Expect a high, hard one near your noggin from Early's spectre if he doesn't go into the HoM early.

Hey, thanks for the "Early" thread! Hahahahahaha

okay....

player career IP . BP translated IP
Wynn..... 4564 ..........4473
Rixey..... 4495 ..........4153

Both loonnng careers. From the BP cards, Wynn has more 'translated' IPs; NOT a league qual adjustment, but adj for typical use in each pitcher's day. Of course, this doesn't necessarily mena it's right. Eppa missed 1+ years for WW1. Early missed 1 yr for WWII. Wynn's ERA+ is lower. He hit a bit better. Even BP, the 'timeline kings', show Wynn's quality-corrected DERA to be higher than Rixey's, 4.38 to 4.22, but Wynn's PRAR are much higher, which confuses me.

Even BP, the 'timeline kings', show Wynn's quality-corrected DERA to be higher than Rixey's, 4.38 to 4.22, but Wynn's PRAR are much higher, which confuses me.

This is how BP adjusts for changing balance between pitchers and fielders in defensive credit. As pitchers get a higher percentage of the credit, their replacement level drops, so being average is worth more runs above replacement. As fielders get a lower percetange, being average is worth fewer runs above replacement.

I think there are some logical problems with this arrangement when it comes to cross-era comparisons, but there it is.

His early "prime" years (ages 19-28) with Washington were very mediocre and generally don't add anything -- like the tail end of other pitchers' careers -- he looks better without them. Apart from the Senators being lousy, any other reasons for this? Are the park factors in ERA+ funny? Was there lousy defense?

His strikeout rates took a huge jump at age 30. His homeruns went up as well. This suggests that he became a different pitcher at age 30, rather than just being a change in defense/park.

Apart from the Senators being lousy, any other reasons for this? Are the park factors in ERA+ funny? Was there lousy defense?

None of that would explain the sharp increase in his strikeout rate. There's got to be some anecdotal evidence about him picking up some key pointer from Feller, Lemon, Hegan, some coach or all of the above.

(Dizzypaco beat me to it) :-)

being on a bad fielding team would increase a pitchers strikeouts because would be putouts are more likley to become hits or errors leaving the out count the same giving the pitcher more chances for strikeouts

For me the guy Early Wynn reminds me of most is Burleigh Grimes. Hey, they rhyme! They both have some nice, big seasons intermingled with a bunch of scrap-yard years. And they both threw an awful lot of innings with an over ERA+ and DERA in the not-much-better-than-average range. They both won 270 or more decisions with similar winning pcts, and they were about as good as hitters as one another. They both had strikeout rates about 10% higher than their leagues.

Two differences: Ironically Wynn hit about two-thirds as many hitters as the league, while Grimes hit a few more than average; Grimes walked about 2% more than the league, Wynn about 7% less.

Grimes threw a spitter, and IIRC, Wynn threw a wide assortment of pitches and may have been a reputed spitballer. I don't know if that's true or not, but I think I remember reading it somewhere, so take a huge grain of salt with it.

Another guy I can see in this same class of pitchers might be Bill Byrd.

being on a bad fielding team would increase a pitchers strikeouts because would be putouts are more likley to become hits or errors leaving the out count the same giving the pitcher more chances for strikeouts

Statistically, this should account for a small increase in strikeouts. Wynn had a massive increase in his strikeout rate, which cannot be accounted for by a bad defense.

baseballlibrary.com....

"Indians owner Bill Veeck obtained Wynn on December 14, 1948 with Mickey Vernon for Joe Haynes, Ed Klieman, and Eddie Robinson, one of the best deals in Indians' history.

Wynn came under the tutelage of Cleveland pitching coach Mel Harder, who taught the portly righthander a curve, knuckleball, slider, and changeup. Wynn threw all his pitches with an easy, effortless motion."

Very strange indeed. Three careers.

1939-49 (age 19-29)--about an average pitcher on OPS+, more BB than Ks (almost 2X in 1948 when he also gave up a career high 18 HR [to that point]). Median ERA+ 96, median OOB .340.

1950-56 (age 30-36)--a nice 7 year prime with the weakest year right smack dab in the middle. As noted above, Ks and HRs jumped, so I figured he had a better defense and he was just throwin' the ball over the plate. No, BBs went up, too. Median ERA+ 135, median OOB .304.

1957-63 (age 37-43)--a pretty steep decline with the worst years at the beginning. Ks went up even more, but so did HRs and the BBs, if anything, increased a bit, too. Median ERA+ 108, median OOB .323.

Clearly once he got to Cleveland, he kept on doing what he had been doing for one year, then learned and did something different. But what? And what, if anything, is more diagnostic--more HR, more K or more BB? I wonder if he just started throwing harder and took his chances. More BB suggest a willingness to be wild, more Ks suggest throwing harder, and more HRs suggest that some of those hard balls went straight throw the zone. Like maybe he'd been trying to be too cute in DC all those years.

As for his decline, he managed one more great year in Chi-town in '59. After 18 years in the bigs, no way does he get into Cooperstown without that one more big year. He was just too mediocre too much of the time to make it without 300 (okay, 295 might have been good enough) wins. If he follows his 14 and 14 wins in '57 and '58, and stays more in line with 13 wins in '60--and gets, say, another 14 in '59 and ends up his career with 292, I don't think he makes it.

I also see him among the long career borderliners. Being a peak/prime voter myself his prime ERA+ of about 135 is not terribly compelling, nor is anything else. And his career ERA+ of 106, which to me is the one single most important number for pitchers, does absolutely nothing for me. I don't think he'll make my top 50.

>Wynn came under the tutelage of Cleveland pitching coach Mel Harder, who taught the portly righthander a curve, knuckleball, slider, and changeup.

So he was a one-pitch pitcher in DC? Who the hell was the Senators pitching coach in those days?

Variations in Wynn's HR allowed, BB allowed, and K are all strongly influenced by league and park environments.

Washington was a very poor home-run park, but as a park with a lot of real estate in play, strong defense was important, and Washington didn't have it.

Cleveland was not a great home-run park, but home runs jumped everywhere in the late 40s and early 50s: a high-offense period.

BBs were also up a lot in this era.

To get an accurate chart of changes in Wynn's pitching style, you'd need at least to chart his BB, K, and HR rates against league averages. To really get it exact, park-adjusting HR would also help.

Two pieces of the picture that are important to Wynn's value that Sunnyday2 doesn't mention are innings pitched and OPS+.

Wynn was top 10 in IP every year from 1951 to 1960, leading the league 3 times. That's a workhorse.

His career OPS+ was 54, so he was helping himself a good bit over the average-hitting pitcher with the stick.

Taken in the context of his time, Wynn is an easy choice for the HoM.

Sunnyday2 wrote:

And his career ERA+ of 106, which to me is the one single most important number for pitchers, does absolutely nothing for me. I don't think he'll make my top 50.

I know it's early in the day for Wynn, but I hope you will re-think burying him like this. I'm not saying it would be a travesty to leave him off your ballot, but he's got to be a serious candidate. Consider this: on your current ballot, you have Rixey at 19 and Dean at 20.

You noted Wynn's 7-year prime. During that prime he had an ERA+ of 127 in 1803.7 IP. And then there was the rest of his career.

Rixey, in his 7-year prime 1921-27 had an ERA+ of 125 in 1906.7 IP. And then there was the rest of his career.

Dean, in his career, had an ERA+ of 130 in 1967.3 IP.

There are reasonable arguments that Rixey and Dean should rank ahead of Wynn. The rest of Rixey's career was better (though Wynn had more big years). Dean at his best was undoubtedly better than Wynn.

