Prelim:

1) Pete Browning-CF/LF (3)

Eleven outfielders and nary a pitcher, John? How do you reach that conclusion about value?

Well, John, you beat me to it...and speaking of beating...gotta beat you up for your vote for Willie Keeler. It is very obvious to me that Wee Willie is one of the most overrated players in history. Elmer Flick is rated #23 among RF by Bill James and Keeler just #35 despite their raw career totals. Flick's peak was vastly better and anyone who cares a whit about peak value has gotta have Flick above Keeler.

Having said that, I don't think either one of them is a top 5er. I'm kind of expecting Flick to come in around #10-12 in fact, but I doubt seriously that I will have the Wee one on my ballot.

John, aren't you the guy who says we tend to overrate players in their first year of eligibility? Maybe not. But the fact is, we do. I think Keeler is perhaps a case in point. He could get elected this year, but if he finishes, well, 3rd, where John has him, he could end up dropping to 10th in a couple years. Let's not be too hasty on this one!

But of course I shouldn't assume that lots of voters will follow the lead of a man who has Tip O=Neill #2!? John, go back to your old system! Aside from Deacon McGuire, you could go from 14 to 1 and I'd like your ballot better!

I noticed an oddity in the shape of Elmer Flick's career.

What was the most intense, hardest to dominate competition of this whole time period? It must have been the post-contraction 1900 NL. There were so few weak players in that league, and the top teams look like all-star teams. That was the single best year of Flick's career. That was his OPS+ of 172.

For 1902 he jumped to the AL. The 1902 AL was still very young, and the level of competition must have been softer than the league Flick left - and yet Flick's own OPS+ dropped to 120 for that year. As the AL gained strength over the next few years, so did Flick compared to the league.

Flick seems to have played better in stronger leagues. Someone who thrived on the challenge?

Can someone tell me what the story of Flick's 1902 season and his trade from Philadelphia to Cleveland was? Was there some sort of injunction against him in Pennsylvania?

But of course I shouldn't assume that lots of voters will follow the lead of a man who has Tip O=Neill #2!?

You're right, Marc. I'm a moron. You can be my proxy for now on.

:-D

But of course I shouldn't assume that lots of voters will follow the lead of a man who has Tip O=Neill #2!?

You're right, Marc. I'm a moron. You can be my proxy for now on.

:-D

as per baseballlibrary.com.....

April 21, 1902: The Pennsylvania Supreme Court, reversing a lower court's decision, grants a permanent injunction (effective only in Pennsylvania) barring jumpers Nap Lajoie, Chick Fraser, and Bill Bernhard from playing for the A's, or any team but the Phillies. Not mentioned, but covered by the decision, are: Elmer Flick, Monte Cross, and Bill Duggleby of the A's; Ed Delahanty, Al Orth, Harry Wolverton, and Jack Townsend of Washington; Ed McFarland (White Stockings) and Red Donahue (Browns).

? May 7, 1902: Elmer Flick signs with Cleveland. Sporting Life says that A's officials "presumably consented" to the move.

Is Cy Seymour eligible in 1916? I see that he played 39 games in 1913.

Thanks, Howie. Would it be safe to assume that Flick didn't travel with the team when Cleveland played at Philadelphia? That would account for his slightly low number of games played that year.

Incidentally, this must have created an unfair advantage for the Athletics, who saw weakened versions of some visiting teams, notably Cleveland. Also: although Duggleby didn't play at all, Cross managed to put in a full season for the A's. The A's won the pennant by 5 games.

Now that I finished my dinner...

1) This is only a prelim. I have been known to make some changes by election time.

2) Elmer Flick just missed (and I mean just missed) my ballot at #17. But since I treat career equally with peak, that hurts him. IOW, Keeler's career advantage is larger than Flick's peak advantage.

3) Willie Keeler, under my old and revised systems, was the best rightfielder of his time. That has more value to me than being only the fourth best leftfielder of his time (Kelley).

There are many here that would (hypothetically speaking) talk about what a great peak a player had because of a 400 OPS+, even if it were only average for his position during his era. What type of value is that?

BTW, Marc, I've noticed that you haven't matched Bill James perfectly yourself on your ballots. Of course, I thought we were supposed to be making our own judgments and not just rubber stamping his or any one else's opinion. :-)

4) As for pitching, I'm just not that impressed with what I have seen. I'm not that much of a peak guy to include McGinnity or whoever else because of their abridged careers. However, as I have stated before, I might revise my thinking on this matter.

5) Regarding the swipe at my selection of O'Neill, I have mentioned the different situation outfielders (and other players) faced during the 1880s compared to other eras. I have explained this reasoning on a number of occasions, but you're still free to deride me for it regardless My skin is tough enough to take it. :-)

Eight straight years finishing no lower than 7th in adjusted ERA? 4 straight years in the top 3? That looks like a great peak to me.

I noticed that you forgot to mention where he placed in innings pitched (or maybe you didn't forget). :-)

Just my position-by-position consideration list so far in this year of fun new eligibles:

C:

Elmer Flick - I like Marc's guy too, though, and may rate him higher than Marc on my ballot.

Marc apparently will make my list. I'm not sure what position he plays, though.

1916 prelim. What a fascinating year. So many good new candidates, and there's sure to be a wide range of opinions on their merits.

I've also big changes to my rating system since last "year". In response to a question on why I had 4 shortshops and 0 pitchers on my last ballot, among other comments questioning the numbers of pitchers on ballots, I reexamined my rating system and discovered that I was not giving pitchers a fair shake with respect to position players. Another change since last time is that the number of points I give for career length for position players depends on position. Since 1B have tended to have longer careers than other positions, they get relatively fewer points. Beckley takes a big drop.

I have five new candidates on my ballot. They break down into two types: outfielders and pitchers.

Outfielders: I'm not worried about rating players highly in their first year of eligibility if that's where I think they belong. Over the next few years, the top OF of the 1900s will leapfrog the second-tier OF of the 1890s and land near the top of my ballot. Flick's peak trumps Keeler's career, and they're split by my perennial favorite Jennings in the top 3 slots.

Pitchers: Let the controversies begin. In reevaluating pitchers, I was gratified to find a perfect cutoff between those pitchers already elected, and the eligible pitchers that had not been elected. (Actually, Splading falls short in 1871+ play, but makes it with credit for pre-1871 play). My ballot includes all of the top 3 newly eligible pitchers, and none of the oldies.

I'm sorry to have to drop Duffy, Van Haltren, and Charley Jones from my ballot.

1. Elmer Flick (N) RF. Prime 1898-1910. 291 WS/6414 PA; 27.2 WS/600 PA. What a difference a few years makes. Most similar batter is Kep Selbach, but they were as different as night and day with respect to league averages.

2. Hughie Jennings (3) SS. 1894-1898: 150 WS/2989 PA; 30.1 WS/600 PA. Players such as Jennings and McGraw deserve to be rated on their own terms, not put into inappropriate one-size-fits-all career value systems. Has the strongest prime of any player on the ballot.

3. Willie Keeler (N) RF. 1892-1910: 333 WS/9594 PA; 20.8 WS/600 PA. Great career length in a position not known for career length.

4. George "Rube" Waddell (N) P. 1897-1909: 136 ERA+ in 2928.3 IP. Reportedly detested his nickname, so I've included his given name. Great combination of career length, good years, and Cy Young-type years (1904-1905).

5. Lip Pike (4) IF/RF/CF. Prime 1866(?)-1878. 158 OPS+ in 2006 PA in NA/NL. He was probably a "character"; nonetheless, I trust the numbers.

