Hey guys, I'm in the process of constructing a modified PHOM of my own, where I'm filling 200 slots by 2010.

I have at least 5 spots I'm unsure about who to fill them with.

Here are 25 names I'm considering: Buddy Bell, Mordecai Brown, Willard Brown, Bob Caruthers, David Cone, Stan Covelski, Rollie Fingers, Tommy John, Goose Goslin, Frank Grant, Keith Hernandez, Tim Keefe, Tommy Leach, Bob Lemon, Bid McPhee, Jose Mendez, Dobie Moore, Dickey Pearce, Lip Pike, Reggie Smith, Joe Start, Ezra Sutton Luis Tiant, Quincy Trouppe, and George Wright.

Anyone's suggestions are greatly appreciated. Thanks and good job as always with the best insight on Baseball history on the net.

Reggie Smith = JD Drew (before his awful 2007 campaign)?

This is a bit unfair to Reggie. I made a different comparison in a previous thread, and I think it's worth repeating: Reggie Smith = Larry Walker. I doubt there are many pairs of hitters who've had more similar MLB careers.

Smith: 17 seasons, 1987 games, 8050 PA, 137 OPS+, .305 EQA
Walker: 17 seasons, 1988 games, 8030 PA, 140 OPS+, .307 EQA

Each player had three seasons in the Top 5 in OPS+ and two seasons in the Top 5 in Runs Created. Smith had five Top 10 OPS+ years and six Top 10 RC years. Walker had six Top 10 OPS+ years and five Top 10 RC years. Most metrics seem to indicate that Walker was a very good defensive RF, while Smith was roughly an average defensive outfielder but with several seasons in CF. There's very little to separate them in terms of value. If you think one of them is a HOMer, it's tough to justify leaving the other one out. If you think one of them definitely comes up short, then I'd imagine the other would also be on the wrong side of the borderline.

Bleed - Goslin, Start, McPhee, Sutton, Wright.

Not to mention marc, these guys are throwing just as many innings in their careers, they are just taking more seasons to do it - you can't take only the value they had in the decade of the 00s or 90s - take the value of their entire careers - and subtract out a reasonable replacement level (which you have to do if you are comparing guys that threw more innings to those who threw less).

heh, I guess those "HOM pitcher per year" charts may come in handy

from the voting list, I like borderliners all.
I guess I go Goslin MBrown KHernandez Keefe Mendez
next 5 Start Caruthers Fingers McPhee Pearce

interesting collection

Bleed, if all of these guys are on the outside looking in, who's in? Ginger Beaumont!? Anyway:

1. George Wright--greatest player in the world at his peak, of course so was Dickey Pearce but for Wright it was after there were a few more guys playing
2. Bob Caruthers
3. Dobie Moore--basically he was Ernie Banks' SS years
4. Goose Goslin
5. Keith Hernandez

Well of course I'm considering their entire careers. I'm just asking the question where they fit among their cohort. I guess Sabes was a cohort of 1?

>sunnyday, but you have to factor in that pitchers are responsible for a much greater portion of run prevention today than they were 100 years ago,

I said that.

>Pitching is surely as valuable today as it was then, maybe more, probably more.

And then I said.

>But the value is just distributed too many different ways

It seems like we don't even care how many innings a guy pitches. I have him ranked pretty high, around #25. But Saberhagen did NOT pitch as many innings as the old guys, nor was he the great pitcher of his generation.

Bleed, I'm not goign to give 5 but I will say that Tim Keefe belongs in any Hall.

After counting Japan, I agree that Walker and Smith are peas of a pod on career value, but Walker definitely had a higher peak. I have his top 5 single seasons as 7.1 (1997), 6.7 (1992), 6.3 (1994), 6.1 (2001), and 4.9 (1998), compared to a peak for Reggie of 6.8 (1977), 5.9 (1974), 5.3 (1978), 4.9 (1973), and 4.5 (1972). Sizable difference there.

Don Newcombe and David Cone

Question 1: Dan R, can you please tell me how the SD for pitcher performance for Newcombe compares in general to historical SD for pitcher performance (if you have this info). Do you know this for Cone too?

Question 2: Pending question 1, is Newcombe a better, comparable or worse candidate than Cone?

I'm struggling with Cone, thinking of adding Newk to the ballot based on Sunny's arguments and the SD findings.

Thanks for the input Joe Dimino, Howie Menckel, sunnyday2, and Rusty Priske.

sunny, you asked if I had Ginger Beaumont in my PHOM (Yest does, but I don't.)

My PHOM is scaled down to only 193 players through 2007. Here are the guys I had planned to "elect" with an approximate timeline of when each member would be elected:

Deacon White, Paul Hines, Dan Brouthers, Cap Anson, Buck Ewing, Roger Connor, George Gore, King Kelly, Billy Hamilton
John Clarkson, Jim O'Rourke, Ed Delahanty, Jack Glasscock, Monte Ward, Kid Nichols, Jesse Burkett, Amos Rusie, Charley Radbourn
Jimmy Collins, Bill Dahlen, Elmer Flick, George Davis, Cy Young, Fred Clarke, Cal McVey, Willie Keeler, Bobby Wallace, Ed Walsh
Christy Mathewson, Nap Lajoie, Honus Wagner, Sam Crawford, Eddie Plank, Home Run Johnson, Joe Jackson, Pete Hill, Frank Baker
Hughie Jennings, Joe Kelley, Jimmy Sheckard, Al Spalding, Sherry Magee, Louis Santop, Heinie Groh, Walter Johnson, Ty Cobb
Eddie Collins, Tris Speaker, Pete Alexander, Pop Lloyd, Smokey Joe Williams, Harry Heilmann, Cristobal Torriente, Zach Wheat
Bullet Rogan, Babe Ruth, Rogers Hornsby, Dazzy Vance, Dick Lundy, Mickey Cochrane, Oscar Charleston, Frankie Frisch
Lou Gehrig, Willie Foster, Al Simmons, Turkey Stearnes, Gabby Hartnett, Lefty Grove, Charlie Gehringer, Jud Wilson, Carl Hubbell
Joe Sewell, Paul Waner, Jimmie Foxx, Martin DiHigo, Josh Gibson, Mel Ott, Hank Greenberg, Joe Cronin, Arky Vaughan, Bill Dickey
Buck Leonard, Luke Appling, Willie Wells, Joe DiMaggio, Lou Boudreau, Ray Brown, Red Ruffing, Johnny Mize, Satchel Paige
Hal Newhouser, Charlie Keller, John McGraw, Bob Feller, Jackie Robinson, Monte Irvin, Roy Campanella, Max Carey, Pee Wee Reese
Enos Slaughter, Larry Doby, Ted Williams, Hardy Richardson, Wes Ferrell, Richie Ashburn, Ross Barnes, Stan Musial, Yogi Berra, Duke Snider, Minnie Minoso, Warren Spahn, Early Wynn, Robin Roberts, Billy Herman, Sandy Koufax, Eddie Mathews, Mickey Mantle
Don Drysdale, Whitey Ford, Goose Goslin, Rube Foster, Ernie Banks, Jim Bunning, Hoyt Wilhelm, Roberto Clemente
Roger Bresnahan, Willie Mays, Al Kaline, Ron Santo, Bob Gibson, Juan Marichal, Frank Robinson, Hank Aaron, Billy Williams
Dick Allen, Brooks Robinson, Harmon Killebrew, Stan Hack, Stan Covelski, Willie McCovey, Jimmy Wynn, Bill Freehan, Bob Lemon
Willie Stargell, Tim Keefe, Carl Yastrzemski, Johnny Bench, Fergie Jenkins, Gaylord Perry, Joe Morgan, Jim Palmer, Rod Carew
Bobby Grich, Pete Rose, Tom Seaver, Reggie Jackson, Steve Carlton, Bert Campaneris, Phil Niekro
Ted Simmons, Darrell Evans, Mike Schmidt, Don Sutton, Graig Nettles, Rick Reuschel, David Concepcion, Dwight Evans
Bert Blyleven, Dave Stieb, Gary Carter, George Brett, Robin Yount, Carlton Fisk, Goose Gossage, Nolan Ryan, Dave Winfield
Lou Whitaker, Alan Trammell, Andre Dawson, Ozzie Smith, Eddie Murray, Ryne Sandberg, Dennis Eckersley, Luis Tiant, Paul Molitor
Bret Saberhagen, Wade Boggs, Will Clark, Ted Lyons, Charlie Bennett, Cal Ripken, Jr., Mark McGwire, Tony Gwynn

My grasp for pre 1890's players and negro league players is limited.