But to have this pair close to your ballot and Wynn outside your top 50? It's hard for me to see that there's a reasonable argument for that.

I should clarify. I haven't run Early through my full system yet, this was/is my preliminary impression.

On one measure I have him rated right in Rixey-Coveleski-Lyons-Ruffing territory. All but Ruffing are in my PHoM but only Coveleski among them ever made my ballot. Actually I see him more similar to Ruffing than the others. Rixey was never as bad as Wynn or Ruffing were for years at a time. Recently my pitcher list looked like this (not updated fully for a few years now).

1. Newhouser
2. Waddell
3. Joss
4. Bond
5. Lyons
6. Rixey
7. Lemon
8. Dean
9. Ruffing
10. Griffith

11. Cicotte
12. McCormick
13. Mullane
14. McCormick
15. Gomez
16. Faber
17. Silver King
18. Wilbur Cooper
19. Grimes
20. Welch

21. Bridges
22. Walters
23. Ferrell
24. Willis
25. Galvin

I think Wynn will be somewhere in the 9-12 range, which might get him in at #30-45ish. So I was probably too hasty throwing out the number 50. But he sure was lousy for a lot of years.

Or I coulda said he'd rank somewhere in the general vicinity of Richie Ashburn, who is around #40 or so.

Here are a couple of quotes from the Neyer/James Guide to Pitchers:

Mel Harder: "We got Early Wynn from Washington in 1949 and he was strictly a fastball and knuckleball pitcher. Those are the pitches he relied on. When he joined Cleveland, Early and I talked quite a bit. I told him he had to throw the curveball and a change-up and forget about the knuckleball. So he ended up being quite a pitcher because he improved so quickly."

Wynn: "I didn't have [a curveball] till I got to Cleveland and Mel Harder showed me how to throw one... Whenever I tried to throw a curve, I'd come up with a nothing ball, a bad slider. So I was just throwing fast balls and change-ups. I had a knuckler, but I just threw it on the sidelines. That's another pitch I didn't really use until I got to Cleveland."

Of course, the two quotes are inconsistent about his use of the knuckleball, but several other quotes suggest that he continued to throw a knuckler during his Cleveland years.

Early Wynn and Burleigh Grimes rhyme?

Am I mispronouncing Early's surname?

Early Wynn and Burleigh Grimes rhyme?

I had the same reaction.

Burleigh rhymes with Early

Early rhymes with Burleigh. Unless I'm mispronouncing them. I don't know if I've ever heard Grimes' first name.

Uhr-lee
Buhr-lee

Right? Or did he pronounce it buhrl?

oh, I thought you meant wynn and grimes. I figured that maybe Wynn was pronounced Whine, which would make it kinda rhyme with grimes.

The lesson, as always....I am an idiot.

How about Grynn and Whimes. Sounds like the authors of some fairy tales or something.

Or how about Grynn and Barrett. Or Gwynn and Barrett.

Of course, we are no longer talking rhymes here, buster, just stringing homonyms together. Homo-what? Never mind.

RA+ Pythpat: 269-238.

That's taking a competition discount for 1943 and 1944. I think we could reasonably project credit as something like 12-12 for 1945; that would bring him to 281-250. I haven't accounted for his hitting; it's possible that might add another couple of wins.

His best years by this were 21-10 in 1955 and 19-11 in 1954.

He's at 8.39 IP per decision, which is low side (or put another way, he had a high number of decisions per inning). That's why his equivalent record in my system has fewer decisions than his actual record.

This is superficially similar to, say Jack Powell (equivalent record 263-225), but Powell pitched when pitchers threw more innings per year; relative to his own times, Wynn was more of a workhorse and had a longer career.

He and Spahn are the two 300-game winners of his time, but otherwise there's little comparison. I have Spahn at an equivalent 340-242, which is a very different neighborhood.

The (likely) election of Rixie would be something of a precedent for Wynn, but I'd probably put Rixie ahead of Wynn.

Everybody's throwing their Wynn comparisons out there, so here's mine:

To me, Wynn resembles Red Ruffing (minus the hitting) more than anybody else.

WARP-1 seasons in parallel (decimal points omitted for brevity):
Wynn(top-11):112, 91, 89, 80, 79, 77, 77, 72, 67, 54, 53,
RedR(top-11): 91, 87, 85, 81, 80, 78, 71, 57, 55, 51, 49,

Wynn(rest): 44, 43, 40, 37, 30, 28, 24, 22, 21, 15, 14, 1
RedR(rest): 48, 47, 43, 43, 39, 32, 30, 23, 22, 1, -3

Wynn was better than Red when comping their top-11 seasons; Ruffing wins the comp for the rest of their careers.

Note: If you threw Rixey into that mix, Eppa wins season 10 (5.7), edges Red for 2nd in seasons 8, 9, 11, and is 3rd all of the remaining years (ties EW in 16 and 17). Rixey's career is both shorter and has less peak (and that's by WARP-1).

That's pretty bizarre, because according to my calculations, Rixey allowed 623 fewer runs than a replacement-level pitcher while Wynn allowed 500 fewer and Ruffing 535 fewer. Apparently, for Wynn's career, they put replacement level as 2.26 runs/9IP below average and at 1.63 for Rixey's. These seem extremely high. I use 0.75 for all periods.

Even BP has Rixey with a lower career DERA adjusted for all-time, which I take to be WARP-3. Where do you get the WARP-1 numbers, because I don't see them on the player pages at BP?

Yes, but doesn't WARP adjust replacement level downward as you get closer to the present?

Look at more modern long-career low peak pitchers:

Tommy John (top-11): 88, 79, 74, 66, 64, 64, 64, 60, 51, 49, 48
Don Sutton (top-11): 85, 78, 74, 64, 61, 56, 56, 55, 55, 55, 53


John (rest): 46, 44, 44, 42, 42, 42, 38, 34, 31, 30, 25, 14, 7, 7, 3, 3, -1
Sutton (rest): , 51, 51, 48, 47, 46, 43, 38, 36, 36, 29, 26, 7

John - 1188 PRAR
Sutton - 1326 PRAR

Rixey - 1019 PRAR
Ruffing - 1072 PRAR
Wynn - 1244 PRAR

Heck, here's everyone over 1000 PRAR who will be eligible through 1980:

Pitcher         IP       PRAA   PRAR
Johnson, W      5914.67  807    1856
Young, C        7356.00  804    1845
Spahn, W        5243.67  335    1605
Alexander, P    5189.00  605    1508
Roberts, R      4688.67  286    1422
Grove, L        3940.00  523    1352
Wynn, E         4566.00   97    1244
Mathewson, C    4780.33  456    1215
Gibson, B       3884.33  328    1195
Nichols, K      5056.33  449    1167
Feller, B       3827.00  253    1156
Lyons, T        4161.00  243    1137
Bunning, J      3759.33  199    1075
Ruffing, R      4344.00  147    1072
Plank, E        4495.67  374    1049
Pierce, B       3306.67  199    1041
Newhouser, H    2993.00  302    1025
Hubbell, C      3590.33  306    1023
Rixey, E        4494.67  206    1019
Faber, R        4087.00  217    1010
Ford, W         3171.00  238    1002

It seems that the long-career modern pitchers can accumulate great PRAR totals despite being relatively average over their careers.

My main conclusion from this data is that I trust WARP for pitchers as much as I trust WS. (However, they eerily offset each other, with WS favoring older pitchers and WARP favoring newer pitchers, so maybe I should just prorate them and add them together :-).)

a few non-WARPian or WS-ese rankings on Wynn
stats from 1943-1961

most wins
Spahn 309
Wynn 279
Roberts 234
Lemon 207

most IP
Spahn 4329
Wynn 4039
Roberts 3740

so he's certainly a fine career-value candidate

By RSAA, runs saved above average (a more prime-friendly measure), Wynn is 15th.