6. Addie Joss (N) P. 1902-1910: 142 ERA+ in 2327.0 IP. 8.5 years at this level is no fluke.

7. John McGraw (6) 3B. 1894-1901: 169 WS/3618 PA; 28.0 WS/600 PA. Best 3B of 1890s. 2nd strongest prime of any player on ballot.

8. Cupid Childs (7) 2B. 1890-1898: 208 WS/5312 PA; 23.5 WS/600 PA. Best 2B of 1890s. Pike and Childs were best at position 5 times, more than any other players on ballot.

9. Frank Grant (8) 2B. The primary basis for picking players from eras and leagues where the statisical evidence is lacking is their career length. Grant had a long career and a reputation as the best black player of the 19th century. I have chosen to give him the benefit of the doubt and assume that his career was nearly at a McPhee/Childs/Richardson level.

10. Dickey Pearce (9) SS. Subjective rating. If I were sure that he should be in the HoM, he would be near the top of my ballot. I am sure that he was at least this good.

11. Joe Kelley (10) LF/CF/1B. 1893-1904: 272 WS/6644 PA; 24.6 WS/600 PA. Strong prime raises him above contemporary OF.

12. Vic Willis (N) P. 1898-1910: 118 ERA+ in 3996.0 innings. My system sees him as McGinnity quality plus two more years of quantity. Those extra years are enough to make a big difference.

13. Jake Beckley (5) 1B. 1888-1906: 318 WS/10348 PA; 18.4 WS/600 PA. Could be counted on to be average to very good for 17 straight years of full-time play. Career is similar in many ways to that of Eddie Murray, though Beckley was not quite as good. Takes a huge hit now that I'm giving less credit for career length at 1B.Close to consensus average now.

14. Jimmy Collins (11) 3B. 1896-1907: 257 WS/6520 PA; 23.7 WS/600 PA. Was very good, but just below the level that would make him an automatic HoMer.

15. Jimmy Ryan (12) OF. 1885-1903: 310 non-pitching WS/9106 PA = 20.4 WS/600, + 6 pitching WS. Long career raises him above contemporary OF.

Not on ballot:

22. Harry Stovey. 1B/OF. 1880-1893: 265 WS/6832 PA; 23.3 WS/600 PA. The AA from 1882 to 1885 was decidedly minor-league. Up through those years, Stovey played at 1B, where he was not as good as the ABC group, then switched to OF, where his level of play does not raise him above the glut.

23. Sam Thompson RF. 236 WS/6470 PA; 21.9 WS/600 PA. Best years in high offense leagues. His best years were better than the best years of contemporary OF Tiernan, Van Haltren, and Ryan, but the balance of his career was not as good.

28. Charlie Bennett. C. 1880-1893: 155 WS/4116 PA; 22.6 WS/600 PA. Without giving special treatment to C, this is where he ranks. From the perspective of 2003, I know I'm not being fair to catchers, but from the perspective of 1916, I would have doubts about Bennett as a historic great.

29. Joe McGinnity. P. Prime 1899-1908. 121 ERA+ in 3441.3 innings pitched. More quantity than quality. W/L record more impressive than Willis'. On the other hand, it is often pointed out, for example, that Nolan Ryan's 8-16 record in 1987 should be ignored because he led the league in ERA. If wins could be incorporated into an ERA+ like rating in a consistent manner, that would be great, but I don't like to make special dispensations for individual players.

My prelim. I see Flick as better than Keeler, but relatively close. Keeler has more counting stats b/c he played much longer. Flick was a top ten-er more often than Keeler in a bunch of different significant hitting categories. In my opinion, Flick would have made more all-star teams. Flick's extra power helps him a bunch and allowed him to create more runs per out. Their 3-year and 5-consec-year WS peaks are comparable, but Flick pulls ahead in a 7-year measure. Keeler gets big points for long career. Who knows? It ain't an exact science, and I haven't looked at the new WARP yet for this list.

1. STOVEY, HARRY

John- I'm curious to know more about your system. In particular, it surprises me that you have Tip O'Neill rated very highly, but John McGraw doesn't make your ballot.

I've got a lot of work to do before I would dare put a prelim. out here, especially with John gunning for me ;-)

John, can you explain Tip O'Neill? Maybe you have and I missed it. And I don't just mean explain what you mean about '80s OFers, though I mean that to. But why O'Neill rather than, say, Charley Jones, who arguably has just as good a peak (adj for season length and league) and arguably has a better prime (if you give any credit for blackball years)? Or for that matter, Harry Stovey?

As for Keeler and Burkett? Both had about 2900 hits. One had 600 XBH, one had 425. Both hit about .340. One's OBA was .415, one slugged .415. The other's OBA was .388 and the one slugged .446. Both scored about 1700 runs. One drove in 950, the other drove in 810. Do I need to say that Keeler came up short in every category? Even OBA, even runs scored (OK by one). OPS+ 140-125. One was a B LFer, the other was a C+ RFer.

But mostly, I don't envy anybody who is determined to put a pitcher on your ballot! Which one?

Well, there's a lot to do with pitchers this time, and there seems to be quite a bit of support for the new arrivals. I haven't examined Joss, Leever, or Phillippe yet, but looking at Waddell and Willis, neither appears to me to be as good as either Griffith or McGinnity among their contemporaries. (I'll leave the 1880s pitchers out of the picture for the moment).

I am comparing pitchers on an extension of the "wins above team" line of analysis. That's an interesting metric, but the number it produces is highly dependent upon the other pitchers on the player's team. If they stink, the pitcher looks great. If they are excellent, the pitcher looks average, or worse. So instead of comparing pitchers to their teams, I am comparing their records to the record of an average pitcher with an average defense behind him would have achieved with the offensive support received, season by season, by the pitcher in question, to see how many wins above an average defense the pitcher (and the defenses behind him) is. I am then determining the defensive efficiency of the pitchers' teams to see how much the fielding may have contributed to the pitcher beating or falling short of average. I've calculated a "hits saved above average during pitcher's career innings" for each pitcher's career as a simple metric for seeing defensive quality. Here's a summary of my findings for the 1893-1910 pitchers now under consideration. I've left tiny seasons out of the reckoning, so ip and W-L records are slightly different from full career totals for some.

Pitcher -- ip -- W-L % -- W above team -- W above avg. defense -- hits saved by fielders

PS. Ron, I played a lot of 3B, a lot of SS, some 1B and a little 2B. But I hate to tell you I won't be eligible until about 1995. I will be honored to be on your ballot then!

John: 11 OF, no pitchers

yest: all OF and pitchers, no IF or C

Marc, you missed Jake Beckley in there. . .

My bad. I meant all corner-hitters.

I was going to lead off with a sarcastic comment, addressing the voters with interests that widen beyond the set of outfielders who amassed gaudy numbers against inferior competition.

But I decided against it ;-).

That's probably was a good idea because I don't know who you were referring to (okay, I do know hwo you were referring to). Speaking for myself, I use standard deviation when comparing each player to his contemporaries (besides adjusting for inter-league competition) so your statement is incorrect.

Good try, though. :-)

As for Keeler and Burkett? Both had about 2900 hits. One had 600 XBH, one had 425. Both hit about .340. One's OBA was .415, one slugged .415. The other's OBA was .388 and the one slugged .446. Both scored about 1700 runs. One drove in 950, the other drove in 810. Do I need to say that Keeler came up short in every category? Even OBA, even runs scored (OK by one). OPS+ 140-125. One was a B LFer, the other was a C+ RFer.

But they're two different positions. They shouldn't just be lumped together in the OUTFIELD cubbyhole that most people throw them into. That's the difference.