Covelski, Goslin, Keefe, Lemon, and Tiant were the five I included in the questionable PHOM pile. I also had Dawson, Herman, Concepcion, Rube Foster, and Richardson in the bottom portion of my PHOM.

Anyone prefer a different 5 to 10 questionable PHOM?
Any suggestions on 5 to 10 players that are omissions from my PHOM?

Thanks for the feedback,
BTF

Sorry for the poor formating, the preview looked great though!!!

Pud Galvin? Phil Rizzuto? Dick Lundy?

Looks very sound to me, considering the smaller numbers. Like seeing Campaneris in there . . .

I definitely have that info. A few caveats on these numbers: Relief leverage is not credited if it exists, I am using a fixed rather than floating replacement level, and I make no adjustment for average career length. The latter is a huge argument in Newcombe's favor since guys from Cone's era lasted far longer than guys from Newcombe's. (I'm adjusting for seasonal IP usage here but not career length, which obviously benefits modern pitchers tremendously). Also, this takes no account of the strength of the NL relative to the AL in the 50's, which also is a very strong case for Newcombe. Without further ado (your mileage will obviously vary on credits):

Glossary

All numbers are straight-line adjusted to a 162-game season.

IPYr: Innings pitched, straight-line adjusted to a 162-game season.
DERA: BP's DERA (the pitcher's RA+, adjusted for his fielders, expressed as runs allowed per 9 innings in a league scoring 4.5 runs per game with a Pythagorean exponent of 2)
PWA1: Pitching wins above average, calculated by the following formula: IPYr * ((2.25/(20.25+DERA^2))-.0556).
BWA1: Batting wins above overall league average, calculated in the same way as they are for hitters.
LgAdj: Ratio of the 2005 standard deviation of DERA to the regression-projected standard deviation of DERA for the year in question.
PWA2: Pitching wins above average, adjusted for the standard deviation of DERA.
BWA2: Batting wins above average, adjusted for the hitting standard deviation.
Replc: Standard deviation-adjusted wins above average (combining hitting and pitching) that a replacement starting pitcher would have had in the same innings and plate appearances. Replacement starting pitching is fixed at 2.1 wins below average per 200 innings pitched; replacement hitting is the offense of a league-average starting pitcher.
IPTrn: Ratio of the innings pitched of IP leaders in 2005 to the ratio of IP leaders for the year in question. IP leaders are calculated as the average IPYr of pitchers ranked #X/2 to #3X/2 in IP in the major leagues, where X is the number of teams in the major leagues.
WARP: Wins above replacement, calculated as (PWAA2+BWAA2-Replc)*IPTrn.

aTTL is career totals excluding sub-replacement seasons.

Note: some numbers do not add up exactly because the league average DERA in any given season is not actually 4.50 as BP claims it is--it's always lower, sometimes by a little, sometimes by more than a little. I've corrected for this so that total pitching wins above average always sum to 0 for every league-season. Also, career totals will not exactly match real-life ones due to adjusting strike seasons for length.

Don Newcombe



3-year peak: 18.1
7-year prime: 38.1
Career: 58.3


David Cone



3-year peak: 19.1
7-year prime: 39.7
Career: 58.4

Well, it's pretty durn close. As you can see, the standard deviation of DERA was about 20% higher when Cone pitched than when Newcombe did. Career value is identical; Cone has a nudge more peak and prime. Nonetheless, the "missing" factors here--league strength and average career length--go *very* strongly towards Newcombe. I don't exactly have the tools to put a quantitative value on those issues yet, but they are worth far more than one win of peak and 1.6 wins of prime. I'd definitely have Newcombe ahead by a good margin.

Dan,

Thanks!

Can you post Walters too just for kicks?

I'm going to post everyone on the pitchers thread in a jiffy.

Death Notices - Mike Coolbaugh was left out.

It's 11/05 to 10/06 for the death notices.

through the end of october then? Buck O'Neil (d. 10/6/2006)

No offense at all to the Coolbaugh family, but does he qualify under our listings?
I thought it was "HOM candidates."

No offense at all to the Coolbaugh family, but does he qualify under our listings?
I thought it was "HOM candidates."

I don't know, Howie. Maybe we should run the MLEs on Coolbaugh's minor league numbers before we dismiss his candidacy too quickly.

Well, if Curt Gowdy and Dick Wagner are candidates, Mike Coolbaugh is a candidate.

Does anybody know how it happened BTW? Was his back to the plate, etc? Seeing a couple rockets just outside 3B the other night with a runner on 3B and the coach with his butt to the plate, you could start to envision....

2007 Preliminary Ballot
1)Cal Ripken (PHOM)
2)Tony Gwynn (PHOM)
3)Mark McGwire (PHOM)
4)Hugh Duffy (PHOM)
5)Ben Taylor (PHOM)
6)Tommy Bridges (PHOM)
7)Kirby Puckett (PHOM)
8)Lefty Gomez (PHOM)
9)Bus Clarkson (PHOM)
10)Dale Murphy (PHOM)
11)Vern Stephens (PHOM)
12)Dick Lundy
13)Gavvy Cravath (PHOM)
14)Bob Johnson (PHOM)
15)David Cone
16)Tony Oliva (PHOM)
17)Dick Redding (PHOM)
18)Bobby Veach (PHOM)
19)Bret Saberhagen
20)Dizzy Dean
21)Orlando Cepeda (PHOM)
22)Reggie Smith
23)Al Oliver
((23a)Andre Dawson))
24)Albert Belle
25)Jack Clark
26)Jim Rice
27)Wally Berger
28)Don Mattingly
29)Dan Quisenberry
30)Lee Smith

Maybe we should run the MLEs on Coolbaugh's minor league numbers before we dismiss his candidacy too quickly.

The Hall of Merit would benefit more if they dismissed you very quickly instead.

Which reminds me - shouldn't Corey Lidle be among the death notices? He died October 11, 2006

The Hall of Merit would benefit more if they dismissed you very quickly instead.

Glad to see that you wasted your 31st post on that. Kiss my succulent Jewish ass.

Prelim:

1. Ripken
2. Gwynn
3. McGwire
4. Perez
5. Walters
6. Ryan
7. Staub
8. Matlock
9. Murphy
10. Saberhagen
11. Trout
12. Johnson
13. Cravath
14. Cone
15. Puckett

Hmm, that's the most interest the necrology list has generated in a long time. I'll try to explain.

For these last couple lists I decided to move towards the sort of list we'll get with our upcoming annual election. The one above runs from 11/1/05 through 10/31/06. The 2008 Ballot Discussion thread will have the deaths from 11/1/06 through 10/31/07.

I include all players who were on the lists of candidates in the Discussion threads, plus a few others with long careers or with some peculiar interesting aspect (Eddie Gaedel was one of these). I considered putting Corey Lidle on the list, but decided his career falls short, that even if he'd lived that he was not going to be on any list of newly eligibles. He is also not eligible for the HOF, having played in only nine seasons.

