a higher ERA+), but 1500 fewer innings. Walsh (whose usage at peak was almost but not quite a throwback to the 80's) still had 500 IP fewer than Whitney - but a 40 point edge in ERA+.

Since we've introduced Griffith into the mix, I think this list of Fibonacci Win Points for 19th Century Pitchers is very relevant to the discussion:

Name Wins Losses Pct. Win Pts.

Here's are two ideas concerning determining the appropriate percentage of pitchers in the HoM. They both assume that one can take an available IES as reliable.

First, we might consider whether the distribution of pitcher values in relation to average is any different from that of position players. The case for 20% states, essentially, that under the 4-man rotation, a team's third and fourth starters don't count, and that among the top half of starters in the leagure, there will be about the same percentage of HoM-quality players as there are at the other positions. If this is true, then the metrics might show a larger pool of indifferent pitchers, out of which a few pitchers spike to greatness. Studying the distribution of values, assuming anyone has the resources to do such a study, could provide some useful information.

Second, it seems to me that so far, pitchers tend to have shorter careers with higher peaks than position players. They wear out their arms, but their playing time increases during their peak performance in ways that the playing time of position players doesn't. So pitchers as a class, being compared to position players as a class, raise the problem of weighing peak vs. career value. Pitchers are going to have higher peaks, so do we discount those peaks when comparing them to position players? Or do we give them full credit for the peaks, knowing that it is counter-balanced by lower career values?

jimd and John, wouldn't it also be perfectly logical to think that 35.1% of HoMers would be pitchers because pitchers earn 35.1% of WS in that system. (This assumes of course that WS has it right and that we all agree to vote according to WS. This is of course not just a large assumption [or two large assumptions], they are both in fact inaccurate assumptions. But you get the point.)

Even if this could all be assumed, however, the problem is that the 35.1% of WS that go to pitchers distribute themselves across those pitchers according to their own logic. Ditto hitters and fielders and their 64.9% (forgetting for the moment that some of the fielding also goes to pitchers).

If four pitchers share their team's pitching WS equally, they each get about 9%. If eight position players share the rest, they each get about 8%. With two pitchers, they're getting 17.5% each and with five, just 7%. With nine position players (and DH), they get 7%. But so what? The

<i> I'd be curious, did you identify any specific comps for Wright's and Pearce's

I was a leading 'enemy of Amos Rusie.'

<i>Clark Griffith 237 146 .619 238

Howie,

For what it's worth, WARP and WS both corroborate the results of OCF's exercise: they see Rusie as having a higher peak and more career value overall than Griffith.

I had Rusie sixth, myself, though now I think I underrated him a bit.

Here is my rough expected HOM pitcher ratio: (I track this and other positions but it really doesnt mean much. 3rd base pre-Sutton-election was clearly underrepresented and Sutton was clearly the best 3rd baseman, but you can check my ballots to see how much that helped him :))

Timelines are approximate.

pre 1900 votes: 11% (1 pitcher: 8 position ratio)

Apparently, I am posting from the Atlantic time zone.

<i>pre 1900 votes: 11% (1 pitcher: 8 position ratio)

>I'm surprised that the group that put Rusie in on the first ballot and elected Galvin

Was that a group? I had Rusie near the top of my ballot and I fought Galvin kicking and screaming all the way. So I don't know which way to go with Cal's daddy.

Leadoff and other batting positions

Principal Fielding Positions of Win Shares leaders

I appreciate that, Paul. It may take me a while to formulate any kind of response. "The leadoff hitter is an outfielder (most likely CF or LF)" and "the catcher bats 8th" are both strong rules, but both do have exceptions.

Here is my thought on pitchers. In todays games, while some teams play a true 5 man rotation, most play a 4.5 man rotation. If everyone stayed healthy, the first 4 guys would get about 36 starts each with the 5th guy getting 18. Fifth starters are like backup catchers or utilityinfielders. You gotta have them, they will get their playing time, but you try to skip them as often as possible. Working on the theory (totally unproven) that all positions are created equal, there are 8position starters and 4 starting pitchers. Plus, the closer position has become a "starter" position. Im not sure about the whole 2nd reliever that Ive added in for post 2000, but Ive got a few years to figure that one out.