But why O'Neill rather than, say, Charley Jones, who arguably has just as good a peak (adj for season length and league) and arguably has a better prime (if you give any credit for blackball years)? Or for that matter, Harry Stovey?

To be honest with you, I haven't incorporated him in the new system yet (I forgot, to be honest with you again :-(. I'll definitely have him done by the election.

As for Stovey, he gets hurt because he played a plurality of games at first (where he was competing with the ABC boys and Orr). If he had played his whole career in left, he would have jumped over O'Neill fairly easily.

John- I'm curious to know more about your system. In particular, it surprises me that you have Tip O'Neill rated very highly, but John McGraw doesn't make your ballot.

I had McGraw on my ballot for a while, but was pushed off.

As for O'Neill, he was the best leftfielder during his era. What I did with him was compare him to the eight best leftfielders (and only for what they did in leftfield) during his time (with weights for most seasons played together). The eight candidates were himself, George Wood (8 seasons played at the same time as O'Neill), Abner Dalrymple (5), Charley Jones (4), Joe Hornung (7), Jesse Burkett (2), Jim O'Rourke (6) and Tom York (2) (in order of best to worst comp). Nobody was close to the Tipster during that time.

As mentioned earlier in a previous post, I use standard deviation so that a player won't stand out unfairly due to inferior competition. The SD figure is what I use when comparing different players for my ballot.

I'm going to add some more comments in the next few days when I have some free time about the '80s and careers..

As for O'Neill, he was the best leftfielder during his era. What I did with him was compare him to the eight best leftfielders (and only for what they did in leftfield) during his time (with weights for most seasons played together).

John- Thanks for explaining your methodology again. I like your system.

I looked at my data to see which LF were arguable as good or better than O'Neill during individual years of his prime and found the following names that are not on your comparison list: Hardy Richardson, Henry Larkin, Harry Stovey, and Darby O'Brien (1887-1892). The first three were not primarily LF for their careers, but did play LF head-to-head with O'Neill during particular years. I would think that it would be essential to throw them into the mix of players that O'Neill is being compared with.

Clear to me that Keeler and Flick should NOT get in on first try.

John, I think you are taken these position comparisons too far. Don't the 1990's versus 1970's shortstop comparison and the 1930's versus 1910's first base comparisons show that this method shouldn't be applied too strictly?

<i>Yest: "Don't you realize a SHORTSTOP that is an above average hitter is a great player?

Prelim

1. Joe Kelley - I bumped him up last week and only the dynamic duo kept him from the number one spot. There is no one in the way this time.

2. Willie Keeler - I am surprised at the negative comments about him. To me he is a guaranteed HoMer.

3. George Van Haltran - Still worthy of a spot.

4. Bob Caruthers - So maybe there is room for one more pitcher from the current crop.

5. Jimmy Ryan - My last sure thing.

6. Frank Grant - Borderline HoM.

7. Hugh Duffy

8. Jake Beckley

9. Mickey Welch

10. Jim McCormick

11. Jimmy Collins

12. Harry Stovey

13. Dickey Pearce

14. Tony Mullane

15. Vic Willis

16-20. McGinnity, Flick, Childs, Thompson, Griffith

John Murphy wrote [re: Burkett v. Keeler]

"But they're two different positions. They shouldn't just be lumped together in the OUTFIELD cubbyhole that most people throw them into. That's the difference."

In my view, there are two "cubbyholes", LF/CF and RF. As far as I can tell, left fielders and centerfielders were about equally important defensively in this time period, and right fielders were significantly less valuable defensively.

So I agree that they should be considered separately. Where I lose you is that Keeler should be considered MORE IMPORTANT by standing out in a LESS IMPORTANT position. It's like saying, "Edgar Martinez may be less valuable offensively than Nomar Garciaparra, but Edgar gets a boost for being the best DH in the league."

I understand giving a boost to a player for being in a challenging area (like boosting a student's GPA for being in an honor's class), but not for playing right field. ("Sure, he wasn't a star, but he DOMINATED remedial outfield.")

Wee Willie won't wash with wily watchers.

In (#11) Rick A. wrote:

'PS. Ron, I played a lot of 3B, a lot of SS, some 1B and a little 2B. But I hate to tell you I won't be eligible until about 1995. I will be honored to be on your ballot then!'

John Murphy wrote:

As for Stovey, he gets hurt because he played a plurality of games at first (where he was competing with the ABC boys and Orr). If he had played his whole career in left, he would have jumped over O'Neill fairly easily.

Stovey seems a lost cause when it comes to positional arguments. That's due to the fact that he would be moved around, both between seasons and even within seasons. Trying to piece together position changes based on baseball-reference shows:

1880: Starts in CF til Worchester gets Buttercup Dickerson so Stovey moves to 1B

I looked at my data to see which LF were arguable as good or better than O'Neill during individual years of his prime and found the following names that are not on your comparison list: Hardy Richardson, Henry Larkin, Harry Stovey, and Darby O'Brien (1887-1892). The first three were not primarily LF for their careers, but did play LF head-to-head with O'Neill during particular years. I would think that it would be essential to throw them into the mix of players that O'Neill is being compared with.

Eric, I did look at all those names that you mentioned above, but the combination of their actual worth in left field with the weights left them out of the running. Obviously, guys like Richardson and Stovey were greater players than Joe Hornung, but they did not have greater value (peak*career) in left field (Richardson) or played as many seasons with O'Neill (Stovey only played four seasons with a plurality of games in left: I don't include any of the seasons that he played in left field if he played more games at another position).

BTW, picking eight players to compare with is totally subjective on my part. I wanted a number where the results would be statistically significant (plus consistent throughout baseball history), but would not bog me down too much.

I looked at my data to see which LF were arguable as good or better than O'Neill during individual years of his prime and found the following names that are not on your comparison list: Hardy Richardson, Henry Larkin, Harry Stovey, and Darby O'Brien (1887-1892). The first three were not primarily LF for their careers, but did play LF head-to-head with O'Neill during particular years. I would think that it would be essential to throw them into the mix of players that O'Neill is being compared with.

Eric, I did look at all those names that you mentioned above, but the combination of their actual worth in left field with the weights left them out of the running. Obviously, guys like Richardson and Stovey were greater players than Joe Hornung, but they did not have greater value (peak*career) in left field (Richardson) or played as many seasons with O'Neill (Stovey only played four seasons with a plurality of games in left: I don't include any of the seasons that he played in left field if he played more games at another position).

BTW, picking eight players to compare with is totally subjective on my part. I wanted a number where the results would be statistically significant (plus consistent throughout baseball history), but would not bog me down too much.

Obviously, guys like Richardson and Stovey were greater players than Joe Hornung, but they did not have greater value (peak*career) in left field (Richardson) or played as many seasons with O'Neill

I should have pointed out that Stovey wasn't nearly as valuable a leftfielder as O'Neill was, too.

I'm with Philip, John. I think positional comparisons within an era can be taken too far because the sample size is too small. You're only comparing him to 8 other players.

Except that the results don't really change that much when you go beyond eight players. Trust me, I have tried bigger sample sizes without much changes in the numbers.

John, I think you are taken these position comparisons too far. Don't the 1990's versus 1970's shortstop comparison and the 1930's versus 1910's first base comparisons show that this method shouldn't be applied too strictly?

Well, I tried it with the ABC boys and they still had high marks in the ranking system so your misgivings seem to be unnecessary.

But let's take this theoretically: if ten more shortstops enter the league next year that have the same numbers as A-Rod, the latter's value at the position will go down. Without a doubt, A-Rod would not help his team win as many games as he did this year. Doesn't that have to be taken into account since wins are the ultimate goal?