One attempt to combine them: give 1 RS for every 20 IP. This means that instead of 'average', it is 'runs saved above 0.45 below-league-average-ERA'

Spahn way lots
Newhouser lots
Pierce 371
Roberts 351
Wynn 337
Lemon 322

One's view of Early will be pretty much dependent on one's pref for peak/prime/career.

As a primer, I've got:

1-2-3. Waddell, Joss, Bond
4-5-6-(7?). Rixey, Ruffing, Griffith
Dean somewhere among Rixey, Ruffing and Griffith

And Wynn probably somewhere among Rixey, Ruffing, Griffith and Dean, but probably toward the tail end of the cluster. He was pretty darn ordinary for 2/3 of his career. I guessed earlier that he would be around #50 on my ballot. Rixey through Griffith are more or less in the 20s so Wynn is probably more like #30 but that is not to say he couldn't PHoM someday as the 6th or 8th pitcher on my list.

Where do you get the WARP-1 numbers

Where they always are, in the Advanced Batting Statistics (which aren't just batting, because they include FRAR and PRAR).

Yes, but doesn't WARP adjust replacement level downward as you get closer to the present?

I think the defensive replacement level remains essentially constant, just like the offensive replacement level.

However, as "the three true outcomes" become a larger and larger percentage of the end product of a PA, pitchers become less and less dependent on the fielders behind them. Defensive responsibility (and value) moves from the fielders to the pitchers. Fielders get less FRAR; pitchers get those defensive runs as PRAR (or vice-versa).

Also, don't forget that the WARP-1 numbers include hitting and fielding. Ruffing was 123 runs better than Rixey as a hitter over his career; Wynn is between the two, 66 runs better than Rixey. Rixey appears to have been a good fielder though, Wynn slightly below average, and Ruffing had a problem with it.

He was pretty darn ordinary for 2/3 of his career.

And that's where you and I differ on evaluation. To me, that's irrelevant. I only care about the good parts. It doesn't matter to me whether he was average or replacement level or in the minors or broadcasting during the seasons that aren't the good parts. Wynn is just making a living when the White Sox want him to eat innings as an average pitcher. Wynn is just doing what he's told to do when the Senators have him pitching in the majors before he is probably ready. I can't count it against him.

In my system, a player can't play himself out of the HOM once he has established a HOM quality portion of a career. It's why I don't use career rate stats (except in arguments to persuade people who do).

I posted this on the ballot discussion, will add it here as well:

I really hope Early Wynn doesn't end up as borderline. The guy won 300 games. Which doesn't count the year of war credit he should get for 1945.

He's got a peak too, from 1950-56. He's got a Cy Young Award outside of his peak (1959), which is typically a sign of a great player (able to win an MVP or Cy Young while not having his best season). He also led his league in pWS twice outside of the year he won the Cy Young.

I know the career ERA+ is only 106. But he threw 4564 innings (without credit for 1945). He had 5 years from 125-154 ERA+, which is the same total as Bob Lemon (with a higher peak, Lemon ranged 133-144). He was a workhorse, 8 times top 3 in his league in strikeouts, 7 times top 3 in his league in innings. He led the league in starts 5 times. He could hit a little too (2.14 career, OPS+ 54, 17 career HR, 7 years over 70 OPS+).

He's not Walter Johnson or Lefty Grove. But he is Red Faber, Pud Galvin, Ted Lyons, Eppa Rixey. He's Bob Lemon, with an extra 2000 or so innings (given war credit).

He should be a very easy choice, once he's on a ballot that doesn't fill elect me spots with no brainers.

I'll also add that this is very important:

You noted Wynn's 7-year prime. During that prime he had an ERA+ of 127 in 1803.7 IP. And then there was the rest of his career.

Rixey, in his 7-year prime 1921-27 had an ERA+ of 125 in 1906.7 IP. And then there was the rest of his career.

Dean, in his career, had an ERA+ of 130 in 1967.3 IP.

He's a notch behind Dean for prime, but not very far behind. Then throw in 1943, 46-47, 59-61 and you have an immensely more valuable overall career, with nearly the same peak.

Sure there are a few bad years thrown in, but he was still eating innings, I can't give him a negative for those years. They just don't add anything.

Take out 1942, 44, 48-49, 57-58, 62. You are left with a pitcher that is 228-137, or basically, Stan Coveleski (215-142), plus a 13-(-5) season. The 72-107 is basically innings eating replacement level, there's no way there should be a negative there.

More perspective: Lemon was 207-128. Wynn sans the bad years at 228-137, is basically Lemon + a 21-9 season. And Lemon's peak is not any higher than Wynn's, as a pitcher. Wynn had a higher best ERA+ season, and his top 5 are in the same range as Lemon's. Lemon was a better hitter (which gives him peak WARP scores about 1-1.5 higher), but Wynn was no slouch.

Joe,

While you make some very good points and your arguemnts are informative, I think that Wynn is destined to go in as soon as there is an open spot. I just can't see enough people leaving him off of their ballot. Especially after electing Ruffing and Rixey in the last five years.

in others words don't worry, be......cheerful?

I bashed Wynn pretty good early on and yet I now have slotted him in as the #5 pitcher on my ballot. Here's where I'm at, with OPS+, IP and WS data. Obviously I prefer pitchers who are effective when they pitch, to pitchers who necessarily pitch a tremendous lot. (Mendez and Redding are also right up at the top but do not fit into the following "chart.")

1. Waddell 180-79-65-53-27-25-23-21-8-2/135 in 2961 IP
#53. Waddell 240/35-33-32/145/30.6

Two knocks on Rube. UER: I don't see them canceling out the rest. Head case: He was, but it's already in the numbers.

2. Joss 205-159-52-50-37-31-30-24/142 in 2327 IP
#80. Joss 191/35-28-25/131/31.5

2300 IP is enough to validate the 142--IOW, enough of a sample to know he was a hell of a pitcher.

3. Bond 167-41-33-26-15-2-(85)-(79)/110 in 2780 IP
#97. Bond 243/60-50-47/225/31.5

I give half his WS to his defense and he still has an incredible peak, even by the standards of his day. And his best years were in a one-league environment--not only that but a six-team, one-league environment. Numbers I've seen suggest the NL of '77-'78 was as good as the NL would be again until after the AA went away.

4. Dean 159-48-35-24-19-14/130 in 1967 IP
#25. Dean 181/37-31-31/145/24.1

OPS+ suggests he is no Koufax or Joss. James' rating of Dean vs. Waddell and Joss tells you everything you need to know about his method. Yes, it is essentially a peak rating, which I like, but his timeline obliterates everything else. You can guess how I feel about that.

(4a. Rixey)
(4b. Ruffing)

5. Wynn 154-42-35-35-26-19-15-10-8-2-(8 seasons below average!)/106 in 4564 IP
#47. Wynn 308/28-25-24/110/24.6

Not much of a peak, so the fact that I would have him #5 is high praise indeed. Surprisingly, more WS/IP than Diz!

6. Cicotte 185-75-74-55-32-29-16-15-1-(4 seasons below average)/123 in 3226 IP
#50. Cicotte 247/35-32-27/124/28.8

Now that I look at it, how is Cicotte not better than Wynn? Well, 1000+ IP by Early. But there was no time at which Wynn was actually as good of a pitcher as Eddie.