Thats the world of today. In the early days, teams had effectively 1

Regarding post #68:

Obviously, Honus Wagner would be among those with WS > 20. In 1900, he was an outfielder. In 1901, he's listed with 54 games in the OF, 61 at SS, 24 at 3B, and 1 at 2B. Where did you wind up counting him in 1901?

I slotted him at shortstop because I don't lump all of his OF work as one position. Obviously, like Ernie Banks and Deacon White for their careers, his numbers for that year need to be massaged somewhat.

I haven't finished Babe Ruth's numbers yet, but I won't be surprised if Honus is the best player of all-time under my ranking system.

So we're the (mythical) sabermetrically-informed fans in the winter of 1911-1912. We've been talking mostly about players who are already eligible for election to the HoM, but there are a few other things we would probably be talking about.

1. What about that huge jump in offense that happened this past year? Are the balls different? Are the big-offense days of the 90's about to come back?

2. The greatest player in the history of the game walks among us. Oh sure, there are a few old fogeys who will want to insist that the moderns are but pale shadows of the true heros of the past like Cap Anson or George Wright, but really! Wagner will be 38 next year, but since he just led the league in OPS and can still play SS, who knows when he'll finally retire?

3. There are sure a lot of hot young outfielders in the AL. In particular, Cobb already has 1200 hits and 600 runs - through the age of 24! His OPS+ has been hovering around 190-200 for three years running.

4. That young 3B for the A's - Baker - will he have the kind of career that will make us forget all about Sutton and Collins and all of the other third basemen of history? Too soon to tell, but he's been awfully good so far.

5. Has Cy Young finally hung up his spikes? He was 44 last season, after all. We would have elected him to the HoM a while back, but we've been waiting for him to become eligible. He's lasted so long that we've started comparing him to Christy Mathewson, who was all of 9 years old when Young was a rookie. Now remember folks - just because Young had another decade-plus in him after the age of 30 doesn't mean that we have any right to expect the same from Mathewson.

6. There's this young power-pitching sensation from out west, goes by the nickname "The Big Train." Flash in the pan, or the real deal? You never can tell with pitchers, can you? Just got a visit from a 21st-century time traveler muttering names like "Prior" and "Beckett" - sorry, we really can't help you with that prediction.

My numbers are an attempt to gradually move from one extreme to the other.

RobC, thanks for answering. Correct me if I'm wrong, but it appears you have a model roster for 1871 and one for 2001 and then linearly interpolate between them.

I would suggest that team pitcher-usage data indicates that the initial evolution was much more rapid, going from 1 starting pitcher in the 1870's to 2 pitchers in the 1880's, 3 in the 1890's, and arriving at 4 in the 1900's. It then remains constant for a long period of time, though the IP distribution between 1 and 4 gets flatter over time; I haven't done the modern roster studies yet but IIRC the 5th starter doesn't become conventional until the 1970's/80's time-period.

this from a Newark Star-Ledger article today, about an '1873 game' in Hohokus (Bergen County, NJ) on Sunday. If anyone is interested...

Every game played by the baggy-trousered Flemington Neshanock Base Ball Club is on a field of dreams. In that misty venue a batter, or "striker," can request a pitch in a strike zone that suits him, foul balls caught on a bounce are an out, calls can't be argued, and no one can address the umpire except a team captain.

Wait, there's more: Everyone barehands, not a glove is in sight. The pitcher (hurler) pitches the ball underhand, the first pitch is a "deadball" that doesn't count unless the striker chooses to swing, a foul ball is not a strike, and a base on balls is awarded after three balls, not four.

Walt, this debate is already over. You lost. Since you missed it, it happened at 1:46 with this:

>Guess that shows using similarity score for players from

A last word on why Clark Griffith will not be appearing on my ballot.