DanG,

Thanks for the answer. I thought we were using last season of 10 or more games to determine eligiblity. I'll have to reread the eligiblilty thread.

He won't be anywhere near my ballot anyway, I just like to take a look at all the new eligibles as they come up.

Rick A.:

Read the thread "Discussion of Token Appearances" if you really want to know why we do what we do. (bedoobedoo)

(#44) - John Murphy:

I should have pointed out that Stovey wasn't nearly as valuable a leftfielder as O'Neill was, too.

That statement sounds like hyperbole, if you believe Win Shares. According to WS, O'Neill is rated grade A in the outfield (3.76 WS/1000, 33.6 career defensive WS, 8,933 est DI). Stovey is rated B+ (3.19 WS/1000, 26.5 career defensive WS, 8,318 est DI).

So, yes, O'Neill was the better defensive outfielder, perhaps due to having the luxury of specializing at one position. IMO, Stovey's deficit of 7 WS is more than compensated for by his versatility and much longer career.

WARNING: Data Dump to Come

Here's some information that people might or might not find interesting. Awhile back, I tried to replicate the statistical analysis Bill James did in the NHBA, in order to see which players were most affected by the subjective factor. I used the numbers as listed in the Abstract, and I believe I did the 1/2 pitching modifier, but nothing beyond that. Since there was some discussion of how he got the SS rankings that he did, I thought since I had the information, I might as well share it, for current candidates and some guys we will or might be considering in the near future.

Here's how I'll list them:

Do I need to say that Keeler came up short in every category? Even OBA, even runs scored (OK by one).

This is from Marc's #22, and the comparison is to Burkett. As long as he mentioned runs scored, let me bring up again R* - my gimmick of runs compared to what an outstanding number of runs would be in that league and year. Five players, this time.

Willie Keeler

That R* number for Joe Kelley, 1902 is way off - it should be 73. Upon revision, Kelley drops back behind (in extended-peak run scoring) such players as Fielder Jones, Mike Tiernan, and Mike Griffin.

John, how about Stan Musial. Will he make your ballot or will he be down there around 16th with Harry Stovey?

John, how about Stan Musial. Will he make your ballot or will he be down there around 16th with Harry Stovey?

1) Since I have Stovey at #7, I have no idea what ballot you are looking at.

2) Since Musial and Williams were the best leftfielders of their time, there is no way in hell that Musial would make it that low. That was an idiotic comment on your part.

3) Marc, if you can't discuss this rationally, maybe we shouldn't discuss it. You've insinuated in a number of posts now that I am an idiot. Maybe I am, but until someone can prove it with detailed analysis of what I am doing wrong, I am not changing my system.

I don't mind some playful jabs here and there. I've been known to do it on occasion (like with Jim above). But I've never been as rude as you have been to me.

I am open to any constructive criticism.

Jim,

Waddell's strikeout totals _are_ impressive, but they aren't all that _important_. In a low-hit, low-power, high-defensive-efficiency environment, strikeouts are much less valuable than in a high-offense environment.

Waddell was a unique talent, and a colorful character. I think he _should_ be in the Hall of Fame. But his talents didn't lead to a truly outstanding results, and that's what I think should matter for the Hall of Merit.

<i>BTW - I'm behind on getting the results up, but for the sake of discussion of 1916, the electees should be obvious to everyone.

Something interesting while looking at Cy Seymour's stats. Here are his year by year OPS+

As a pitcher

The methodology that Jim has been putting on the "pitcher questions" thread is quite friendly to Waddell. Among all pitchers up to about 1930, Waddell ranks 8th on Jim's scale. Those ahead of him are the 4 obvious 1st ballot choices, then Brown, Walsh, and Plank. Waddell is just barely behind Plank and clearly ahead of McGinnity, Griffith, Willis, Reulbach, Wood, and all the leftover 1880's guys.

If I understand Chris correctly, Chris is arguing that the sort of numbers that Jim is feeding into his scale overvalue Waddell, because Waddell wasn't as able to turn his physical prowess into wins for his team as were some less-gifted but smarter pitchers like Griffith.

From the distance of this many years, what sort of psychiatric or medical label would we put on Waddell if we were to dare to try? Perhaps the label would be "mentally retarded" or "developmentally disabled", but maybe it's not that at all - maybe it was more of an emotional-behavioral thing, and the heavy drinking didn't help.

If I understand Chris correctly, Chris is arguing that the sort of numbers that Jim is feeding into his scale overvalue Waddell, because Waddell wasn't as able to turn his physical prowess into wins for his team as were some less-gifted but smarter pitchers like Griffith.

Yes, that's what I'm arguing. I think there may be three factors involved here.

Rick A.

Well, we're going to have a guy in a couple of decades who comes off as this:

As a pitcher: 50, 189, 122, 162

I would think the SO would be more important then, since advancing runners and scratching out runs was important, and the SO does nothing at all for the offense.

That makes sense.

I would think the SO would be more important then, since advancing runners and scratching out runs was important, and the SO does nothing at all for the offense.

Not so.

You're looking at the matter from the hitter's side, not the pitcher's, and those perspectives are not the same. In the modern game, strikeouts are highly valuable for pitchers, but they're no big deal for hitters. In the deadball game, the situation is reversed. The hitter loses his chance if he strikes out, but if the pitcher allows the ball to be put into play, he doesn't lose very much.

Consider this: If you're the pitcher, what does striking out this guy accomplish?

1) It prevents him from hitting a home run. In the 1990s, that's a big deal. In the 1900's, it's not -- he's not going to hit a home run anyway.

2) It prevents him from putting the ball in play. In the 1890s, that's a fairly big deal, since the league may be hitting .300 on balls in play. In the 1900s, the league is hitting .260 on balls in play. The odds are still heavily in your favor if you let him put wood on the spit-laden, scratched up hunk of mush you're tossing up there.

From the pitcher's point of view, a strikeout is still a better outcome than letting the hitter make contact, but the consequences for the pitcher of _not_ getting a strikeout are substantially less negative, so the relative value of the strikeout goes down significantly. And if you tire yourself out by throwing hard all the time to get strikeouts, or if you increase the percentage of hits recorded on balls in play by giving the batters free kinetic energy to use, your pursuit of strikeouts might disadvantage you in other ways.

Without the home run as a likely outcome, the pitcher has less to lose by allowing the batter to hit the ball. We can argue about the finer points of just how likely a ball in play is likely to become an out, looking at rates of double plays, hits on balls in play and errors. I _think_ that such analysis would show that pitchers had less to lose by allowing the batter to hit the ball in the dead ball era than in later times, but maybe not. However, the lack of HR absolutely lowers the importance of the strikeout.

Consider

AL 1905 -- 43000 PA, 5000 K, 156 HR

This looks like the toughest year yet. I would consider Waddell a top flight candidate based on his seasonal ERA+ and IP, but I have a lot of doubts based on his W/L record.

Whadaya know - an issue I can address. Last year I invented my own stat - Run Support Index (RSI) which uses the game logs at retrosheet to determine a pitcher's run support for all his starts in his career, adjust them for park & league & centered on 100 (It's RS by his team divided by GS divided by Batter Park Factor divided by League R/G Avg, times 100 - round to an integer for a single season & to the hundredths place for career value). Now, with the pitcher's entering on the ballot I can't give perfect info, because the game logs at retro only list the starting pitchers for the game beginning in 1901, but that'll cover most of the careers of a lot of guys entering on the ballot. (NOTE: major flaw with RSI: it completely ignores relief appearances the pitcher may have made) Without further adu, here's the "career" RSI for some pitchers we're voting on:

Leever 118.77

I was working on a reply to Chris Cobb's #68. He partly scooped me with his #70, but I'll run it anyway.