To call the list "HOM Candidates" is, admittedly, a bit of a misnomer. Over time, I've gradually tossed in a few interesting non-players into each list. This goes along with the original intent, to maintain a little historical flavor to the project (the HoM simulated induction weekend games were also a nice touch along these lines). It helps remind us where we are in time.

I would not expect that I'll have Mike Coolbaugh on the next list. He scarcely had a career in MLB and made only a minor contribution beyond that.

Not that I see it affecting this election, but it looks to me like there has been a WARP update. FRAA numbers seem to be much less conservative than I remember.

sunnyday, zop, and Dan R,
This is getting way too boring. If you guys can't generate an argument about something soon, I'm going to have to do something desperate like pick a flame war with karl by insulting Bob Caruthers's mother. Don't make me do this.

This is getting way too boring.

I'll take boring.

Also, I think I remember Jake Beckley's mom wore army boots.

Hahahahah. I'm sure there will be some debate when I post my pitcher numbers, but I decided to hold off on them until I can get a working career length adjustment in place--just adjusting for seasonal IP norms as Joe Dimino does is way too harsh on the deadballers IMO.

Jim Sp, if you post a preliminary ballot maybe I can find something to quibble with. But we tend to see eye to eye on a lot of things.

Need a quibble? Got a quibble.

George Van Haltren - 70 'years' ago, we thought he WAS something. He ain't much (17th in voting) now. Did we learn.... or did we forget?

GVH's case has been hurt, IMHO, by BP's WARPed view.
1 According to WARP, GVH was not mch of a hitter; but BP's EqA metric does not adjust well for the high-OBP, speed-based offense of the 1890s. It underrates GVH, whose calling cards are OBP and speed.
2 BPO also disses GVH's fielding, calling him a significantly below-average OFer. I honestly don't know how good he was, but for a guy who played lots of CF, ran real well, and had a arm good enough to pitche, you'd think he was OK out there.

GVH has never doen well by peak voters; I understand that. But if you're a prime guy, here are the MLB Win Share leaders for 1888-1901

Nichols, Kid
Young, Cy
Hamilton, Billy
Van Haltren, George
Delahanty, Ed
Burkett, Jesse
Rusie, Amos
Duffy, Hugh

I'd say he was HOM quality for those years, wouldn't you?

And if you are a career guy; well, George pitched almost 700 innings. And still is ranked 33rd all-time in runs scored, which last time I checked, was a fairly important part of the game of baseball.

I'm not GVH's best friend. But he used to have a lot more friends. I think some of them became fair-weather.

Yay! A discussion!

I dunno about EqA, but I don't think OPS+ underrates Van Haltren's 1893-99 years. Van Haltren's game certainly involved OBP and speed, but he's not NEARLY an extreme case like McGraw, who is just so far removed from normal that OPS+ can't handle him at all. Van Haltren's .101 career ISO is actually higher than the league average of .096 when he played, while McGraw's is significantly below average. And Van Haltren only had one season (ignoring the weak 1891 AA) when he finished in his league's top *eight* in SB--he was not posting the extremely high league-relative SB numbers that tend to correlate with an above-average success rate.

When I convert my BWAA+BRWAA to a projected OPS+, they match Van Haltren's actual OPS+'s almost exactly. The McGraw argument doesn't apply to him--OPS+ is a very accurate measure of his offensive value.

No idea about his fielding, although I could find out what DRA thinks of him if you'd like.

A few other important points about GVH.

1) Most of his career was spent in the contracted NL. This shaved down his peak, making it harder to post stratospheric OPS+, win shares, etc.
2) Career win shares: 381. Eleven years 25 or more. That's after cutting his pitching WS in half and adjusting to 162 G schedule. Does WS overrated CF play that much, to make him unHoMable?
3) I think voters simply became disinterested in him. He lacks that Wow Factor. These 27 players somehow became better than him since 1972:

George Sisler
Joe Gordon
Cannonball Dick Redding
Minnie Minoso
Dobie Moore
Jake Beckley
José Méndez
Hugh Duffy
Ralph Kiner
Cupid Childs
Joe Sewell
Pete Browning
Billy Pierce
Bucky Walters
Rube Waddell
Tommy Leach
Nellie Fox
Charley Jones
Quincy Trouppe
Edd Roush
Gavy Cravath
Alejandro Oms
Roger Bresnahan
Bob Johnson
Charlie Keller
John McGraw
Phil Rizzuto

Some of these were ranked behind GVH for decades. How could we have had GVH so wrong for so long? Doesn't that imply that a good many HoMers deserve to have their plaques ripped down as well?

Did Van Haltren more score runs than an average baserunner of his era knowing the kind of hitters he had batting after him?

On VH's fielding: my recollection is that his range and assist totals are good, but his error totals are exceptionally high. I don't know the cause -- was his arm strong but scattershot, or did he have "bad hands"? Or am I totally misremembering the data?

Count me as a voter who cooled on Van Haltren. It seems to me that Jimmy Ryan is better, among 1890s outfielders.

In that 1972 election, GVH ranked 7th, Duffy 15th. All of the other top 20 have now been elected. I wonder if the presence of two 1890s center fielders in the consideration set hurt both of their causes, similar to how having 2 candidates for league MVP from the same team sometimes hurts both players in that voting?

I wonder if the presence of two 1890s center fielders in the consideration set hurt both of their causes, similar to how having 2 candidates for league MVP from the same team sometimes hurts both players in that voting?

GVH, Duffy and Ryan were collectively known in the 1920s as "the CF glut". The electorate's inability to decide on how to rank those three indeed split each others votes. The late 20s and early 40s were those guys' best shots at getting elected before the backlog of OF-ers got to be too huge. More "unique" candidates like Thompson, Sheckard and Pike ended up being inducted instead.

yes, GVH made many errors for an OFer

And Van Haltren only had one season (ignoring the weak 1891 AA) when he finished in his league's top *eight* in SB--he was not posting the extremely high league-relative SB numbers that tend to correlate with an above-average success rate.

He finished 1st, 9th, 9th, 10th in SBs IN THE COMBINED MAJOR LEAGUE (excl 1891), not NL or AL separately; which is about 1st 4th 5th and 5th in other eras.

batting runs / wins calculators NEED to adjust pre-1910 for how much more speed was part of the game (correlated to team runs scored). I cannot recall how your BWAA works, DanR.

Going back to something I did in the infancy of the project, I looked at runs scored in terms of league-leading runs scored - and I used runs scored as a key element in my arguments in favor of Harry Stovey. All three of Ryan, Duffy, and Van Haltren look good on this measure (and Pete Browning doesn't, but that's another issue). To be honest, Ryan probably looks the best on this measure.

I hear Chris - the separation of Ryan's fate from GVH and Duffy probably shouldn't have happened, but it did.

1) Most of his career was spent in the contracted NL. This shaved down his peak, making it harder to post stratospheric OPS+, win shares, etc.

This is inaccurate. Standard deviations (which are the measure of how often and how high one sees "stratospheric OPS+, win shares, etc.") were extremely *high*, not low, in the 1890s (although defense formed a much bigger part of that variance than it does today). There is certainly an argument to be made that the single-league 1890s represented a higher-*quality* league, and therefore that a pennant in 1895 was "worth" more than a pennant in 1901. But it took *more* wins (as measured by WS, WARP, etc.) to win a pennant in the 1890s than in almost any other era of baseball history.

GVH, Duffy and Ryan were collectively known in the 1920s as "the CF glut". The electorate's inability to decide on how to rank those three indeed split each others votes.