Comparing Griffith directly to the five pitchers who received multiple votes in 1911 (Caruthers, Mullane, McCormick, Welch, and Whitney):

Wins: A crude measurement, but generally tells you something worthwhile. Griffith is 4th of the 6, better only that high-peak Caruthers and hard-luck Whitney.

Innings pitched: Fifth out of six. Only Caruthers pitched fewer innings. Sure, tell me usage was different when Griffith pitched, but that doesn't tell me why he is more valuable than those who pitched more innings. Especially considering:

K/9: Dead last. If he pitched fewer innings, and didn't strike people out when he was in, and didn't get wins other ways, he's beyond thin ice in my book.

BB/9: Fourth of six. He walked fewer per 9 than Welch and Mullane.

Hitting: Bad. It's easy to overlook grades of mediocrity when comparing pitcher hitting, but this doesn't help Clark either. His OPS+ of 69 tops only Welch (68) and his Batting Runs over Replacement by Position (BRARP on Prospectus) tops only Welch and McCormick.

WARP-1. Not surprisingly, therefore, in the holistic WARP measurement, Griffith comes in a convincing 6th out of 6, barely beating out the vote-poor but still over-valued Jim Whitney.

I have Caruthers and Welch on my ballot now. I would bring McCormick and Mullane on before I would add Clark Griffith. Nothing person against him, but compare him 1 on 1 to any of those four and he simply fails to stack up on any substantive standard.

MattB, before you make up your mind, consider that virtually all of the comparisons you are making are heavily influenced by the change in pitching distance, which not only reduced the number of innings pitchers could throw, it raised the average walk rates and lowered the average strikeout rates. Of course Griffith is going to look worse by these measures!

To assess Griffith's value, you need to look at his rate stats relative to the pitchers who were working under the same conditions he was. Griffith was actually a very fine control pitcher, relative to other folks working on four-ball walks from 60 feet, six inches. If you compare rate stats across the mound-move (and the other changes that separate 1880s pitching from 1890s pitching), all the 1880s pitchers are going to look better.

If you want to value players in terms of their absolute impact on the game, all 1880s pitchers were much more valuable than 1890s pitchers, as a group. That's a defensible position to take.

Since you pretend Joe Start played 162 games a year, I thought maybe you could pretend I played 21 seasons.

I hope this is not one of the voters here because I don't like the tone and the cowardly way of handling this dialogue. If you have a reasonable argument against Start, fine, let's hear it. If not, then accept the inevitable.

I didn't piss and moan about Rusie getting the keys for induction on his first try. I made my objections and went on. We all want our selections rubber stamped, but we all have different ideas about how to rank the players. I think we should all learn to accept everybody's different viewpoint or we shouldn't participate.

BTW, I made the Joe Start rebuttal and a couple of the other celebrity posts.

MattB, before you make up your mind, consider that virtually all of the comparisons you are making are heavily influenced by the change in pitching distance, which not only reduced the number of innings pitchers could throw, it raised the average walk rates and lowered the average strikeout rates. Of course Griffith is going to look worse by these measures!

To assess Griffith's value, you need to look at his rate stats relative to the pitchers who were working under the same conditions he was. Griffith was actually a very fine control pitcher, relative to other folks working on four-ball walks from 60 feet, six inches. If you compare rate stats across the mound-move (and the other changes that separate 1880s pitching from 1890s pitching), all the 1880s pitchers are going to look better.

If you want to value players in terms of their absolute impact on the game, all 1880s pitchers were much more valuable than 1890s pitchers, as a group. That's a defensible position to take.

Wins: A crude measurement, but generally tells you something worthwhile. Griffith is 4th of the 6, better only that high-peak Caruthers and hard-luck Whitney.

Because it is a crude measurement, when I look at wins, I try to give some credit for Wins Above Team. I don't have any numbers in front of me, but I recall that Griffith has a very good WAT number (approximately 30). From memory, Welch also has a good WAT number, but not at Griffith's level. Just something to consider.

I hope this is not one of the voters here because I don't like the tone and the cowardly way of handling this dialogue. If you have a reasonable argument against Start, fine, let's hear it.

I'm glad you said this, John. I don't like the cloak and dagger approach.