2) It prevents him from putting the ball in play. In the 1890s, that's a fairly big deal, since the league may be hitting .300 on balls in play. In the 1900s, the league is hitting .260 on balls in play. The odds are still heavily in your favor if you let him put wood on the spit-laden, scratched up hunk of mush you're tossing up there.

I had to check to see how close Chris was with the numbers he threw into this.

In the 1896 NL (6.03 R/G) there were 55571 AB and 16136 H (a BA of .290). Just taking AB-H gives us 39435 known outs - 24.9 per game. Obviously, there are many outs not accounted for (as well as games shortened by weather and darkness). There were 404 HR and 3522 SO - the strikeouts account for 8.9% of the known outs. In AB that did not result in a strikeout, the BA was .310. If you remove the HR from that, the BA was .305 otherwise. (Of course, many of the HR were playable balls.)

In the 1906 NL (3.57 R/G) there were 39649 AB and 9693 H (a BA of .244). Just taking AB-H gives us 29956 known outs - 24.4 per game. The drop in known outs/G from 1896 to 1906 must come from the increase in outs invested in 1-run strategies. There were 175 HR and 4277 SO - the strikeouts account for 15.1% of the known outs. In AB that did not result in a strikeout, the BA was .276. If you remove the HR from that, the BA was .273 otherwise.

So Chris was about right with his numbers. Still, notice the overall change in strikeouts - from 9% of outs to 15%. There was a great deal of commentary at the time on the effects of the foul strike rule. What difference would that rule have made? Its main effect would have been to increase strikeouts.

John, I will try to speak english this time.My comments/questions concerning your prelim. ballot and the follow-up discussion are two in number.

1. Why is Tip O'Neill second? The only discussion I've seen is that OF in the '80s had some unusual circumstances and that O'Neill is better than Stovey.

2. Somebody else, not me, asked the following question, so I'm not the only one who is confused by your discussion of Stovey. That question was, is Stovey being penalized for being versatile? (I think that would have been a response to your comment that he was not as valuable as a LF as O'Neill, when I would wonder, well, was he not as valuable as O'Neill, period?)

Those were the two questions. The fact that I asked them was not meant to imply the answer or that your answer would be idiotic. Maybe I am the idiot because I don't already know the answers myself, but I don't.

Chris J., that's very interesting data! If it can be demonstrated, and it appears it can, that Waddell's run support, adjusted for park and league, was unusually low, that would account for some of the low performance of Waddell vs. an average pitchers with the team's run support, since his particular run support was, perhaps, lower than his team's.

Do you happen to have these numbers for Eddie Plank? It woudl be very helpful to see if more examples of how the numbers vary from pitcher to pitcher on the same team. McGinnity and Mathewson would be another interesting pair. From Leever and Willis, it looks like significant differences between the run support pitchers receive is typical.

Do you happen to have these numbers for Eddie Plank? It woudl be very helpful to see if more examples of how the numbers vary from pitcher to pitcher on the same team. McGinnity and Mathewson would be another interesting pair. From Leever and Willis, it looks like significant differences between the run support pitchers receive is typical.

The biggest teammate split I know of is still decades away - Gaylord Perry & Juan Marichal (Perry's support bad, Marichal's was Vic Raschi-esque).

McGinnity & Matthewson's RSIs from 1902 (not including McG's Baltimore half-season that year) until 1908:

......CM...JM

Chris J, that's great data. As I understand your method, a pitcher who is a good hitter would show up as having good run support. Any idea how much of the differences can be attributed to the pitchers themselves (as hitters)?

I have no idea whatsoever. Newcombe has the best RSI & he could hit (his 1955 season is tied for the 2nd best single season RSIs of the 1300+ I've gotten w/ at least 25 starts - & he hit .350 that year). My guess would be that a pitcher's impact is minimal. The phrase "good hitting pitcher" is a relative term. Don Drysdale had worse run support that Sandy Koufax while being a much better hitter. I'd WAG it's worth a few percent, but that's about it. It would proabably mean more back in the older days (re: the years were voting on now) because some pitchers hit good enough to get jobs as hitters (yea, Smokey Joe - I'm looking at you & your 112.39 RSI).

I was one of the original FOTO, and I used to vote for him in a block with Stovey and Browning. I gave up in 1910 because I thought there was no chance in the world that he'd ever see the light of day.

I am torn over this comment. Part of me thinks that we should NEVER be strategic in our voting, and views this statement as a break with the rules. Another part of me thinks that it is not strategic voting, since strategy can truly happen at the top of the ballot. Yet another part of me thinks that the "no strategy rule" is designed so that you don't leave a Cal McVey off your ballot just to get Bid McPhee elected, even though you think McVey deserves a higher spot.

I am not picking on Clint, because I have also partly used the "well, the group doesn't seem to support him very much" rationale for leaving a player off the bottom of my ballot. Herman Long, Mike Griffin, and Ned Williamson come to mind.

Anyone have any thoughts on this? Judge Scalia, I know that a literal interpretation of our constitution would require that Clint and I be shot, but how about some of the other voters? Or am I making too big a deal over spots 11-15 on my ballot?

FWIW, if anyone's interested in seeing the data themselves e-mail me & I'll try to send you the stuff (as much of it will fit on a disk - 2 of the 3 sheets of the excel file) when I get a chance. Sheet 1 is season-by-season, & sheet 2 is career. 119 pitchers (plus some random seasons in Sheet 1) including all 20th century HoF pitchers, all 20th century pitchers in the Top 100 of the NBJHBA, & all 20th century pitchers with more starts than Doyle Alexander plus whoever I felt like doing.

One problem I initially had with RSI was that as much as I like it, I initially couldn't figure out a way to convert it into W/L. Then inside a week, I figured out two different methods - neither are perfect but both have their advantages - we'll call them the Pythag method & the RSI method.

Pythag's easier (you don't even need RSI) so let's start with that. Take, for example, Don Drysdale's 1964 season. He went 18-16 with an ERA+ of 149. That's crud. The method here is the pythag therom (pretend I can spell, please) using Drysdale's RA/9IP & League R/G avg times Batter Park Factor. Then take the percentage & multiply by decisions. You find out he should've had a W/L percentage of 23-11 that year. He lost 5 wins. Ah, that was easy - no problems there, right?

Well, there is a problem & that problem is why I have a 2nd method to determine W/L. You see, in that 1964 season, the year Drysdale's run support cost him 5 wins, his RSI was actually 115. His run support was 15% above league average & taking those runs away from him isn't likely to help him win games. The problem with the pythag method is that it's actually accounting for two separate things: run support & variation. And that second factor can have some major impact. And for some pitchers, for some reason (& no, I don't want to get into that debate) tend to veer toward one direction throughout their entire careers. In English: some pitchers notably overshoot or undershoot their expected wins based solely on the pythag method.

So in comes the 2nd, much messier method. First, let's start off with the above math. Let's call the RA/9IP by PF*LR/G formula "A." You use pythagriam theorum here also. You pythag the pitcher's RA/9IP with his actually run support by his GS (for Drysdale in 1964 this would be 171 runs scored for him in 40 games). Put those through the therum (if I keep spelling it enough different ways maybe I'll get it right sometime) & you have a pitcher that should've had a .738 winning percentge - the closest W/L percentage would be 25-9. Not done yet now - this is just the prelims. It looks like his actual WPct was only 72% of his expected WPct based on his actual level of run support. So in figuring what his W/L mark would be if he had league average run support, multiple 72% by "A" (remember A - from the top of the paragraph?) Now we've got Drysdale going 17-17 if his run support dropped.