This seems to me to be exactly what is going on with shortstops today.


TomH, well, there were 12 teams not 8, so the boost should be 50% not 100%, but you're right. Nonetheless, my research on the relationship between SB success and attempt rates shows that when you don't have CS data, it is only the people who are very, very far out on the SB attempt rate curve (Eddie Collins in 1910 leaps to mind) that are almost guaranteed to have excellent success rates. Van Haltren's SB attempt rate was not nearly far enough from the league average to credit him as a far above average runner. (For all the fuss I make about McGraw's baserunning, I actually don't give him *that* much credit for it either--his SB totals were so high in part because he was on base so god damn much, and I derive estimated CS from the SB per time on base ratio).

Speed isn't part of my BWAA but rather my BRWAA, not that it matters. I don't give SB a different "weight" for the 1890s than I do for other eras because I don't use a linear run estimator for those years. SB/estimated CS and estimated EqBR are counted in BaseRuns just like they are in any other era. However, the fact that the 1890s were a high-SB era means that the standard deviation of BRWAA (even after adjusting for the overall standard deviation of performance in the league) is much higher in the 1890s than in, say, the 1950's, so it impacts WARP totals much more. The best runners, like Billy Hamilton, get about half a win a year credited to them for their work on the basepaths, which is (according to my regression equations) about all you can say with confidence without CS data.

Thanks for re-sparking the discussion, Tom.

I agree with Chris that Ryan is preferable to GVH. I rank them

Duffy
Ryan
GVH

with Duffy on my ballot, Ryan having a strong argument for making my ballot, and GVH being in the high backlog.

GVH being the 3rd highest in the 1890s backlog is not a bad showing. It was just in 2006 that McGraw passed him to put someone else behind Duffy.

Van Haltren problem might be he's grouped with the bats and there has been a groundswell to get the gloves voted in higher numbers here at the end of the project. Would he be the worst electee? Probably not, but there is little exciting there to make even his backers stand on the soapbox and yell for election.

Solid work, Dan. Your handling of lack-of-CS data is I think gorgeous. But here is something I'm not sure is captured:

Bill James's work in the original Historical Abstract (not mentioned much in the update) showed that pre-1910, to correlate team runs scored to team stats, you have to over-emphasize SBs much more than post-1920. Because there were more forced errors and extra bases, it was obvious that the fast runners generated more offense than in seen in traditional stats. I'm sure most fans at the itme knew this, which was why Cobb was thought by many to be better than Ruth (and while I disagree, it's closer than many might think). Today if we didn't have CS stats, we would use a basic RC equation and come real close to team runs totals. In fact if yo utry to use SB totals and NOT CS totals, you'll find that you have to weight SB's really low to make it work. Not so in 1900! Teams that ran scored a lot of runs, more than hits+TB+walks etc.

Win Shares captures this. WARP and LW does not; nor does the "OWP" that is on the BP cards.

I forgot to mention that SB attempt rate also obviously forms a major part of my regression equation to estimate non-SB baserunning runs, so high SB totals help you doubly there.

As for the greater relative importance of SB in the 1890s, I did my own study on it and did not find any support for James's claim. I looked at every team from 1893 to 1900, did a regression of their SB per time-on-base rate to the error of BaseRuns' prediction for their runs scored. The correlation was under .1. If I were either meaningfully underweighting or overweighting SB in that era, there would clearly be some relationship between team SB rate and the error of the formula.

I can send you the data on this if you'd like; it's very straightforward.

wow, that IS different from my understanding of James' conclusions. Sure, send me a file.

This is inaccurate. Standard deviations (which are the measure of how often and how high one sees "stratospheric OPS+, win shares, etc.") were extremely *high*, not low, in the 1890s (although defense formed a much bigger part of that variance than it does today). There is certainly an argument to be made that the single-league 1890s represented a higher-*quality* league, and therefore that a pennant in 1895 was "worth" more than a pennant in 1901. But it took *more* wins (as measured by WS, WARP, etc.) to win a pennant in the 1890s than in almost any other era of baseball history.

Why?

It seems counter-intuitive; if contraction is forcing most of the worst players out of the league, shouldn't standard deviations be lower?

If you've already answered this somewhere, please point me to it.

It seems counter-intuitive; if contraction is forcing most of the worst players out of the league, shouldn't standard deviations be lower?

I don't know what the structural reasons for this are; perhaps Dan does. But just empirically, the 1890's NL was a league were there were multiple >.650 teams and multiple <.300 teams in every season. That kind of gigantic spread in team quality is consistent with, and in fact can probably only exist with a very high standard deviation league.

How much is "syndicate baseball" messing with the stdev's? (How much of the 1890s felt the effects of syndicate ball?)

Well, I'm not comparing the 1890s to the 1880s--my research on standard deviations starts in 1893. I'm sure stdevs were higher still in the 1880s. Compared to the following periods in baseball history, though, the following elements are what account for the high stdev of that era:

1. Low player population per team.
2. Very awful teams--the 1890s were really a 9-team league with three AAA franchises (Louisville, St. Louis, and Washington) making everyone else's stats look good.
3. High run scoring.
4. A very low home run rate (with virtually no runs being scored via the HR, position players have more opportunity to distinguish themselves from each other via their fielding).

Note that the 1899-1900 contraction and 1900-1901 expansion most definitely did show the exact pattern you'd expect: the NL stdev was 3.20 wins per player per 162 games in 1899, 2.69 in 1900, and 3.42 in 1901. It may very well be that the same was true from 1891-2. But my research starts in 1893, and the 1890s were certainly high stdev compared to everything that came later. You didn't have any 200 OPS+ seasons, but you *did* have of 150 OPS+, 30 FRAA, 70-SB seasons. Moreover, because baseball was still developing and skill sets had not yet fully specialized, you still had something of what I call the "fat kid plays right" syndrome--some of the best overall athletes were both the best fielders and the best hitters, and so you get *tremendous* within-position variance at SS, where you have Jennings/Dahlen/GDavis/BWallace on the one hand and guys like Germany Smith and no-hit-no-field Monte Cross on the other.

DavidFoss, certainly the Spiders' .129 winning percentage in 1899 leads to high standard deviations (I'd loove to see McGraw's splits that year; he must have had a .750 OBP vs. the Spiders and .500 vs. everybody else). I don't know how prevalent that was from 1893-1898, but I measure that effect by using the winning percentage of the worst team in the league, which definitely has a statistically significant relationship to the standard deviation of individual player performance.

Dan, your research focuses on 1893-on. Have you gotten the chance to examine the effects on the demise of the AA on stdevs?

No, I haven't, but I certainly imagine it had a gigantic effect. I'm sure stdevs before 1892 were far higher than they were afterwards. If I didn't include *both* a league quality adjustment *and* a wild-guess stdev adjustment for Browning and Charley Jones, they'd have been 1-2 on my ballot every year.

Again, I don't work with pre-1893 seasons because a) run estimation errors are too high and b) all the "fat kids playing right" mean that you can't use the worst-regulars average to calculate replacement level.

OK, that makes sense. Essentially, it was super-contraction, with all the good players being compressed into nine teams. So what does that mean, win shares-wise for the second-tier stars (GVH, Beckley, Ryan, et al)?

On the one hand, they have more WS, because they'll be playing on teams getting more wins from pounding on the patsies. OTOH, they'll be getting fewer WS, since they'll be competing with all-strong teammates for the team's allotment. I would guess the depressing factor would be the stronger one here. Any idea?

Well, now that I think about it, the AA had been bleeding talent (and teams) to the NL for a few years before it finally died. A look at the 1892 NL standings makes it look like the four teams brought over from the NL had the immediate effect of an "expansion" (though guys from teams that did contract like Brouthers & Duffy no doubt helped the bolster the NL). So, although stdevs certainly dropped since the mid-1880s, they might have gone up in the NL between 1891-1892?