Before anybody gets too bent out of shape over Walter's comments, I'd like to point out that I found them amusing. I even took the first one as a veiled criticism of similarity scores, rather than a direct attack on FOJS.

Walter's second comment could be taken as a veiled criticism against fans of players with undocumented careers. This may be objectionable to some, but people should generally be entitled to their opinion. I wasn't a voter a few years ago, but there seemed to be some heated discussions regarding the support of Negro Leaguers.

I don't agree with Walter (Start is 3rd on my ballot, McVey 4th, Grant 9th) but I don't think he can or should be censored or vilified for his minority position. I submit that support of Joe Start may once have been a minority position which was carefully examined. Support of Walter Holke may be just as carefully examined and the majority will decide whether to accept or reject that position.

Sorry, Walter, but I will probably be an EOWH for the HOM.

Before anybody gets too bent out of shape over Walter's comments, I'd like to point out that I found them amusing. I even took the first one as a veiled criticism of similarity scores, rather than a direct attack on FOJS.

I think "Walter's" posts showed that his problem is with Joe Start, not with similarity scores. Which is fine, BTW. But I would like to see one baseball historian that would compare Start and Holke equally as players. Holke was a journeyman player, while Start was a star. Frankly, it's a stupid comparison.

But I would have left it there (except for my Start rebuttal) until the second post. That was just plain dishonest and inflammatory. His or her bitterness shone through there due to Start's imminent election. Well, I doubt one of us here will have all of his choices picked by the rest of us. That's life.

Humorous anonymous posts are fine by me here (I have made my share) as long as they make a legitimate point and are not made to be offensive. "Walter" failed that test.

I think the cross-era comparisons and validation of our methods (yes it does make sense to normalize Start to 162 games, but not Holke to 21 seasons) are useful. The Start/Yastrzemski parallel is interesting; I'm an FOJS, but even more of a FOCY and it tells me that JS is very nearly CY but not quite, because there's no 1967 (if there had been a CY-1967 for JS in the 1860s, we might not have the stats for it, but we'd sure as hell have the media excitement.)

karl, I said "years" ago that Yaz was an interesting comp for Joe Start. I'm not a Red Sox fan, sorry, but there is no denying Yaz. I can't believe he won't be a 1st ballot selection, but of course it will help his cause immensely if his comp Joe Start goes in, too ;-)

Chris wrote:

"To assess Griffith's value, you need to look at his rate stats relative to the pitchers who were working under the same conditions he was. Griffith was actually a very fine control pitcher, relative to other folks working on four-ball walks from 60 feet, six inches. If you compare rate stats across the mound-move (and the other changes that separate 1880s pitching from 1890s pitching), all the 1880s pitchers are going to look better."

I completely agree with Chris's point that an additional factor to consider is how pitchers rate in comparison to peers. Unfortunately, I don't see how this adds anything to Griffith's case. Comparing the same six players on the pitching "grey ink" test, Griffith finishes fifth out of six, over only Caruthers (and would finish dead last, of course, if I included Caruthers' hitting grey ink). That number includes ZERO top ten finishes for most strikeouts per 9 innings for Clark, compared to, for example, 8 for McCormick, and 4 for Welch. He has 5 top 10 finishes in ERA+ (tied for last with Whitney and Caruthers, who had higher peaks), including a first and a fourth, but also an 8th, 9th, and 10th. His strong suit in in BB/9, where he has 11 top ten finishes, but none higher than fourth (and in one year in which he finished 10th in BB/9, he actually allowed the most walks in the league!).

If Clark were striking out fewer than the 1880's guys, but more than his peers, I could see an appropriate adjustment. But in 1895, for example, Clark had 26-14 with 79 K's and 91 walks, while Amos Rusie led the league with with 201 and another 5 pitchers had over 100!

I am leaning toward the belief that pitchers were more important in general in the 1880s than in the 1890s. But even if we normalize value across the decades, Clark Griffith still trails the pack.

OCF(#71):

As a Redsox fan and a fan of players named 'Carl', I will be a huge FOCY.