Which method is better? I may change my mind from one day to another, but in general I like the latter one because it's only focusing on run support & not variation. If a pitcher doesn't have much substantial variation in his career then it's a wash, but if he does have variation - maybe that says something about the pitcher. However, the most howling bad single season W/L come from the 2nd method. It's adjusts in extreme cases can be bizarre. Sometimes a small change in a pitcher's run support could have a large impact on W/L Pct & this 2nd method let's small changes stay small. An example:

1987 Nolan Ryan had a miserable year of run support - but not as miserable as you might think. His RSI was 77, which is bad, but by my reckoning 1 in every 28 pitcher seasons is worse - it's literally the sort of thing that happens every year to someone. But he went 8-16 with a league leading ERA. The first W/L method - the pythag method says he should've gone 16-8 w/ league average run support while the 2nd method - the RSI method - says he should've gone 10-14 with league average run support. I can't believe that a pitcher with a league leading ERA would ever have a losing record with league average run support. But such bizarre seasons are very rare which is why I generally look at both W/L methods.

In a little bit - or a much longer bit if I'm too sick of typing (expect the latter) - I'll post career W/L marks for the 119 pitchers (the turn of the century ones because we're voting on them & the others to give them more context) I've gotten. Maybe not today or anytime soon, but hopefully in this thread.

Hope that made some sense.

I am not picking on Clint, because I have also partly used the "well, the group doesn't seem to support him very much" rationale for leaving a player off the bottom of my ballot. Herman Long, Mike Griffin, and Ned Williamson come to mind.

Anyone have any thoughts on this? Judge Scalia, I know that a literal interpretation of our constitution would require that Clint and I be shot, but how about some of the other voters? Or am I making too big a deal over spots 11-15 on my ballot?

I think it's important that you show your best 15 in order. That doesn't mean that you should never be guided by the consensus. If you think a guy merits staying on the ballot, but the fact that no one else does leads you to doubt or change your judgment, then it's ok to drop him down. But if you're sure that he deserves the spot, then you should put him there.

There are two reasons for this. First, we're all trying to keep track of a lot of players, so there's a real danger that I'll forget about someone, or that I won't bother to go back and adjust my view of this or that old player, and if someone else is putting that player on the ballot and making the player's case, I'll be more motivated to keep my assessment of that player up to date. Mark McKinniss ranked Silver King #15 last election, reminding us that he looks good in new WARP, so I'm going to include him in my review of pitchers in new WARP. I may not look again at Bill Hutchison, however, because he's dropped off the ballot.

Second, in a close election, and we've had some very close ones, the 11-15 votes could make the difference, so the integrity of the election depends on the whole content of each ballot. So far no one who has lost a close election has not gotten in shortly thereafter, but I can imagine circumstances in the early 30s where, if a player doesn't get elected in a particular year, it may be a decade before the player has another chance, if ever.

And in any case, Ned Williamson and Herman Long _need_ your support :-) .

Long as I'm thinking of it - here's the career W/L for turn-of-the-century pitchers. First I'll do it using the Pythag method. They're ordered by Fib. Win Points. NOTE: All W/L these pitchers had before 1901 have not been altered in anyway & therefore their W/L marks are only partially adjusted.

Cy Young 471-225 (565 FWP, 98.61 RSI)

"Cy Young 471-225 (565 FWP, 98.61 RSI)"

No - except for the RSI, that's Walter Johnson's info. Cy Young was 524-303 by this method.

Chris J. - when you do the "Pythag" and other such W-L records, are you using the pitcher's ERA or RA?

OK John, your turn! I am not big on any of the newcomers, frankly. I see the best of them (Flick) as second-tier. So mostly, my 1915 ballot moves up, except that I have noticed that there are new WARPs floating around!

Must-Be-HoMers

1. Charlie Bennett (was 3)--perhaps the greatest position player 1881-83, second best peak of any eligible position player

2. Sam Thompson (4)--second best prime of any eligible player, short career maybe but his 9 year peak is as good as most and at a higher level

3. Dickey Pearce (5)--more total career value as of 1871 than anybody, and I would guess that no more than 3 or 4 eligible players had more career value as of 1916

Very-Deserving

4. Jim McCormick (6)--dropped a lot in "new" WARP but I will adopt "new" WARP slowly if at all; right now, I see him having the most career value among the eligible pitchers; heck, he led all active pitchers in career WS as late as 1886

5. Bob Caruthers (not rated)--I once had Bob as high as 3rd then became a skeptic; here is one case where "new" WARP made me take a fresh look; maybe I'll regain my skepticism, even by next Monday, but right now he looks like the best choice from the AA, not only among eligibles; I also think we're being too hard on pitchers

6. Harry Wright (8)--second greatest player pre-1871; what other eligible player, other than Pearce, is even the second greatest player of his generation?

7. Lip Pike (7)--huge peak and NOT a short career among the pre-1890 set

8. Tommy Bond (9)--highest peak on the board, and we're too hard on pitchers

9. Hughie Jennings (11)--#1 peak among position players

Deserving

10. Charley Jones (10)--his high rating was based largely on his WARP peak, I will need to think about whether "new" WARP supports such a rating, but for now I'll leave him where he is

11. Elmer Flick (new)--actually a nice comp for Charley Jones, so put him here for now; not a bad comp for Lip Pike either, but I frankly can't see Elmer move up that much

12. Harry Stovey (12)--not much to add to Harry's legend

13. Joe Kelley (13)--one of just three position players in my top 10 for peak, prime and career, all three (the others being Pearce and C. Jones, Kelley's claim to that status is actually the strongest of the three--i.e. least speculative--so he could move up) but he is not anywhere near #1 in any of the three either

14. Ed Williamson (14)--fourth best peak among position players; his credits for best player are as credible as J. Collins' as the greatest 3B of all-time, death-bed or no

15. Jimmy Collins (15)--this is every bit as speculative as my rating of Dickey Pearce, in that the numbers don't really support this high of a rating, but peer praise does; what I said about his accolades as the greatest 3B of all-time is a reference to the fact that men (e.g.

I'm using RA, not ERA. Those unearned runs count on the scoreboard, too.

1916 Preliminary Ballot, top half

I'm far from ready to place pitchers, but since I don't _think_ any are going to be in the top half of my ballot, I'm sending it out. It includes preliminary placements for Keeler and Flick.

Tough ballot this week makes me want to post a prelim for the first time in awhile. I've shaken things up a bit, but I'm still not happy with the number of outfielders on my ballot. . .

1916

1. Joe Kelley (6)--I?ve been underrating him I realize, and thus he shoots up to the top of the ballot. He beats Stovey (and Browning) on just about every measure if you discount the AA realistically.

2. Harry Stovey (3)--I think some people have been applying an awfully harsh AA discount to him. He was a tremendous hitter and looks great in WS pennants added and in the baserunning info that?s been posted. More career value than any of the other ?hitters? further down the ballot. I think he clearly stands out from the lower glut.

3. Lip Pike (4)--Tough to get a handle on him: not as good in the NA as McVey, but better before; much better in the NA than Start, not as good before.

4. Charlie Bennett (5)--Great defense and hitting (for a catcher) moves him in-between the top and middle parts of the OFG.