Just thinking out loud from observations (no stats in front of me), I could be wrong.

No idea. You could easily calculate the stdev of BP WARP for those years and find out! (I wouldn't recommend using OPS+ or WS for this due to their exclusion/underweighting of defense, which is a big part of the high 1890s stdev).

So I finally was able to create a list for the pitchers after crunching it through DanR's WARP2 data. There are some big gainers/losers on the pitching lists compared to BP WARP.

For all data I'm simply using PWAA2 + WARP2 for DanR and PRAA+PRAR for WARP. Negative seasons are usually deleted.

Pitchers who DanR's data likes better than BP data (in descending order)

Kevin Appier
Steve Rogers
Dennis Martinez
Frank Viola
Wilbur Cooper
David Cone
Ron Guidry
Herb Pennock
Tommy John
Dwight Gooden
Kevin Brown (top 20 of all-time)
Jim Bunning
Jimmy Key
Eppa Rixey

Bert Blyleven also jumps into the top 10 all-time.

Pitchers DanR doesn't like
Luis Tiant
Dutch Leonard
Kid Nichols
Wilbur Wood
Jerry Koosman
Eddie Cicotte
Joe McGinnity
Red Faber
Bob Lemon
Eddie Plank

Pitchers who don't move much
Burleigh Grimes
Catfish Hunter
Wes Ferrell
Frank Tanana
Nolan Ryan
Gaylord Perry
Don Drysdale
Don Sutton
Lefty Grove
Warren Spahn

Top 15 eligibles according to Dan R's data (in order)
Tommy Bridges
Bret Saberhagen
Rick Reuschel
David Cone
Kevin Appier
Urban Shocker
Dick Redding (est)
Tommy John
Bucky Walters
Dizzy Dean
Virgil Trucks
Ron Guidry
Wilbur Cooper
Luis Tiant (!)
Frank Viola

Kevin Brown would top that list. Relievers are excluded.

Sorry, Bridges just didn't pitch enough. Even at his prime--3 years > 250IP and 3 others > 200. And 2826 total, about 228 IP per year for his top 10 years, 265 for his top 5 @ 140-137-120-119-115 or 10 ERA+ points LESS than Walters. And his 3 20 win seasons were ERA+ 137-120-119, in about 850 IP, his team appears to have helped him out a bit. That's only a decent workload compared to the ? 1980s. I'm sure it was hard to dominate.... The rest of Bridges career is a bunch of 140-200 IP seasons, but god bless him, he held his ERA+ up.

Walters threw 300 more IP for his career, but a ton more IP in his prime, 265 per year for his top 10 years. > 300 3 times, > 250 another 4 times, and > 240 another 2 times. Top 5 = 1437.2 with ERA+ 168-153-145*-127-107. Only one of those (*) is WWII (1944). Throw that out altogether, you can still replace it with 253 IP at 126. His 3 20 win seasons were at ERA+ 168-153-145* with over 900 IP. The rest of Walters career was more of a mixed bag, true. But as late as 1946, he had a very Bridges-like season of 151 IP @ ERA+ 131, still chuckin' it pretty good at age 37.

And don't forget to discount Bridges' 1943 (191 IP at 147), too.

I guess for a peak voter, Walters was just a better, more valuable pitcher at his best than Bridges was. 20-30 more IP per year AND a better ERA+ in their primes. Of course, if the AL was a lot tougher than the NL....

I do think Dan is right about Cicotte and Tiant. Over-rated. And Wilbur Cooper is under-rated.

How do you guys separate Rixey and Faber? Well, league, I guess?

Whoa, Bridges? What kind of war credit are you giving him?? I do NOT support Bridges and I don't think my data does either--although of course one can interpret the numbers any way one wants...In terms of career WARP2, Bridges has 52.1, and I figure the war discount for '43 and the credit for '44 and '45 roughly cancel out, maybe give him another win or two, so that's say 54.0 with a *very* low peak (top seasons of 6.0 and 5.1). That's way below a lot of other candidates...you sure you don't have a typo somewhere? Just off the top of my head, Reuschel and Saberhagen have more career WARP2 and MUCH higher peaks.

Tiant is doing so badly because I don't have his reliever seasons (including the very valuable 1972) in the spreadsheet, DL from MN. I warned about that. I have posted Tiant's numbers for those years treating him as a starter on one of these threads.

The two cautionary notes I'd stress about these numbers are 1) they are based on BP's defensive adjustments, which crushes someone like McGinnity perhaps unfairly and 2) they make no adjustment for career length but do adjust for seasonal IP norms. This obviously favors modern pitchers tremendously, since their careers last longer than those of prior eras.

I definitely think Reuschel and Saberhagen are the two top backlog pitchers. Cone, Appier, and Tiant I would definitely ding a bit for the long career lengths of their cohorts. Bridges really isn't close on my radar screen and I'm not sure why he popped to the top of your list. Shocker has under 50 WARP2, hard to see how he's a serious candidate.

Bridges just kept putting up above average PWAA. I don't care if his seasons are short, he's clearly above what you get in value from an average pitcher.

Year PWAA
1930 0.5
1931 -1
1932 1.9
1933 2.4
1934 2.3
1935 1
1936 3.3
1937 0.7
1938 1
1939 3.2
1940 2.9
1941 2.3
1942 3
1943 3.2

"I figure the war discount for '43 and the credit for '44 and '45 roughly cancel out"

I strongly disagree. That's a 66% discount. You really think 3 war years are equivalent to 1 pre-war year?

I'm giving him 56.5 WARP2 and 29.7 PWAA (his career plus another 1942). That's a little ahead of Coveleski.

Since Bridges was already 37 in 1944 and 38 in 1945, normal aging would make his war seasons not terribly productive. Ted Lyons is a similar case for me.

While I do give rate a premium for hitters, I don't do the same for pitchers, since I think a P who doesn't throw a lot of innings really does hurt the rest of his team by tiring out the bullpen, etc. Using PWAA will bias you towards high-ERA+, low-IP hurlers, and away from their opposites who may accumulate equal value.

If you like that type of pitcher, then why Bridges over Saberhagen? Saberhagen has exactly the same PWAA2 and WARP2 as Bridges, but with a *much* higher peak.

While I do give rate a premium for hitters, I don't do the same for pitchers, since I think a P who doesn't throw a lot of innings really does hurt the rest of his team by tiring out the bullpen, etc. Using PWAA will bias you towards high-ERA+, low-IP hurlers, and away from their opposites who may accumulate equal value.

It's interesting that this subject has come up. I am doing a comprehensive pitcher review also, and the conclusions that I am reaching, focusing on performance over a given set of innings, are similar to DL's. The situation with pitchers may be analogous, I think, to the situation we started facing up to with hitters a few years back: the players with great peaks have gone in, and the players with the better peaks remaining don't have outstanding enough peaks to justify picking them against players whose peaks are lower but whose overall value above average is greater.

It's not the case, for example, when looking at Walters vs. Bridges, that they are pitchers who accumulated equal value, with Walters having the better peak. Bridges accumulated quite a bit more value above average. Using BP's numbers I get Bridges with 222 RAA (that's PRAA + BRAP + FRAA) in 2607 xIP during his prime, while Walters has 199 RAA in 2553.4 xIP in his prime. Does Walters' higher peak (with whatever wartime deductions are appropriate) make up for 23 runs above average? (Replacement level shouldn't be an issue, since this is looking at very similar playing times.)

Now in addition to those primes of very similar length, Bridges has 40 above average xIP, 224 below average xIP, and whatever war credit one wants to give for 1944-45. Walters has 426.2 above aveage xIP and 162.2 below average xiP. That's probably a career edge for Walters.