RobC: A roster driven ratio for the 1880's indicates 8:2. (20% pitchers)

jimd: pitcher-usage data indicates (. . .) 1 starting pitcher in the 1870's to 2 pitchers in the 1880's, 3 in the 1890's, and arriving at 4 in the 1900's. It then remains constant for a long period of time, though the IP distribution between 1 and 4 gets flatter over time; I haven't done the modern roster studies yet but IIRC the 5th starter doesn't become conventional until the 1970's/80's time-period.

Two data points: In 1901, NL rosters normally included 5-6 pitchers and 10-11 others (16). AL rosters normally included 4-5 pitchers and 9-10 others (14).

Two informal observations: The Fred Clarke pirates and Frank Chance cubs commonly used 5 or more starting pitchers. For extreme example, see Chicago 1906-1907.

If (a big IF!) our top two holdovers are inducted this year, I'll be very interested to see where top newcomer Jake Beckley slots in next year as the clear top first baseman eligible.

He looks like a clear Top 10 guy to me, but it sounds like lots of folks are down on him. He strikes me as the next most likely player to pull a Hardy Richardson (receive at least one vote for every position from first to fifteenth and be left off of at least one ballot entirely.)

I bored discussion whether Griffith should be 10th or 20th on my ballot. Anyone know where they will be putting Beckley?

MattB: a rough estimate has Beckley at about 12-14 on my ballot.

Another note on Beckley: I think it may be safe to say that he has the worst peak of any player we have seriously considered. Its not even close to McPhee's peak. He wont even come close to the ballot of primary peak voters. His "peak" knocks him down 3-4 spots on my ballot, and I barely consider peak.

There hasn't been much discussion of Deacon McGuire, although karlmagnus just posted a ballot suggesting that he considers McGuire to be his #16 - nipping at the edges of the ballot.

If you like longevity, this guy's got it. He played from age 20 through age 42, and had 1611 games caught, which is a huge number. His playing career crossed over multiple large changes in the game. My first thought was "Bob Boone." Maybe, but McGuire was probably better. Boone played from age 25 through 42, and caught 2225 games (with much better protective equipment than McGuire had). Boone had a career OPS+ of 82, topping 100 in only 5 of his seasons. McGuire had a career OPS+ of 101, with 10 years above 100, 6 of those above Boone's best of 115. Not that it matters much, but when McGuire was old and making token appearances, he hit a Forsteresque .400/.400/.667.

I still have no idea what I'll do with him.

Well, Beckley's career OPS+ was 125, a mark he TOPPED 11 times!

Jim Spencer wrote (out of order):

"ERA+: Caruthers is the only one higher, in fewer (unadjusted) innings, in a much weaker league."

Griffith's ERA+ was 121 in 3386 innings. McCormick's was similar (and within the margin of error) with 118 in about 900 more innings. Mullane's was also 118 in over 1100 more innings. I'm all for a higher ERA+, but I'd gladly sacrifice 3 points for 1000 more innings.

"Hitting: Bad???? Griffith's OPS+ of 69 is very good for pitchers in his era, it can't possibly be considered a negative."

I was comparing Griffith to the other eligibles, against whom he stacks up badly.

Comparing him to his peers, however, I randomly chose 1896 because Griffith's OPS+ (73) was closest to, without going under, his career ERA+ (reverse Price is Right rules). Looking at #1 pitchers (defined as "listed first in b-r.com")that year, Bill Hoffer had a 110 OPS+ that year. Cy Young had an 89. Frank Dwyer was 82. Frank Killen was 84. Joet Meekin was 114.

Of the 12 #1s, Clark Griffith finished 6th. (Below him, Kid Nichols had a 36 OPS+. Jack Taylor was 20. Brickyard Kennedy was 10. Win Mercer was 54. Ted Breitenstein was 69. Bill Hill was 20.) So yes, some pitchers were horrendous hitters, but it certainly was not the norm. At best, his hitting does not help his case relative to the other pitchers on the ballot.

IP: Look at the context.