5. Willie Keeler (-)--Lotsa career value and a better peak than Ryan/Van Haltren puts him at the head of the OFG2.

6. Hugh Duffy (8)--Career value edge on Flick, Peak edge on Browning after the AA discount.

7. Elmer Flick (-)--A kind of jumbo-sized Sam Thompson: long, high peak with no padding to his career.

8. Pete Browning (7)--AA discount and short career drops him behind the rest of the mid-ballot OFG.

9. Jimmy Collins (9)--Comparable Peak and Career values to the OFG3. I don?t want to have positional quotas, but the fact that he?s the best at his position of his time (clearly) has to make him stand out from the identical outfielders.

10. Dickey Pearce (10)--The best shortstop of his time stays ahead of the OFG3. I really doubt Harry Wright will ever make it on my ballot though. . .I think I?m gonna pair him with Collins the way I?ve paired Grant with Childs, seems like a good comp to me. . .

11. Bob Caruthers (-)--Parisian Bob?s triumphant return to my ballot. Three high-peak players in a row here. Decided to rank them in order of their defensive value for this prelim.

12. Hughie Jennings (15)--Like Sam Thompson, only a slightly better peak and he was a shortstop instead of a right-fielder.

13. Sam Thompson (13)--Decided to rank Thompson?s peak ahead of Ryan and Van Haltren?s Career edge this year.

14. Jimmy Ryan (11)--I could see any of these Oufielders in any order, they?re that similar. Slight, very slight, edges in peak and career value place him ahead of Van Haltren.

15. George Van Haltren (12)--This gives me ten Outfielders on this ballot and that just feels really wrong. I think I may bump these last three off for Childs/Grant and Beckley, but I haven?t decided yet.

Re "Not Top 30" players:

I'm using RA, not ERA. Those unearned runs count on the scoreboard, too.

That would have been my choice. You're talking me into moving McGinnity up the ballot to some place he can be seen.

Ron said:

"If somebody tells me that's strategic voting, I'll stop doing it and I'll put Tip O'Neill, Dave Foutz, Jack Stivetts and others back on my ballot. But I hope that's not the case, and that this continues to be viewed as a prudent way to prune out the huge numbers of players at the tail end of the ballot."

In the 1911 and 1912 election, Bob Caruthers had dropped to 20th place, on only 11 ballots. I felt he was HoM-worthy, and continued to fight for him. The few other people with Caruthers on their ballots joined in, and eventually I think people are coming around. Don't know where he'll be this year, but it'll be a big jump from 20th.

I have had other players on my ballot (Ed Williamson, Bud Fowler), who I felt like I couldn't articulate as strong of a case for. They have since dropped out.

This is how I see it: I consider a lot of things besides pure numbers. For example, I put a lot of weight on whether a player was the best or second best at his position in a given year. I also give weight to Bill James' rankings. I also give weight to the Top Player lists of other experts. I consider the voting results of our electorate as an "expert" worthy of consideration. If I think a player is 10th or 11th, I will factor into my considerations that the Group doesn't consider him ballotworthy. That will effect my voting not because I am being Strategic, but because I respect the group decisions (except, as in the case of Caruthers, where I don't.)

yest, your comment was more about Bill James than about Willie Keeler. I should clarify. When I say that James didn't care about/did a poor job in dealing with the 19th century, I mean through the mid-'90s or so.

--He especially had trouble dealing with the really really big inning pitchers pre-'93 and that problem then goes away, so his rankings of Rusie and Nichols, e.g., are pretty good.

--He also makes no seasonal adjustment, short-changing the short-season guys. The first 140 game seasons were in 1891 and the first 150 game seasons in 1898 though there were still a few more <140 after 1898. But generally players whose careers began in the mid-'90s are not getting much of a penalty for short seasons. So I think his treatment of Keeler and Flick is pretty accurate.

To a very large degree I think the term "19th century" in MLB is shorthand for pre-60' 6", at least I often use it that way.

Pre lim ballot:

TomH - that's a fair point. Keeler is here as an offensive player, and what he's selling as an offensive player is lots and lots of batting average and very little else, apart from stolen bases. His .074 isolated power, given that he peaked in the 90's, is pretty puny. But batting average does have value. When the time comes, I'll take Tony Gwynn seriously. Now that I think of it, I think I'll rank Keeler behind Kelley, but probably ahead of Duffy/Ryan/Van Haltren.

December 9, 2003 (#88) - Chris J.

Extending myself ;-)

Prelim comments now, more to follow later:

I will have Flick and Keeler in my top 5, with Waddell in my top 10. Having taken another look at Caruthers, I will add him to my ballot but probably in my bottom 5.

Even if you believe that Keeler is over-rated, he was a great player nonetheless. I try to refrain from such statements, but you've got to be pretty crazy not to include Keeler on your ballot somewhere.

Flick was probably better than Keeler but his overall career stats look somewhat mediocre due to illness-induced early end of his career. Flick is a strong candidate for the least known Hall of Famer.

December 9, 2003 (#53) - OCF

As I mentioned in last "year"'s discussion, I have been following along with the process, but this is my first year submitting a prelim ballot. I would like to take part in the discussion, and then if the powers that be approve, begin voting in 1917.

For my analysis, I use the bare bones of WARP and modify them somewhat. I do this for 2 reasons: 1. Tango's insight into the proper use of FRAA instead of FRAR to generate wins above replacement, and 2. I wanted to ensure an accurate schedule adjustment, so I use the first-place games played figures from Paul W.'s website.

As I want to reward all kinds of excellence in my voting, I have tried to put together a rating system that considers both career and peak value, and value both as a whole and as parts. For example, consider 2 players, one who is a superior defender and an average hitter, the other who is above average in both areas of the game. While their WARP totals may be the same, I feel that there should be some bonus for being the best at a particular facet of the game, and want to rate the Ozzie Smith type higher.

My system takes every eligible player, as well as those already elected, and all players scheduled to come eligible by 1930 (in addition to any player who could be considered a star in the active period that I'm considering, in this case up to about 1916-1918). Currently I have 246 players in the database. I then rank them from 1 to 246 in each of 49 categories, then sum the rankings - low number wins. The high number of categories reflects an effort to balance peak vs. career, as well as pitchers vs. position players.

My 1916 Preliminary Ballot:

1. Willie Keeler

max, if you are going to put that little thought into your rankings, you might as well pcik your choices out of a hat.

Welcome, Max! Sometime it would be interesting to see the list of 49 categories that you use. My simpler numerical management skills limit me to many fewer, but I'd like to see what I'm missing.

It looks to me like your approach is producing highly defensible results.

I figured that with the change in Warp (which is the baseline of my system) I would make another change I was considering. Its basically a reweighting of different aspects. I reweight virtually every year, but this one is bigger. Anyway, I think it helps the peak guys out, at least, it helps Jennings. Caruthers is also helped, although finally getting new Warp stuff for pitchers into my system also made a difference.

1. Willie Keeler (-)

Max- Great looking ballot!

You must find it a challenge to compare position players with pitchers in your system, since I presume that not all your rating categories apply to both. How do you do it?

Whittling down my consideration list and applying various PA, adjWS, adjWARP numbers, Jim Spencer's pitching rankings, and gut feelings, gives the following prelim:

1. Willie Keeler - He may be overrated, but he has the counting stats, WARP, and WS numbers to rise to the top of this weaker class. I retract my earlier comment that he was like Burkett. Burkett was much better.