7. Griffith 191-34-31-30-28-20-18-16-14-9-(81)/121 in 3386 IP
#70. Griffith 273/34-32-30/143/29.2

His #70 rating is serious timeline abuse. Don't see how he could be above Cicotte however. The Old Fox was better in seasons #1 and #7 on, but #2 through #6 is all Eddie. Still I could see Griffith ahead of Wynn, too, though I understand there are about 1200 reasons to prefer Early.

8. Mullane 159-59-35-34-32-26-12-7-4-(95)/118 in 4531 IP
#82. Mullane 401/58-55-46/183/29.4

Serious competition issues (AA discount) but did 159-26-7-4 in NL after age 31 and 3000 IP. Better than Mickey Welch even with the AA discount.

9. Gomez 191-74-49-35-28-27-22-6-5-(97)/125 in 2503 IP
#67. Gomez 185/31-29-20/106/28.4

Light workload means light WS peak, but very very effective.

10. Jim McCormick 171-28-27-22-20-18-10-7-3-(89)/117 in 4276 IP
X. McCormick 334/54-53-42/223/23.4 acc. to my calc

This is after I threw out his OPS+ 166 in the UA. Not dissimilar to Mickey Welch, which is why Welch isn't near the top of my ballot despite his W total.

(10a. Faber)

11. Welch 160-42-30-19-17-14-12-11-(3 seasons below average)/113 in 4802 IP
X. Welch 354/57-46-42/191/22.1 (my calc)

McCormick threw from 1878-87, Welch from 1880-92. For their careers I don't see any significant difference in competition, but for peak McCormick's came in the early '80s when the AA was still a very inferior league, Welch's came after the AA had caught up--IOW in a NL that was probably not quite as good as in McCormick's peak years. I don't see how anybody can have Welch above McCormick. Yes he started an additional 70 games, that's cool, and pitched an extra 500 IP, and won an extra 40 games. The 40 wins reflect team quality, and McCormick was just as effective, and against better competition at his peak.

There you have it. Just for the record, the deeper backlog is: W. Cooper, S. King, Bridges, Walters, Grimes, Willis, (Galvin), Will White, Luque, H. Vaughan, Warneke, Mays, Ferrell, Shocker and Whitney.

Upcoming newbies who are on the radar are (in order) Spahn, Koufax, Ford, Roberts/Drysdale, Bunning, Pierce. All but Pierce and maybe Bunnng are top 10 material (I mean relative to the pitchers listed above).

OK, since I made the Burleigh Grimes comparison, I should be the one to do this...

I'm not a FOBG, nor an EOEW, i'm more like PBEW&BG;---Puzzled by Early Wynn und Burleigh Grimes.

Doing the good seasons vs. bad seasons thing.

Here's Grimes in all seasons where his ERA+ was above 99:

11 years, 326 GS, 193-116, 2711 IP, 775 BB, 984 K, 3.12 ERA, 119 ERA+

And Grimes in the bad seasons:

8 years, 171 GS, 77-96, 1469 IP, 520 BB, 528 K, 4.28 ERA, 83 ERA+


Now comparing that Early Wynn.

Years above 99 ERA+:

14 years, 371 GS, 211-123, 2861 IP, 1103 BB, 1507 K, 3.03 ERA, 119 ERA+

Poopoo years:

9 years, 241 GS, 89-121, 1703 IP, 672 BB, 827 K, 4.40 ERA, 83 ERA+

So what I get from this is that there's basically no difference between them in terms of prime effectiveness, but there's about 100 more good innings in Early Wynn's arm and 250 more bad innings in Early Wynn's arm.

Both pitchers have ten full seasons of starting in their good years (meaning 200+ innings). Wynn's best five ERA+s average out around 128 (rounded up from 127ish). Grimes's best five ERA+s come out at 129 (rounded up from 128ish). After that, Grimes has one other full season above 108 (a 123 season). Wynn has three full seasons above 108 (110, 115, 118). Grimes has one additional quarter season at 114. Wynn has two half seasons at 110 and 111 as well as two quarter seasons at 258 and 159.

Wynn has more total effective seasons, but Grimes is packing more of his best work into fewer total seasons and coming up only 100 innings short of Wynn's total in 3 fewer years.

In addition, Chris J's RSI numbers for the two don't suggest much other differentiation: Grimes had a 107 RSI, and Wynn had a 106 RSI. These numbers were probably influenced a bit by their own hitting. Grimes sported a stellar (for a pitcher) 58 OPS+ and Wynn a similarly good 54 OPS+.

So I can see Wynn being placed just ahead of Grimes by timeliners or on the strength of having a year or half a year more good years, but only by a slot or three. He's not any better than Grimes, especially considering that Wynn had more bad years. And not just bad but really bad. Here's their bad years with innings and ERA+ noted:

EARLY       BURLEIGH
ERA  INN    ERA  INN
--------    --------
96   208    98   311
96   165    87    83
95   252    85   259
89   240    85   181
87   168    83   247
86   263    80   194
75    20    79   141
74   198    67    53
71   190

Anyway, I guess I'm saying that if you're taking the approach that Wynn's bad years are just "eating innings" as has been written a couple times in this thread, then to avoid discontinuous logic you should do the same with Grimes and rank them near to equally.

Sorry, the Early/Bearly table again. I didn't eralize the code was case sensitive.

Doc,

I don't think you can do a simple IP comparison between Grimes and Wynn during their primes. During Grimes prime, starters had more IP per season than during Wynn's prime. This has to be taken into account otherwise you are during an inadvertant reverse timelining, kinda like taking WARP3 at face value when comparing them. How often did Grimes lead his league in IP or was at least in the top 5. I dont' have an answer but I bet that it was less than Wynn did. This has to be adjusted for and once it is, Wynn's good seasons become more valuable.

Plus, while we have no way of figuring this out, shouldn't Wynn get a boost because he played in an era where pithers were more valuable in comparison to their fielders? I don't feel that this is timelining so much as it is a reflection of how the game was played.

During Grimes prime, starters had more IP per season than during Wynn's prime

This is the big issue, IMO

IP ranks:

BG - 1, 1, 1, 3, 3, 4, 7, 9, 9, 9
EW - 1, 1, 1, 2, 2, 3, 3, 6, 6, 6, 6, 7

They were equal workhorses at their height, but the longer you go, the more Wynn separates himself in terms of workload in context.

The average #1 starter in the 1920s NO was tossing about 275 innings a year.

I don't have that data for Wynn's time, so I can't compare, but I'd bet that while the innings did likely decline somewhat, they didn't go down much. 25 tops, probably less.

The big decline in innings happened in the early portion of Grimes' career, then IIRC, things kind of plateaued a while before rocketing back up in the 1960s.

Of course league leaders were throwing more.

On the other hand, it may be worth noting too that Grimes finished 94 games, Wynn 66, and Grimes pitche more in relief, potentially effetive his innings totals.

Anyway, what I'm not trying to say is that I know they were exactly the same, what I'm trying to get at is that they were very similar, and I don't see how someone could figure Wynn to be #3 but have Grimes off in never-never land. Same strenghts and weaknesses, same kind of career arc. Was Wynn a little better? I wouldn't be surprised, but I also don't think he was tons better.

He's not Walter Johnson or Lefty Grove. But he is Red Faber, Pud Galvin, Ted Lyons, Eppa Rixey. He's Bob Lemon, with an extra 2000 or so innings (given war credit).

Considering that none of the pitchers in that second sentence but Galvin (who may be a mistake) have made, or are terribly close to making, my pHOM, I feel more justified than ever in burying Wynn. Then again, I'm perhaps the most extreme peak voter here, so I suspect you weren't talking to me, really. :)

Given the voting patterns of late, I'd be very surprised if it took Wynn very long to be elected. Medwick may be the peaksters' last hurrah (other than no-brainer candidates) for a while.