I'm not saying that Bridges is definitely to be preferred here, but I don't see it as an obvious call.

When you go look at Urban Shocker, who put up 264 RAA in 2503 IP, I think the call becomes even less obvious. It may be that adding SD to the picture dings Shocker considerably (I am working from WARP1), but since DL's list uses that adjustment and rates him highly, I'm not yet convinced about that.

That said, I won't attempt to make a serious argument for or against any of these pitchers until I have my full data set ready to show and analyze, and it's not quite there yet.

I will say that my findings are not incompatible with the view that Reuschel and Saberhagen are the two top backlog pitchers, though there are others who have an argument to be up there.

I hope that we can take a good long look at full pitcher numbers from Joe and from Dan, with some other thoroughly developed perspectives like sunnyday's study of peak WS above median and the prime above average study that I am working on. By taking that look, we might be able to build some new consensus on pitching candidates.

Standard deviations would help Shocker--the 1920s was a very low-stdev era for ERA+. (I have Adolfo Luque's as the seventh best pitching seasons since 1893 for that reason--he had a 201 ERA+ in a year where the #2 guy was at 139, *and* he was second in the league in innings. Pretty impressive.)

Chris Cobb, do you really mean to use xIP (which is just weighted by decisions for pitchers)? Certainly their translated IP would be a better bet...I can email you the spreadsheet with my preliminary pitcher numbers if you'd like. Just remember the caveats--adjusting for seasonal IP but not career benefits modern pitchers; reliever seasons I have to do manually; and I'm dependent on BP's fielding adjustments which you may not like.

Yes, I mean to use xIP.

The study that I am doing is _not_ attempting to compare pitchers from different eras directly, which is what translated IP do.

It is also not attempting to make minute comparisons of value, so the generally small differences between xIP and IP for starting pitchers are not a significant issue for the study that I am doing.

I am using xIP mostly because I am looking at PRAA and by implication DERA for a number of seasons, and since BP uses xIP in their PRAA and DERA calculations, my calculations are a heck of a lot easier if I use xIP.

Thank you for the offer to send your preliminary numbers, but I will wait patiently for the finished version! I don't think that I could use them in the study that I am doing (which is _nearly_ finished: I hope to post over the weekend), so I'll just continue forward with that using runs above average derived from WARP1.

My study is not an attempt to generate a comprehensive measure, like your work does and Joe's does. It's an effort to examine what the HoM's standards for pitchers have been so far and to identify the unelected pitchers who appear to meet those standards, so that their cases can be more closely examined using more sophisticated measures. It focuses on runs above average over 2500 IP as a way of working around the problems of adjusting innings pitched totals by era.

"Since Bridges was already 37 in 1944 and 38 in 1945, normal aging would make his war seasons not terribly productive."

I don't think that conclusion is drawn on data. He had yet to experience a dropoff at age 36, was added to the World Series roster immediately upon his return from WWII and pitched in the PCL for _four_ years after that. He was dominant early in the PCL including a no-hitter. Eventually alcohol and age deterioriated his skills (he was not a drinker when young but seemed to have picked up the habit in the war) as his career and his personal life fell apart.

He was a little guy and didn't lead the league in innings on a seasonal basis (outside of 34-36) but I think the lower number of innings could have helped to keep him productive past his prime. I don't think the Tigers of that era were particularly hurt by a lack of pitchers. They had Trucks, Rowe, Auker, Newsom, Trout, Newhouser and at times the bullpen was a Tiger strength with Hogsett and Benton.

You say he didn't have a great peak, do you factor in his post-season performance into that analysis? His 1934-36 is consistently about 295 IP/season if you add in the post-season work. That's at the top of the league 3 straight years. He also led the league in Ks over those 3 seasons. He put up 9 more good WS innings in 1940 which bumps that season over 200IP.

Also, you mentioned that the relief innings are missing - Bridges was regularly brought in to relieve - over 60 games over the course of his career.

1. No, I only count the regular season.

2. The relief *innings* aren't missing per se. I used a rather brute approach to separate out starters and relievers--GS/G > 50%. So all of Bridges' work is in there. What's missing are years like Tiant's 1972 (43 appearances, 19 starts, 179 innings). I can do those by hand easily where I miss them, although I'll be treating them as if they were starters' innings. Figuring out how to do relievers will be a huge challenge, because you can't know how much to credit them for leverage until you know how bullpen usage has changed over time (see my posts on "chaining.")

It seems to me that IF you give Bridges a ton of war credit for '44 and '45 AND don't dock him too much for '43 AND credit him for postseason work AND think that a 130 ERA+ in 150 innings is worth more than a 120 ERA+ in 200 innings, you STILL couldn't get Bridges higher than Saberhagen unless you don't care about peak at *all*. I guess I just don't think wins above average are a very useful measure for pitchers--today, a league-average innings muncher costs you $11 million a year to put up 0 PWAA!

But I'm very happy to see that you disagree, actually--one of the most rewarding things about doing this research is seeing how many intelligent voters use my data to draw conclusions I find practically appalling. :)

>That's at the top of the league 3 straight years.
>didn't lead the league in innings on a seasonal basis (outside of 34-36)

These comments sound like he led the league, which he never did.

I looked at WS "above the median ace." That is, I took the #1 WS starting pitcher from each team and identified the median, the #4/5 in Bridges day. This is based on the belief that a "great" or a HoMable pitcher would be at least all-star caliber at some time, and so compares him to what might be considered a minimally all-star level performance. This is further based on the belief that the accumulation of career totals is not as important as the seasonal totals you get along the way, since pennants are awarded on a seasonal basis, not apportioned based on career value.

Career WS over the median "ace"

Pitchers mentioned above

Walters +39
K. Brown +28
Saberhagen +26
Reuschel -4
Bridges -9
haven't done Shocker

Some other guys

Dean +39
Wi. Cooper +33
Heshiser +29
Grimes +25
Cone +20
Newcombe +20 in 7 ML seasons, no MiL (integration) or mil(itary) credit
Tiant +15.5

Willis, Joss, Jack Morris, Gooden are <+10, while Cicotte and Gomez < 0. Gomez and Bridges (and Joss) strike me as similar, effective but with low workloads.

Looking ahead: Maddux beats Clemens +103 to +101, Glavine beats K. Brown +42 to +28. I think Pedro is going to end up in the +60s or so. Randy Johnson is closer to K. Brown than Glavine, and Smoltz trails all of these '90s-'00s candidates as a starting pitcher--i.e. that does not include his closer days.

Note that Hershiser beats Saberhagen.

But I don't get Bridges. Why not Gomez? Or Hershiser?

On Bridges vs. Gomez:

Look, Gomez does have the big seasons that get him points in an "above the median ace" ranking, but outside those about 4 seasons, he was pretty much a flat ordinary pitcher. And he was a _really_ bad hitter, too.

Bridges was really good year after year. If look at Bridges over appx. 2500 IP, he is 222 runs above average.

If you look at Gomez over appx. 2500 IP, he is 115 runs above average.

A little bit of peak advantage for Gomez just can't outweigh that much difference in quality.

I say it's not obvious between Walters and Bridges.

It is obvious between Bridges and Gomez. Bridges was unquestionably the more valuable player, unless you look only count value above a very seasonal high baseline-- "median ace" is a baseline that is well above an average starting pitcher, even--and disregard the rest of the pitching record. Value above median ace is a useful measure for some comparisons, but not for comparing a Bridges to a Gomez, because it misses the fact that Bridges was way better than Gomez in all the years except Gomez's few big seasons.

Of course, yes, Saberhagen over Bridges, I should add.