Context tells me that pitchers were throwing fewer innings per season. But Griffith only appears on the top ten for innings pitched TWICE in his career. (cf. 9 for Welch, 8 for Mullane, 7 for McCormick, 3 for Caruthers). I don't give Griffith credit that pitchers were generally throwing fewer innings in th 1890s if Griffith was pitching even fewer than his contemporaries!

K/9: Preventing runs is the object, not striking people out.

BB/9: Are runs important, or walks?

Griffith's RA was 4.92. McCormick's was 4.41. Caruthers' was 4.43. Welch's was 4.79. Mullane's was 5.01.

I'm not sure exactly what stat is supposed to make Griffith stand out. He's got a better winning percentage? Sure, he was throwing 100 fewer innings per season than his contemporaries were.

His stats are not better than the others on the ballot. His stats are not better than his peers. Under either measure, he should not be placed above the more valuable pitchers of the 1880s.

I am leaning toward the belief that pitchers were more important in general in the 1880s than in the 1890s.

Can you flesh this thought out a little more? Since fielding was improving rather rapidly, I'd be interested to hear the theory that pitchers were taking on a smaller role. I would have guessed otherwise.

In 1880 NL, earned runs had a 1:1 relationship to unearned runs. By 1890 NL, earned runs had a 1.6:1 relationship to unearned runs. By 1895, the ratio was 2.3:1. To me, this indicates more of defense was attributable to pitching than fielding.

So in the early 1880s NL, unearned runs accounted for roughly 50% of all runs scored. By 1895, unearned runs accounted for only 30% of all runs scored. Although there are some other factors at work, it seems like fielding was generally improving (which makes logical sense), and therefore more of the runs allowed in a game were attributable to the pitchers' performances than to the fielders' performances...for better or worse.

Is that right, or do I have it backwards?

Was Clark Griffith close to being the best pitcher in the game for any one season? His peak wasn't that extraordinary when you put his career in context.

I agree with Matt that his lack of true durability within his era really hurts his HoM chances.

In reading my own post #104, the following thought occurs:

I can't speak for MattB, but I think individual pitchers were more important in the '80s though pitching was not.

John, I think it's a bit unfair to dismiss Clark Griffith because he was not Cy Young or Kid Nichols. Was there ever a period with two more dominant pitchers than that? Later on it's Johnson-Matty-Alex and Walsh and Brown, too, so worst case we elect five and of whomever was sixth we will say, "Was he ever the best pitcher in any one season?" Or in the '60s it will be Koufax, Marichal and Gibson, at least, and of whomever is fourth we will say, "Was he ever the best?"

But to say that of possibly the third best pitcher of his time (and I'm not saying he was, BTW, but possibly) before he's really been evaluated is a very tough hill for any pitcher to climb.

It's like saying, well, was (who besides Babe Ruth and Rogers Hornsby during the 1920s) ever really the best hitter?

I'd agree that pitching an exceptional number of innings in any single season was not one of Griffith's particular strengths, but for his career he was rather durable.

At the end of the 1912 season, he ranks 22nd among all pitchers in lifetime innings pitched.

Of the 21 pitchers ahead of Griffith, 12 played almost exclusively in the period of 1871-1893 (Galvin, Keefe, Mathews, Welch, Clarkson, Radbourn, Mullane, McCormick, Bond, White, Whitney, Buffinton), 3 had careers that straddled the watershed year (Weyhing, Rusie, and Adonis Terry), and and 6 had careers primarily after 1893 - 1912 (Young, Nichols, Mathewson, Powell, Willis, McGinnity).

Given that the frequency of pitchers reaching this number of innings clearly declines significantly after 1893, Griffith's career seems to me to demonstrate pretty significant durability.

However, given that so far we've inducted only pitchers with really outstanding durability (the top 10 IP is made up of 5 HoMers, 2 ineligible but will-be-first ballot HoMers in Young and Mathewson, and Mathews, Welch, and Mullane) or who have really top-notch, sustained peaks (Rusie and Spalding), Griffith clearly doesn't fit that standard.