>Medwick may be the peaksters' last hurrah (other than no-brainer candidates) for a while.

;-(

Mark, are you considering burying Wynn because he didn't have a strong peak, or because he pitched a long time outside of his peak? The former might be defensible, but I don't think the latter is.

I'll repeat and extend some comparisons between Wynn and Dizzy Dean that I made, noting that Dean placed 11th on your ballot in 1968.

Wynn has a seven year peak in which he had an ERA+ of 127 in 1803.7 IP. And then there was the rest of his career.

Dean, in his career, had an ERA+ of 130 in 1967.3 IP.

Let's cut this to 5 years, since that's what's generally agreed to be Dean's peak.

Dean, 1531 IP, 130 ERA+, ranked 1, 2, 3, 1, 1 in IP

Wynn 1315.7 IP, 126 ERA+, ranked 3, 6, 1, 7, 2 in IP

Dean's peak is better, but is the difference such to justify dismissively burying one player while putting the other on your ballot, especially when the player to be buried has three more seasons at a similar level of quality (1950-51, 59), while the on-ballot player has no other full seasons pitched?

What are you looking at when you consider Wynn's peak?

Anyway, I guess I'm saying that if you're taking the approach that Wynn's bad years are just "eating innings" as has been written a couple times in this thread, then to avoid discontinuous logic you should do the same with Grimes and rank them near to equally.

That's also my conclusion, Eric, though I keep coming up with Grimes over Wynn, mainly due to the effect of the Lively Ball Era had on the arms of those early 1920's pitchers. Of course, WWII affected its generation of pitchers in a different way, but we can give them pitching credit for that.

I'm still going over the numbers, though. BTW, there will be some changes on my next ballot.

Of course, WWII affected its generation of pitchers in a different way, but we can give them pitching credit for that.

With the one year of war credit, the two years of war discount and the overall mediocre level of his play in the 1940s, the whole war thing turns out to be a wash for Wynn in my opinion.

Mark, are you considering burying Wynn because he didn't have a strong peak, or because he pitched a long time outside of his peak?

The former; a strong peak is pretty much the be-all and end-all for me.

As for comparing him with Dean, I'm more a WS and to some extent PRAA guy; I'm not a big fan of ERA+. And Dean's 37-31-31 three-year WS peak trounces Wynn's 28-24-24. With PRAA it's something like 49-38-37 to 38-24-24. Yes, there are league strength issues here, but I don't find them significant enough to come close to erasing that difference. And the extra low-20s WS seasons that Wynn has do nothing for me.

I know peaks this short aren't enough for most voters, but they are for me. (Obviously, Koufax will fare well with me when he comes around.)

I haven't set Wynn's final placement yet, but it looking to me now like he'll end up at the bottom of my top 50 or just outside it, somewhere around Lemon and Newcombe and Matlock and Carl Mays. I have him slightly above both Rixey and Ruffing, if that makes you feel better. ;)

And Dean's 37-31-31 three-year WS peak trounces Wynn's 28-24-24. With PRAA it's something like 49-38-37 to 38-24-24.

Let's try RA+-equivalent FWP. Several pitchers sorted best year to worst (not necessarily consecutive).



"Ruffing*" is the offense-adjusted version. I haven't made an offense-adjusted version of Wynn; it wouldn't be as dramatic, but he would improve a little.

Let me emend my previous post--I shouldn't imply that I ignore ERA+ entirely. But I do weight it far, far less than either WS or PRAA.

Chris's post points out that ERA+ isn't high on Dean, and that's a good point; I'll take it into account in my placement of Wynn, and in a general pitching reassessment I'm doing this election. But I don't think that's going to do much more in the long run than get Wynn just above #50, rather than just below.

Rixie's advantage on Wynn in the RA+ system lies in about his 8th best through 15th best seasons. Wynn had a better top few seasons, and both have some junk at the bottom end. Of course, they are of such different generations that all such comparisons are suspect for several reasons.

It should be noted that I have Robin Roberts at a career equivalent record of 295-226, which is no-doubt better than Wynn's 281-250 (which includes a hypothetical 12-12 for 1945). Roberts ahead of Wynn for both peak and career. (And, of course, Spahn ahead of Roberts, at least for career.)

Dr. Chaleeko good job on the Grimes/Wynn comparison. I've got Burleigh in the mid-20s/low-30s IIRC, I'll be moving him up some.

A minor nit to pick - it's 150 IP difference at their peaks, not a 100.

Using translated IP, which I think adjust for era very well, Wynn threw 2976.7 tIP in his +99 years, Grimes just 2515.3. That's two full seasons as a good pitcher more, about 1/6 more. That's pretty signficant. Top 5 each (only counting the +99 years):

Wynn 290.3, 288.3, 285.3, 280.7, 276.0
Grimes 308.3, 287.0, 271.0, 267.3, 261.7

I don't think it's a factor at all that Wynn had more bad years. The bad years are basically a 0. When you multiply a bigger number by zero, you still get zero.

So I still have Grimes signficantly behind Wynn. And I feel much more solidly justified in that ranking. I'll move Grimes up a little bit, but Wynn should be significantly ahead.

OCF, agreed on the ordering of Spahn, Roberts, Wynn.

I had Rixey No. 1 last time, and am leaning to Wynn at No. 3 this time. I lean to Wynn over Rixey, but they'll quite get compared directly.

I think the electorate is beginning to realize the wisdom of neither penalizing the long-career guys nor especially rewarding them.
By the former, you vote against a guy because his ERA+ or W-L Pct isn't that great, for instance.
By the latter, you go by counting stats like Ws and IPs too much.

I've always tended to 'zero out' the meanningless years. That used to seem to some to be a burden on the long careers. But it also boosts up their rate stats, and clarifies that in many cases they are Doppelgangers for 'peak' guys who didn't have tails to their careers.

guh, "not quite get compared directly"

Why wouldn't a guy like Mickey Welch compare more favorably to Wynn?
He got his 300+ wins in 10 less seasons with an additional 7 points of ERA+
Mickeys's bbref #s look better than Wynns; HoF stds and HoF monitor are much stronger for Welch.
Was it harder to win 300 in the 1940's and 50's than 1880's?
Seamingly so I'd say.
That said, Early hung around through a bunch of lousy years to get his 300.
1959 is Wynn's salvation in my opinion; it won him the right to have 4 more mediocre years to reach his 300 total.

Hung around?

He had two years at 37-38, where he had an 86/89 ERA+. It's not like he had an ERA of 7. He threw 500 innings those years and Chicago traded Minnie Minoso for Wynn and Al Smith in the middle, so they must have thought he had some value.

Then he goes out and wins a Cy Young Award for a team that won it's first pennant in 40 years, and follows that up with 2 more productive years, though he got hurt in 1961.

In 1962 he tried to come back and pitched badly but not God Awful, and then he won his 300th game in 1963 while posting a 159 ERA+ in relief.

The whole Early Wynn hung around myth is great exaggerated. Also, give him back 1945 and he has his 300th win sometime in 1961.

***********

As for Welch, I think it's very hard to compare 1880s pitchers with post 1893 pitchers. It was a different game. Fielders had a lot more to do with defense, making ERA+ much less reliable.

IMO, Welch just doesn't stand out amongst his contemporaries enough.

Joe, I just want to challenge the wisdom of zeroing out the bad years using Wynn as an example.