Hershiser, maybe. It's not obvious, but Hershiser looks to me to have a value profile that is a lot like Bucky Walters' -- strong but short peak, not as effective a _pitcher_ as contemporaries Saberhagen and Cone, but making up a lot (all?) of the difference by being a good hitter for a pitcher and an above-average fielder, also.

I'd say Hershiser merits more discussion than he has gotten so far.

Walters does so well in this approach, sunnyday, because there was basically no other P active in the NL during his prime who could even be called HoVG. All the good ones were in the AL--Lyons, Feller, Ruffing, Gomez, Bridges, early Newhouser and Trucks...I think if you did the major leagues rather than just NL, you might get different results. And also, does WS factor out the Reds' superlative fielders?

I understand what you're trying to do here, but I think the methodology is sooooo subject to star gluts and droughts as to be quite unreliable--particularly in the eight-team-league era. Have you considered doing WS above the median SP, rather than above the median ace? That would give you enough of a sample size that--especially using the median--I think your results could be quite reliable.

Also, I think the way you do it is biased towards high IP guys rather than high ERA+ guys (the opposite of DL from MN's approach). The "correct" way (at least as far as I can see) to measure against the median is to calculate the WS per IP of whatever your baseline is (replacement, average, median, median ace), multiply that by the IP of the pitcher in question, and subtract, as opposed to simply subtracting the actual WS total of the player you select as your baseline from the WS total of the player you want to evaluate.

Let's take a hypothetical example of an eight-team league. Here are the WS and IP of the team leaders:



Now, if I understand what you're doing, you would rank these pitchers 1-2-3-4-5-6-7-8, giving them scores of 9.5 WS above the median ace, then 7.5, 4.5, 1.5, -1.5, -4.5, -7.5, and -9.5. Using what I consider to be the "proper" approach of rating vs. average, you'd calculate the median WS/IP of the group at .107, multiply that by each pitcher's innings, and subtract, giving you this quite different ranking (WSAPM is Win Shares Above Positional Median):



This is what DL from MN does with PWAA. However, I think this goes too far the other way in rewarding ERA+ at the expense of IP, as you can see with Bridges.

What I think the way to go is, of course, to measure above replacement. If a replacement SP is roughly a .400 WPCT, that means a team of all replacement SP and average position players would be 16.2 wins below average per 162 games, which is 2.2 wins below average per 200 innings pitched, or 6.6 Win Shares below average per 200 innings pitched. If an average team wins 81 games and the pitchers get 70% of 52% of its 243 Win Shares, then an average pitching staff should have 88.5 Win Shares, or 12.1 WS per 200 innings pitched. So replacement is thus 5.5 WS per 200 innings pitched. (Does this sound right, WS experts?) So, my own ranking of this hypothetical group would be as follows:



I hope this helps to clarify the various ways to weigh rate and IP for the group.

I understand all the "caveats" about WS vs. "median ace." It's certainly not the only measure I use. It's a new toy. But frankly, where Bridges (among many many others) is concerned, it has reinforced what I was already thinking. The guys that surprised me are:

Better than expected--Hershiser, Cone, Saberhagen, the recent low IP guys, plus Walters, Newcombe, Wi. Cooper

Worse than expected--Joss and Cicotte, Cicotte and Joss (waaaaay worse), Gomez, Tiant

With Reuschel and Bridges, it was pretty much as expected, since I've always been a peak/prime voter.

My 2007 ballot is likely to have 4 pitchers on it.

Dean +39
Newcombe--probably +35 with various adjustments
Bond--haven't done
Walters--+39

What's this? A 19er on the agenda as a career candidate?

DanR
Van Haltren only had one season (ignoring the weak 1891 AA) when he finished in his league's top *eight* in SB--he was not posting the extremely high league-relative SB numbers that tend to correlate with an above-average success rate.

What is the time period for the study of SB counts and success rates?
Is Max Carey ~1920 the earliest important outlier?

DanG:
3) I think voters simply became disinterested in him. He lacks that Wow Factor. These 27 players somehow became better than him since 1972 [list deleted]

140. Delorians Posted: October 31, 2007 at 02:54 PM (#2601704)
In that 1972 election, GVH ranked 7th, Duffy 15th. All of the other top 20 have now been elected.

Interested readers should also revisit the
1932 Election Results

139. Chris Cobb Posted: October 31, 2007 at 02:42 PM (#2601683)
On VH's fielding: my recollection is that his range and assist totals are good, but his error totals are exceptionally high. I don't know the cause -- was his arm strong but scattershot, or did he have "bad hands"? Or am I totally misremembering the data?

What data do we have that helps distinguish "scattershot arms" from "bad hands"?

I did the study on every year that CS data is available for. No, Fritz Maisel went 74-for-91 (81%) stealing in 1914, when the league success rate was 55%. I have Maisel with a cool 1.6 baserunning wins that year.

Thanks, DanR.

Regarding the 1932 HOM election, let me note:
Louis Santop arrived that year, ending the long drought at the top (no first-tier candidates).

I have some measures of ballot strength meaning overall quality of eligible players in the estimation of the electorate. The early 30s show up as the trough, with 1932-33 below the late 1920s despite the first-tier new arrivals. The measures are based on election results. Briefly, if GVH fares better in 1932 than in 1929, the 1932 ballot is weaker.

Rollie Fingers is going to look crappy in Win Shares v. the median ace...

Is there any evidence that Bridges was rested for the playoffs in 1934-1935? If so then I think you'd have to consider his postseason in the analysis. I happen to think 34-36 is a pretty nifty 3 year consecutive peak.

'36 was Bridges' best season; it would have been Saberhagen's fifth-best. And Bridges' DERA was 4.13 in 1935--which is a 109 RA+. That can't possibly be called a peak season no matter how many innings you throw.

Regarding the 1932 HOM election, let me note:
Louis Santop arrived that year, ending the long drought at the top (no first-tier candidates).

Indeed, that is a grim, grim backlog. In fact, it doesn't even seem so much that the estimation of Van Haltren declined as
1)the quirky open door provided by chronological voting closed
2) the electorate has inexorably evoloved to a slightly peakier preference.

Other than Childs, did any 19th C guy leapfrog Van Haltren in many ballots?

GVH was third that year. Beckley, Jennings, and Griffith all thus finished behind him and eventually got in, as well as the aforementioned Childs.

From now on every time I read that Win Shares replacement level is too low I am linking to this thread:

Minor League Free Agents

I know, I saw that, it's the best list of FAT players I've ever seen--because, of course, they are by definition freely available!

Regarding the 1932 HOM election, let me note:
Louis Santop arrived that year, ending the long drought at the top (no first-tier candidates).

Indeed, that is a grim, grim backlog. In fact, it doesn't even seem so much that the estimation of Van Haltren declined as
1)the quirky open door provided by chronological voting closed
2) the electorate has inexorably evoloved to a slightly peakier preference.

Other than Childs, did any 19th C guy leapfrog Van Haltren in many ballots?