But I think we're going to need a new standard, as we can't fill even 20% of the HoM with pitchers in the Young, Nichols, Mathewson class, and I don't know if Waddell and Walsh will match up well with the Rusie/Spalding peak model. Will Griffith meet that new standard? I can't tell yet. I don't think we're going to be able to rank Griffith really accurately until we have more eligible pitchers of the post-1893 generations to compare to him: Powell, Willis, McGinnity, Waddell at least. Three Finger Brown is building up a pretty similarly shaped career, working for a Chicago staff that spreads innings around rather than riding a #1 workhorse.

John, I think it's a bit unfair to dismiss Clark Griffith because he was not Cy Young or Kid Nichols. Was there ever a period with two more dominant pitchers than that?

Of course. You lived through the Clemens/Maddux era, didn't you? The only difference is there are many more good pitchers now so they don't stand out like they would have had they pitched one hundred years earlier.

Young and Nichols were, without a doubt, great. However, they weren't OH, MY GOD, GREAT!!!! If they were both around today, we wouldn't be going around in total disbelief at their achievements.

At the end of the 1912 season, he ranks 22nd among all pitchers in lifetime innings pitched.

He did have a fairly long career for his era. While he didn't pitch a great many innings per season, he was able to last longer than most of his peers.

I still have him as the fourth best pitcher of his era (Young, Nichols and Rusie being better).

Jeff wrote:

"In 1880 NL, earned runs had a 1:1 relationship to unearned runs. By 1890 NL, earned runs had a 1.6:1 relationship to unearned runs. By 1895, the ratio was 2.3:1. To me, this indicates more of defense was attributable to pitching than fielding."

Marc wrote:

"I can't speak for MattB, but I think individual pitchers were more important in the '80s though pitching was not."

This is exactly what I was thinking. Any individual inning was more pitching-dependent in the 1890s, but top pitchers in the 1880s pitched many more innings. In my estimation, the larger quantity of innings more than counteracts the lower value of any given inning pitched.

I'll write some more on this later.

Refresh my memory on the story on the 1885 Providence Grays and the 1886 Washington Nationals.

This is exactly what I was thinking. Any individual inning was more pitching-dependent in the 1890s, but top pitchers in the 1880s pitched many more innings.

But the top pitchers of the 1880s pitched many more innings than the 1890s pitchers because it was easier to do, not because of greater skill. Should they then be rewarded (via more elections to our Hall) for this difference?

An odd occurance last night... I was idly channel-surfing and I happened on the PBS show Antiques Roadshow. This is the show that invites people to bring out the odd-looking old stuff around the house so the host can explain what it is and tell the owner that it's surprisingly valuable. The bit I happened on involved a man from Kansas City who had a painted portrait of Kid Nichols, along with an array of other Nichols and Beaneater memorabilia: a team photograph, an individual photograph of Nichols, a variety of printed materials. The owner of this stuff didn't act like he was particularly familiar with Nichols, so the host ran down some of the Kid's statistical accomplishments (e.g., he won 30 games in a season more times than anyone else). The person I was watching with, who is a knowledgable baseball fan, said something to the effect, "If he was that good, how come I haven't heard of him?"

>But the top pitchers of the 1880s pitched many more innings than the 1890s pitchers because it was easier to

The glut of 1880s pitchers is a bit of a myth, because we have a dearth of 1870s pitchers (1 - Spalding) and I don't think anyone would argue that Matthews was better than Walsh or Caruthers. The 1880s were just a good decade for pitchers, as were the 1960s -- there is a LOT of room for random fluctuation in numbers like 1,3,5 or 7, both up and down.

As for 3B, I buy the argument that it was a difficult but unimportant fielding position -- anybody good enough to field 3B was moved to 2B or SS, where his talents were better used.

The 1880s were just a good decade for pitchers, as were the 1960s -- there is a LOT of room for random fluctuation in numbers like 1,3,5 or 7, both up and down.

The 1960s has the same problems as the 1880s. The conditions were perfect for both eras to chalk up high win totals. If the top pitchers of the 1960s had started pitching during the seventies, there is no way that we would have had as many 300 game winners.

Context, context, context.