Wynn's 408 worst innings were at 75 ERA+. I don't know where ERA+ corresponds with replacement level, but his teams surely could have found a AAA pitcher to do that in those two and a quarter years. His performance is killing his teams. Granted the Senators weren't exactly going anywhere in those years, but Wynn hardly helped.

If those 400 innings (about an eleventh or so of his total innings) were it, then I can see overlooking them, but Early Wynn threw another 1300 or so innings of below-average to poor baseball. 1700 innings is about as long as Joe Wood or Dizzy Dean's entire careers, and Wynn had an ERA+ of 86 over those nine years.

Any way you cut it, that's awful, but choosing to ignore it short-shrifts someone like Eppa Rixey. Rixey who had only five bad years, but they were closer to the league average, and his peak ERA+s are not much lower than Wynn's. (Understandbly, ERA+ isn't the end all be all, but I'm using it here as a proxy). Ditto Red Faber (only two bad years) or Ted Lyons (four bad years).

It short shrifts them because it begins the comparison between them and Wynn (or Ruffing or Grimes) with the good stuff then forgets the rest. I'm not suggesting that Wynn is a better or worse candidate than these guys, only that in drawing comparisons among them without talking about every season, it will lead to an overselling of Wynn in comparison to players who offered better value outside their primes.

It would seem to depend on who you're comparing a guy like Wynn or Grimes to. When comparing them to short-career peak guys like Dean, it probably makes some sense to zero out the bad years, since otherwise you're almost giving Dean and the like extra credit for flaming out and never having the chance to have bad years. (Of course, the peaks--at least the short ones--are so much lower that it doesn't really make much difference to peaksters like me.)

But when comparing them to other guys with an equal number of years, you're right, it gives these guys an advantage--and one that makes little sense--over people like Rixey or Faber who didn't have as many down seasons.

I agree with Mark on this one. When compared to Dizzy Dean, for instance, Wynn sh ouldn't get points taken away because he was below average in an number of seasons. Soemthing tells me that if Dean had the opportunity to be below average for that many seasons at the end of his career he would take it. It isn't Wynn's fault that they kept running him out there when he wasn't any good.

On the other hand, those seasons need to be considered when comparing him to Rixey, Faber, et al., as Mark stated.

.
Number of pitchers with at least 20 Win Shares, NL and AL
year=, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,
1890s, , , ,16,15,14,14,16,18,20, (12 teams; 154 games 1898-1899)
1900s, 9,25,17,20,24,25,24,22,22,20, (1904, permanent increase to 154 games)
1910s,14
1920s,16,
1930s,11,
1940s,10,
1950s,15,10,12,11,10, 7,13, 6, 9,10,
1960s, 7, 8

Number of pitchers with at least 25 Win Shares, NL and AL
year=, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,
1890s, , , , 8, 9, 9,10, 7,12,11, (12 teams; 154 games 1898-1899)
1900s, 2, (8 teams 1900 only)
1910s, 8,
1920s, 9,
1930s, 5,
1940s, 4,
1950s, 6, 3, 3, 3, 2, 2, 4, 1, 1, 0,
1960s, 2,

Someone may be able to complete those tables semi-automatically, using a digital source for win shares by player-season. I counted player-team-seasons (no summartion for multi-team players) in the printed book.

I believe that much of the decline in pitcher usage and thus standards of excellence measured by win shares occurred during the 1950s. I believe that much occurred during the 1920s, too, as someone said.

1950 may be an outlier w/r 1940s as it is w/r 1950s.

1960 is an outlier. Wynn finished tie for 4th to 9th in AL pitcher win shares with 16.

What Win Shares database is available, perhaps on the web, perhaps commercially? Is there one that enables a query such as "year 1900, pitchers, order by WS"? If so, how much capability covering multiple seasons?

5. Wynn 154-42-35-35-26-19-15-10-8-2-(8 seasons below average!)/106 in 4564 IP
#47. Wynn 308/28-25-24/110/24.6

Not much of a peak, so the fact that I would have him #5 is high praise indeed. Surprisingly, more WS/IP than Diz!

6. Cicotte 185-75-74-55-32-29-16-15-1-(4 seasons below average)/123 in 3226 IP
#50. Cicotte 247/35-32-27/124/28.8

Now that I look at it, how is Cicotte not better than Wynn? Well, 1000+ IP by Early. But there was no time at which Wynn was actually as good of a pitcher as Eddie.

Maybe you should apply the dimino maneuvre, a la
Stan Coveleski plus a 13-(-5) season or
Bob Lemon plus a 21-9 season.

7. Griffith 191-34-31-30-28-20-18-16-14-9-(81)/121 in 3386 IP
#70. Griffith 273/34-32-30/143/29.2

His #70 rating is serious timeline abuse. Don't see how he could be above Cicotte however. The Old Fox was better in seasons #1 and #7 on, but #2 through #6 is all Eddie.

Griffith leads Cicotte on all four Win Shares measures despite playing shorter seasons. What's up?

<i>I agree with Mark on this one. When compared to Dizzy Dean, for instance, Wynn sh ouldn't get points taken away because he was below average in an number of seasons. <>

I totally agree. IMO, there's zero reason to penalize any player for below-average seasons.

What Win Shares database is available, perhaps on the web, perhaps commercially? Is there one that enables a query such as "year 1900, pitchers, order by WS"? If so, how much capability covering multiple seasons?

I believe The Hardball Times has made their Win Shares database available for download, but not sure if that is for "everyone" per se....

Clearly Wynn and Dean must be compared by a similar number of seasons and innings, but I do wonder about using Wynn's best years versus Diz's ten best years, because that comparison is all about which ten years or 1975 innings of His Earlynees you pick. If you pick Wynn's poop years plus one other 250 inning year, you've got the entirety of Dean's career. Perhaps ideally you would compare Wynn's ten best consecutive years. Or his best ten consecutive adding up to around 2000 innings?

I totally agree. IMO, there's zero reason to penalize any player for below-average seasons.

Sure there is, John. If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years, it wouldn't be imprudent to recognize that the extra two bad years not only don't improve the candidate's case but probably hurt his team's ability to win. If you call that "penalizing" him I'm OK with that. ;)

<u>All else being equal</u>, the pitcher with the two extra bad years should probably be ranked after the other guy in the theoretical instance I've described. It's just a question of ranking by attrition: the positive value is equal, so whoever did more damage goes second. If a pennant is a pennant, then so too is an also-ran an also-ran.

You could also consider a case like Wynn and Rixey where Wynn's got a little higher peak, but Rixey has several fewer bad seasons/innings AND his bad years are much closer to average than Wynn's. You can't wash that away in a comparison of them, it's part and parcel of who these pitchers are.

At a more realistic level, this stuff hardly ever comes into question, it's just somewhat borderline guys like Wynn, Ruffing, Grimes, Kaat where there's a LOT of bad seasons in their portfolios so that ignoring the bad stuff means dismissing a third of their careers that it becomes a big issue.

John Murphy wrote:

IMO, there's zero reason to penalize any player for below-average seasons.

The Good Doctor wrote:

Sure there is, John. If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years, it wouldn't be imprudent to recognize that the extra two bad years not only don't improve the candidate's case but probably hurt his team's ability to win. If you call that "penalizing" him I'm OK with that. ;)

All else being equal, the pitcher with the two extra bad years should probably be ranked after the other guy in the theoretical instance I've described. It's just a question of ranking by attrition: the positive value is equal, so whoever did more damage goes second. If a pennant is a pennant, then so too is an also-ran an also-ran.