However grim it was, we elected quite a few from it over the years. Taking a "Where are they now?" look at the 1932 top 30 non-elected players is interesting:

1932 rank. Player -- 2006 rank or HOM year elected
3. George Van Haltren -- 17
4. Clark Griffith -- HOM 1971
5. Lip Pike -- HOM 1940
6. Jake Beckley -- HOM 1998
7. Hughie Jennings -- HOM 1960
8. Rube Waddell -- HOM 1986
9. Hugh Duffy -- 15
10. Roger Bresnahan -- HOM 2004
11. Mickey Welch -- 22
12. Cupid Childs -- HOM 1988
13. Jimmy Ryan -- 67
14. Pete Browning -- HOM 2005
15. Tommy Leach -- 14
16. Jose Mendez -- HOM 1985
17. Bill Monroe -- 46
18. Charley Jones -- HOM 2003
19. Spotswood Poles -- not ranked in 2006
20. Larry Doyle -- 35
21. Bobby Veach -- 79
22. Harry Hooper -- not ranked in 2006
23. Frank Chance -- 72
24. George J. Burns -- 64
25. John McGraw -- 9
26. Addie Joss -- 56
27. Ned Williamson -- 47
28. Gavvy Cravath -- 12
29. Eddie Cicotte -- 68
30. Dobie Moore -- HOM 1991
31. Vic Willis -- 26
32. Fielder Jones -- 96t

Yeah, I don't accept it as "grim." We certainly had lots of other options by '03-'04-'05 and yet we elected 3 guys from this backlog, plus Beckley in '98 and 3 others after 1980. I don't think any of them is our worst selection. Well, ok, there is Beckley. And it's a bit of a shock to notice that we elected Lip Pike how many years before Griffith and Browning and Jennings and Waddell, guys who most of us would probably rate ahead of the Lipster today.

And peakier? I mean, jeez, we elected Beckley. But yes, it seems that among the old timers there is a slight peak preference. I don't think it carries over (peakiness) to more modern electees. I would guess it is more of a coincidence that voters who like 19C and dead ball guys after all these years also happen to be peak/prime voters. It's not like the electorate "as a whole" elected any of these backloggers, they were niche candidates and the overlap in niches is just a coincidence. I think.

And it's a bit of a shock to notice that we elected Lip Pike how many years before Griffith and Browning and Jennings and Waddell, guys who most of us would probably rate ahead of the Lipster today.

Not me.

top 3 eligibles, plus roughly others of interest (cribbing as usual from DanG....)

2007 —elect 3
WS W3 Rookie Name-Pos (Died)
427 169.1 1982 Cal Ripken-SS
398 124.3 1982 Tony Gwynn-RF
342 109.5 1987 Mark McGwire 1B
307 102.4 1980 Harold Baines-RF/DH
280 105.4 1984 Tony Fernandez-SS
259 98.6 1987 Paul O’Neill-RF
205 94.2 1987 David Cone-P*

2008 —elect 3
WS W3 Rookie Name-Pos (Died)
390 131.8 1981 Tim Raines-L
213 98.3 1986 Chuck Finley-P
231 80.0 1991 Chuck Knoblauch-2B


and after a year's break.....
2009 —elect 3
WS W3 Rookie Name-Pos
535 178.1 1979 Rickey Henderson-LF
294 96.5 1988 Mark Grace-1B
245 92.0 1987 Jay Bell-SS
241 91.0 1987 Matt Williams-3B

and then

2010—elect 3
WS W3 Rookie Name-Pos
376 132.7 1988 Roberto Alomar-2B
346 122.2 1986 Barry Larkin-SS
341 104.6 1987 Fred McGriff-1B
305 108.3 1989 Edgar Martinez-DH/3B
272 110.1 1990 Robin Ventura-3B

2011—elect 3
WS W3 Rookie Name-Pos
395 137.4 1987 Rafael Palmeiro-1B
388 135.4 1991 Jeff Bagwell-1B
301 115.2 1990 John Olerud-1B
311 106.6 1990 Larry Walker-RF
241 106.0 1989 Kevin Brown-P

According to Win Shares (>300)

1. Rickey
(big gap)
2. Ripken
(small gap)
3. Gwynn
4. Raffy
5. Raines
6. Bags
7. Alomar
(gap)
8. Larkin
9. McGwire
10. McGriff
(gap)
11. Walker
12. Baines
13. Edgar
14. Olerud

According to WARP (>100)

1. Rickey
2. Ripken
(then the gap)
3. Raffy
4. Bags
5. Alomar
6. Raines
7. Gwynn (big drop)
8. Larkin
9. Olerud
10. Ventura (big gainer)
11. McGwire
12. Edgar
13. Walker
14. K. Brown (gains)
15. Fernandez (gains)
16. McGriff (big drop)
17. Baines (drop)

My Gut (and I'm a peak/prime voter)

1. Ripken
2. Rickey
(not such a big gap)
3. McGwire
4. Bags
5. Raines
6. Larkin
7. Alomar
8. Gwynn
(gap)
9. K. Brown--not sure he's a HoMer, Saberhagen is just as good, but maybe
(in/out line?)
10. Fernandez--not quite
11. Ventura
12. Raffy--no boycott, but not quite the peak of these other guys
13. McGriff--and the rest of these guys, well, there's too many better hitters
14. Edgar
15. Walker
16. Olerud
17. Baines

Wow, my system REALLY disagrees with both WARP and WS on the relative ranking of a lot of these modern guys. I have (career WARP2; no innings translation for pitchers; numbers in parentheses exclude sub-replacement seasons):

2007
Cal Ripken 101.8 (102.7)
Tony Gwynn 76.0
Mark McGwire 65.7
Harold Baines 37.4 (39.5)
Tony Fernández 53.1 (53.5)
Paul O'Neill 50.0
David Cone 59.6 (61.0)

2008
Tim Raines 80.0 (80.3)
Chuck Finley 51.9 (52.1)
Chuck Knoblauch 36.3 (37.3)

2009
Rickey Henderson 111.0 (111.4)
Mark Grace 43.2 (44.4)
Jay Bell 43.3 (45.6)
Matt Williams 43.6

2010
Roberto Alomar 60.1 (61.7)
Barry Larkin 85.2
Fred McGriff 48.7 (49.4)
Edgar Martinez 56.6 (57.5)
Robin Ventura 55.3

2011
Rafael Palmeiro 68.1
Jeff Bagwell 79.4
John Olerud 50.2
Larry Walker 63.4 (63.7)
Kevin Brown 67.2 (68.2)

So, I went to Cooperstown this weekend and met Luis Tiant. He seems like a great person, and he loves to tell baseball stories. It was a small private reception with Tiant and Benny Ayala, so we really got to talk to both them.

On the 23rd, Tiant is going back to Cuba for the first time in 46 years.

If anyone is interested, it appears that in the latest iteration of WARP BP has substantially increased the standard deviation of outfield FRAA.

To wit, Babe Ruth's 1923 comes in at a cool +41 FRAA. #### you not.

This new FRAA may make some bad fielders less attractive (Manny: -138 for career!) and bump up good fielders (I think Dom DiMaggio has a legit case with war credit now).

Manny's a perfectly decent fielder; that projection simply shows up what rubbish these sabermetric numbers are. 9 out of 10 agree quite well with "real" numbers but don't add anything; the tenth is cuckoo like Babe 1923 in your example.

Whoa...if I might chime in on THAT subject...http://www.nytimes.com/2007/03/18/sports/baseball/18score.html.

Manny's a perfectly decent fielder

Only through the rosiest of rose colored glasses.

If anyone is interested, it appears that in the latest iteration of WARP BP has substantially increased the standard deviation of outfield FRAA.

To wit, Babe Ruth's 1923 comes in at a cool +41 FRAA. #### you not.

This new FRAA may make some bad fielders less attractive (Manny: -138 for career!) and bump up good fielders (I think Dom DiMaggio has a legit case with war credit now).

Gee if we wait one more week to vote maybe they will change again and Mike Hargrove will look electable too!

Of course the Manny being Manny, and being a trully horsebleep fielder, was written in March of 2007, since which time the Sox won another world championship. I'm not sayin' Manny is a "perfectly decent fielder" by any means. Just that on balance, he's a pretty valuable guy to have around.

Speaking of voting, aren't we supposed to have a ballot this morning?

Speaking of voting, aren't we supposed to have a ballot this morning?

It's up, Ron.

bump