2. Charlie Bennett - He has fallen slightly out of favor with the group. The group rates Stovey, Kelley and Collins a bit ahead of Charlie. He is so much better than the other catchers, and will continue to be that way until Bresnahan. (Sorry, Johnny Kling)

3. Jimmy Collins - The best 3B, for an underrepresented HOM position. Similar arguments as Bennett, we won't see another electable 3B until Frank Baker. (Sorry, Larry Gardner)

4. Elmer Flick - His prime and peak, sometimes in very tough leagues, are outstanding. He far outshines Sam Thompson.

5. Harry Stovey - Harry is lucky he is this high. The new OF are making a strong showing at the top of the glut. Joe Kelley is pushing ahead as well.

6. Frank Grant - I agree with TomH, we will most likely overlook 19th century players when comparing them to 20th century stars. Grant has been acknowledged to be the best, he played for a long period of time, that's all we have. I for one will give him strong consideration. Will Sol White be paired with him next year? (albeit below Grant)

7. Joe Kelley - Could easily move higher on my ballot. I am more impressed by Kelley with a direct comparison to Keeler, although both are below Burkett. Keeler may be getting a slight underrepresented RF boost.

8. George Van Haltren - New WARP moves him ahead of Ryan, but not by much. His lesser peak keeps him down.

9. Jimmy Ryan - He just doesn't have enough of a peak, and his prime is lower than Van Haltren's.

10. Lip Pike - I am beginning to sour on Pike (and Stovey, for that matter). I suspect that it is the influx of 1890's and 20th century talent. I will work to keep Lip on the ballot. Excellent peak, against lesser competition.

11. Clark Griffith - Returns to my ballot for the first time since 1912 as the best available pitcher, thanks to new WARP and Jim Spencer. I am really interested in how PA treats Waddell, Willis and Joss.

12. Jake Beckley - No peak, decent prime, good career. I could leave him out entirely.

13. Bob Caruthers - I can't raise him much higher, because of his very short career in an inferior league.

14. Lave Cross - He may drop out, as I don't see him ever getting elected, and don't feel strongly that he needs a spot. I have been pairing him with Beckley.

15. Vic Willis - The new PA numbers could vault him higher or lower.

Just missing out

16. Fielder Jones - Both new WARP and WS like him a lot.

Thanks very much, red! I forget sometimes that you guys don't really know me, you don't know my sense of humor (I know, such as it is). Heck, some people who know me don't get my sense of humor. This is fun and every different point of view just makes it more fun. I sometimes try to inject a little light-hearted needle and I guess it doesn't come across as light hearted. Sorry, John. I really did NOT mean to imply that you are any less thoughtful or intelligent than anybody else here.

But you know, I'm here in Minnesota, and yes, I am sure the sun will come up. But I am vastly less certain that I'm going to SEE it.

As for his low win totals, his Cleveland team finished as high as 2nd, one time in the 9 seasons Joss excelled. I?d love to see his Run Support numbers as compared to others on that staff.

Here's R/G for Joss from 1902-10 alongside those for the rest of the team in the same years & in a third column Joss's OPS+ from that season:

......AJ......rest..OPS+

Paul, I was imprecise in my comments about season length, but the original question was about Bill James' treatment of the 19th century and whether he gives Keeler and Flick a fair shake. My response was to say that the obvious issue for position players is season length and James' apparent failure to account for it.

I didn't get the season lengths quite right but I think the larger point was substantially correct, acknowledging that everything is a matter of degree. To compare some of the OF glut:

Lip Pike (representing the '70s): in his prime (documented) seasons of 1871-77 his teams played an average of 56 games or 35 percent of a 162 game schedule.

Charley Jones (moving from the '70s to the '80s): prime 1876-86 (though only 9 years) his teams played an average of 91 games or 56 percent of a 162 game season.

Pete Browning and Harry Stovey (representing the '80s):

redsox, some consideration of IP in seasons and in career would be a helpful extension of your categories, which seem to favor short-career pitchers who lack any sort of decline phase. It's rather like the ranking you'd get of position players if you ranked them purely in terms of OPS+.

The differential between his career ERA and the league ERA was .79, which is the best on the ballot. The league ERA during his career was only 2.68, which is lower than any other pitcher is being compared to. It would seem to me that as that number drops it would compress other's ability to surpass it, making his .79 even more astounding.

Several pitchers besides Joss were able to post exceptional ERA differentials despite the low league ERA: Ed Walsh's differential was .81, Mathewson's .75, Waddell's .75 . Joss's ERA+ is great, but it's not so great as to separate him from other outstanding contemporary pitchers.

Offensive Winning Percentage Leaders

In any case [Addie Joss] needs to be seriously considered. That was the intent of my post.

I certainly agree.

Can someone explain to me why people consider "prime" to be a significant category. Why does it matter how people performed in their prime, which I take to be a certain age period around 27? Or do people mean something else, and a something else that isn't the same as peak? Or does prime mean consecutive year peak (which in fact is my definition of peak to begin with).

An old habit I just can't shake..

yest, I don't think I said Bill James' rating of G. Davis did NOT hold. He's got Davis #14, Keeler #35. Career WS 398-333.

I reserve the right to disagree with James. My larger point was that the earlier the player (in the 19th century) the less reliable James' WS numbers and/or ratings. By the '90s, his formulae seem to work a lot better. And I think he gets Davis, Keeler and Flick pretty much right. I think he rates Dahlen too low, BTW, but oddly enough Dahlen's numbers are very high, Bill dropped down due to the "intangibles."

daryn, I define prime as a players numbers in nonconsecutive above average full time years. Essentially, I drop the below average and partial years entirely when considering prime. GVH - big prime. JR - not as big as GVH, because he has lots of years below. I would be interested to hear others' definitions of prime.

I use a system not unlike Max's though certainly not with as many variables. Basically I rate peak, prime and career, with anywhere from 6 to 8 variables each.

For peak, I use the more or less customary 3 and 5 years though I flip-flop James' any 3 and 5 consecutive--I use 3 consecutive and any 5--and I use WS and adjWARP1 and a couple other measures.

But (in answer to daryn's question) because 3 and 5 year peaks are somewhat arbitrary--i.e. perhaps a given player does better on 4 and 6--I also use a flexible prime. That is I figure out the player's prime, again using both WARP and WS, which often disagree by the way. I figure a rolling average, and I figure generally that the prime has ended if and when the player falls below 75 percent of the average. I will keep 1 below average season if it then buys 2 more prime seasons. For most players, the prime will be 7 to 9 years, but there are plenty of 10-11 year primes, not many longer than that. In the prime category, I give credit for the length of the prime as well (obviously) as the average level.

I am still trying to refine career. What is awkward is including both counting stats and rate stats, because they often disagree, and create situations where a player's career rating sort of defies expectation. Maybe I will simply break the rate numbers out as a separate category--i.e. peak, prime, career and rate.

But in answer to daryn's question, peak is somewhat arbitrary though I will continue to use it because it is apples to apples. Prime is more flexible and conforms more to the shape of a career, but it is sometimes apples to oranges--i.e. comparing a 7 year prime vs. an 11 year prime at a lower level is awkward.

That's also how I define prime. Where peak is measured for a fixed number of years (most people look at 3 to 5 years), prime can be any number of years above a certain threshold. Prime is therefore also a measure of consistency at a very high level. BTW, I don't look at consecutive years. I don't see any more value in having peak seasons lumped together as opposed to spread out.

In (#121) redsox2004 asked:

How about throwing a shutout once for every 6 starts?

In the heart of the deadball era, was Joss tossing shutouts at an extraordinary rate?

Best Shutout Pct 1902-10, 20+ ShO

thanks for the explanation -- what you guys call prime, i call a sustained peak -- i'm a big fan of the sustained peak or the long prime.

redsox2004: I don't see Clark Griffith on your list. Is he really below Jack Chesbro in your system?