I agree with John. If the team had a better option than the below-average player, the below-average player wouldn't be on the field, in almost every case (Neifi Perez's career notwithstanding). If you are the best option a team has, even if you are below averge, you are not hurting your team. If Player A and Player B are entirely equal over 13 seasons, matching each other in value year by year, win share for win share, except that Player B added two more below average seasons, the fair conclusions are (1) the two players are essentially equal or (2) Player B should rank slightly ahead.

In Wynn's specific case, it is wrong to regard him as a bad pitcher in 1957 and 1958, for instance. He was a somewhat below average in effectiveness, but he was among the league leaders in innings pitched (#2 and #6, respectively). If I remember rightly, 1957 is the year Herb Score was hit by the line drive early in the year. I would guess that Wynn's innings helped to pick up the slack. He was certainly a better option than Bob Lemon, who threw 117 innings at a 5.44 DERA to Wynn's 4.86, which was only slightly above the team average of 4.78.

How do you evaluate Greg Maddux's 2005 season? Nobody would say he was a great pitcher in 2005, like he used to be, and he was a bit better than Wynn in 1957-58 (101 ERA+). He doesn't go all that deep into games, but he led the team in IP and was 7th in the league. With all the injuries to the rest of the staff, where would the team have been without him? To contend, they needed Wood and Prior to be healthy, and they weren't, but Maddux giving them a chance to win every fifth day sure wasn't hurting the team.

I disagree with Doc in taht the pitcher with 5 bad years should be just about equal to the pitcher with three bad years (assuming their best 8 years are equal). I can't blame a guy for wanting to stick around and play baseball a few more years or for being called up too early.

However, where you shouldn't just zero out all below average play is when comparing players of equal length. Let's take the two pitchers with eight really good years again. Say instead of three bad years the first pitcher has five below average years (say average amount of IP but an ERA+ of 85) and the other guy has five bad years (AVG IP, ERA+ of 70). Zeroing them out would make the first look equal to the second when he was not.

I have no problem with taking out extra bad years when comparing long career guys to short career guys, but taking them out when comparing two long career guys is probalby going to lead to soem problems.

So I would say this...
A pitcher should not be penalized for extra belwo average pitching outside of his prime BUT it should count if that player is being compared to someone with as long (or longer) a career and better outside of prime pitching.

I suppose I'm ultimately arguing for Loss shares....

the fair conclusions are (1) the two players are essentially equal or (2) Player B should rank slightly ahead.

Yes, the players are essentially equal, but no in fairness player B can't rank ahead. Not if we're talking about merit. Two terrible years, even if a guy is the best available to his team, is not something to hang a HOM hat on. Which is to say that I don't agree that a pitcher posting 400 75ish ERA+ innings can realistically be seen as the best a team can do in a given moment. (It's worth noting that WARP disagrees with me on this matter---Wynn never had a negative WARP.) More likely it was the best the team chose to do. I'd hate to see bulk badness push any candidate toward HOMularity above another equal one.

but his teams surely could have found a AAA pitcher to do that in those two and a quarter years.

Um... They did. His name was Early Wynn, hot prospect that wasn't developing because it was a bad organization ;-)

Actually, the answer depends on where you place replacement level. 75 ERA+ would be a typical pitcher on a .350 team. WS gives positive credit to both batters and pitchers at this level.

If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years,

What do you mean "roughly equal"?

If you mean they have the same value in the good years, then the guy with the two extra bad years is more valuable because it's still above replacement, hence positive value, even if not positive by much. He's probably a better pitcher during those two years than the other guy who was so bad that we have no record of him because he was forced to retire when nobody would pay him to pitch.

If you mean they have the same value after you include the bad years, then the other guy is more valuable because he has more of his value in the good years and the guy with the extra bad years needed those extra years just to catch up.

Sure there is, John. If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years, it wouldn't be imprudent to recognize that the extra two bad years not only don't improve the candidate's case but probably hurt his team's ability to win. If you call that "penalizing" him I'm OK with that. ;)

In my system, a pitcher may get negligible points (IOW, basically zero), but I never deduct value from what he has done prior or after his sup-par seasons.

If Bob Feller had never retired and was still allowed a roster spot in 2006, none of his "contributions" should be seen as a negative to his career as a whole. Maybe the past fifty years would deserve a big fat zero or close to it, but it shouldn't TPI his value to the point that he's perceived as a below-average player historically (obviously, this is an extreme hypothetical ;-)

"It short shrifts them because it begins the comparison between them and Wynn (or Ruffing or Grimes) with the good stuff then forgets the rest."

Agreed - I was only comparing him to Lemon, Coveleski. As Mark said, I think it's perfectly reasonable when comparing someone to a pitcher with a much shorter career.

In the Grimes comparison, Wynn threw more innings in the good years (relative to his peers). In the bad years they both had 83 ERA+ over a similar number of innings (Wynn threw more innings in the bad years, but the rate was the same). So I don't see an issue there. It worked in that specific case. I agree that below average years shouldn't be zeroed out - heck, I'd actually say that the guy with the extra below average years was MORE valuable than the guy without them. I was just trying say, 'even if you zero them out . . . " for the peak guys.

With Rixey, etc. I could absolutely see where it would be an issue.

"Sure there is, John. If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years, it wouldn't be imprudent to recognize that the extra two bad years not only don't improve the candidate's case but probably hurt his team's ability to win."

I couldn't disagree more. I'll never give a player a 'negative' if his manager thought he was worth running out there. I ALWAYS assume if they could have found someone better, they would have.

>I'll never give a player a 'negative' if his manager thought he was worth running out there. I ALWAYS assume if they could have found someone better, they would have.

The problem with saying that of course Early Wynn had value based simply on the fact that they kept running him out there is this. This assumes an efficient market.

I don't remember who said what over 70 years of HoM discussions, but...

I would think that those who supported Buzz Arlett and who give MLE for MiL play (Cravath, Averill, etc.) would NOT want to argue that "the market" says that Wynn or other OPS+ 75 pitchers necessarily had MLE (much less HoM case additive) value. If the functioning of the talent market is that efficient then the judgements made about Arlett and Cravath by the market would be decisive.

True that the market is not perfectly efficient, but it's possible to recognize unusual cases where a player of above-average quality does not play in the majors while still acknowledging that in most cases teams will not use a player if they can find a better alternative.

To my mind, the more important point re Wynn and Grimes is that in seasons that some voters would zero out, they were still quite valuable to their teams. Wynn in 1957-8 was not just hanging around: you can find starting pitchers on his own team worse than he was, and I would guess that you could identify ten starting pitchers around the league with worse ERA+/DERAs, and that's not even opening up the subject of the value of his durability.

I agree that seasons with an ERA+ of 75 are bad, and that he had little value to his teams in these years.

I agree that seasons with an ERA+ of 75 are bad, and that he had little value to his teams in these years.

But as you would agree with, Chris, there's a difference between little value and negative value.

I guess what needs to be done here is jsut set the zeroing out level below average, something like 2 wins below average or whatever.

Yeah, I don't believe in negative value either. It's just hard to argue that Buzz Arlett should get MLE on account of the market being inefficient, and Early Wynn should get credit because the market is efficient.

Yeah, I don't believe in negative value either. It's just hard to argue that Buzz Arlett should get MLE on account of the market being inefficient, and Early Wynn should get credit because the market is efficient

I think this is why economics is a social science . . . :-)

Murphy:
IMO, there's zero reason to penalize any player for below-average seasons.

Chaleeko:
Sure there is, John. If two roughly equal pitchers each have 8 good seasons and one has 5 bad years and the other 3 bad years, it wouldn't be imprudent to recognize that the extra two bad years not only don't improve the candidate's case but probably hurt his team's ability to win. If you call that "penalizing" him I'm OK with that. ;)

Dimino:
I couldn't disagree more. I'll never give a player a 'negative' if his manager thought he was worth running out there. I ALWAYS assume if they could have found someone better, they would have.

Wendt:
That doesn't make it true and it is demonstrably false.

It's a stochastic world. Realized value may be less than expected value; indeed, that happens about half the time. JoeD and JohnM know that, but the D-M postulate implies the opposite about players with expected baseball value zero (replacement) and implies the opposite approximately about players near zero. Here "baseball value" refers to helping the team win, and I mean helping as a player.

There is non-baseball value. Maybe Dave Concepcion draws some fans to see him in action because of what he represents. Maybe the Cinci club judges that retaining him and playing him half time is "treating him well" in a vital sense.

Those two points have nothing to do with efficient markets, nothing to do with whether economics is social, natural, or logical. They show why the postulate is false even in economic equilibrium.

Besides, the world is not in economic equilibrium. Marc argued along this line and it is a familiar one here.

A false postulate isn't necessarily bad, I don't even say that whole hearted commitment to a false postulate is bad. Theories manage the world partly by saying false but useful things about it.

Cobb:
To my mind, the more important point re Wynn and Grimes is that in seasons that some voters would zero out, they were still quite valuable to their teams. Wynn in 1957-8 was not just hanging around: you can find starting pitchers on his own team worse than he was, and I would guess that you could identify ten starting pitchers around the league with worse ERA+/DERAs, and that's not even opening up the subject of the value of his durability.

Wendt:
Indeed, the evaluation of particular player-seasons is practically important. For instance, Wynn 1957 and Wynn 1958. (Dr.Chaleeko has observed that WARP disagrees with him.) For instance, Wynn 1942 and Wynn 1948 (ERA+ 75). For instance, some Burleigh Grimes seasons.

Chaleeko:
I don't agree that a pitcher posting 400 75ish ERA+ innings can realistically be seen as the best a team can do in a given moment. (It's worth noting that WARP disagrees with me on this matter---Wynn never had a negative WARP.)

Wendt:
But Davenport (WARP) agrees on the general point that a player may be hurtful, not helpful, to his team's winning. A player-season may have negative baseball value.

jimd:
Actually, the answer depends on where you place replacement level. 75 ERA+ would be a typical pitcher on a .350 team. WS gives positive credit to both batters and pitchers at this level.

Wendt:
And WARP, too, with a higher standard. Right?

But the D-M postulate that every player-season helps the team win (or it is precisely neutral), that no player-season has negative value in other words,

162 IP minimum, listing all 100 ERA+ seasons:

BoLemon 144 39 36 34 33 13 12 08 03 01

Ferrell 146 35 33 30 26 24 24 06

EpRixey 144 43 42 39 36 29 24 15 15 13 10 09 09

BGrimes 152 44 38 36 31 23 08 08 08 03

EarWynn 154 42 36 35 26 18 15 10 09 03

ERixey top 10 in IP: 1 3 3 3 4 7 8 8 9 9
Grimes top 10 in IP: 1 1 1 3 3 4 7 9 9 9
EaWynn top 10 in IP: 1 1 1 2 2 3 3 6 6 6 6 7

Rixey never gets beat after a 1-pt loss in '2nd-best season.'
Grimes and Wynn are even more similar.

I had Rixey 1 and Grimes 12 last year, with Welch at '16.'
Relatively speaking, Wynn was a little more of a workhorse than Grimes, so he probably noses out Grimes and Welch. Wynn misses ages 25 season, probably deserves a 'bonus 103' or so. Grimes pitches in weakened 1918, with his 131 ERA+. Little stuff starts to add up for Early.

Wynn beats Lemon by this measure, but that horse has already left the barn. And Lemon has the 82 to 54 OPS+ lead for those who like that tiebreaker.

Wendt:
But Davenport (WARP) agrees on the general point that a player may be hurtful, not helpful, to his team's winning. A player-season may have negative baseball value.

I should first say that I don't always believe that each team picks the best players they can find to be on their rosters.

Secondly, I also belive a player can negatively hurt his team.

But for this project, my first two paragraphs shouldn't apply, IMO. To allow negative value as part of the total value of a player's career, you wind up negating the positive that the player has accomplished.

I had posted this a few days ago on this thread:

If Bob Feller had never retired and was still allowed a roster spot in 2006, none of his "contributions" should be seen as a negative to his career as a whole. Maybe the past fifty years would deserve a big fat zero or close to it, but it shouldn't TPI his value to the point that he's perceived as a below-average player historically (obviously, this is an extreme hypothetical ;-)

This is using negative values to its most absurd lengths, but the underlying truth is still there. IOW, I refuse to penalize a player for the sins of the general manager.

Howie,

Re post 81: Grimes missed some portion of 1918 and/or 1919 due to the war, he's listed on the Hall's website as a veteran. So this is likely a washout with Wynn, particularly since Grimes was pitching better in those seasons than Wynn was in the mid 40s.

John,

This is why I think Loss Shares would be useful here. The negative value or contribution to losing would be helpful to know not as part of how much a guy helped his team win, but in another seperate column where his positives aren't washed out by his negatives.

Does Win Shares show how much someone's positives outshine their negatives? If they don't then the player will have a really small number of Win Shares and be nearly worthless. WARP would be the same way. Aren't teh things a player does to help his team lose the same things that will garner him fewer Win Shares? It isn't like these are seperate skills, making outs helps a team lose and prevents it from winning.

This is why I think Loss Shares would be useful here. The negative value or contribution to losing would be helpful to know not as part of how much a guy helped his team win, but in another seperate column where his positives aren't washed out by his negatives.

I could handle that, Eric.

Does Win Shares show how much someone's positives outshine their negatives?

The Runs Created formula (the new one anyways for sure) has a negative contribution for outs. Linear Weights most certainly does have a negative contribution for outs. Without this, run estimators wouldn't be able to account for players left on base. I don't think you can fully separate win shares from loss shares because of this.

The issue with Win Shares is that it actually has a baseline performance where and 'zeros out' players that fall below. See the Win Shares critique by Patriot on his "Walk Like a Sabermetrician" blog for further details.

Also the "zero-level" performance is quite low and below replacement. Some sort of Win Shares Above Replacement measure would be interesting for guys like Wynn.

I just want to add that I agree with John's post #82.

So I guess Paul is correct, I shouldn't (and don't) assume that if they could have found someone better, they would have.

I should have said, I don't penalize the player if the GM was willing to run him out there.

I thought that's what I was saying, but it's not what I said :-)

"Also the "zero-level" performance is quite low and below replacement. Some sort of Win Shares Above Replacement measure would be interesting for guys like Wynn."

That's an easy one!

I drop off about 6.5 (IIRC it's somewhere around 6-7) WS a full 162-game season for position players (I actually take, league plate appearances per game, divide by 9 and multiply by team games to get a 'full season') and IIRC, that works out to about 6.5 per 220 IP for pitchers. Subtract that each season and you'll get WSaR (Win Shares above Replacent).

that works out to about 6.5 per 220 IP for pitchers.

I believe that you meant "330 IP". A league average pitcher with 220 IP gets 13 Win Shares. 6.5 is half of his value. A league average full-time batter/fielder gets about 19 WS. 6.5 is one-third of his value. The adjustment as written kills pitchers.

The above assumes that WS replacement is the same for both offense and defense. Many agree that WS is imbalanced with a much lower replacement level on the offensive side of the equations. (Which IIRC is why tangotiger proposed a 39/61 offensive/defensive split, to balance the replacement levels.)