Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Monday, December 19, 2016

2018 Hall of Merit Ballot Discussion

Vladimir Guerrero (267), Sammy Sosa (255), Luis Tiant (189), Jeff Kent (179), Vic Willis (161), Kenny Lofton (151), Bobby Bonds (133), Ben Taylor (130), Buddy Bell (130) and Tommy Bridges (106) will be the top 10 returnees for 2018.

Jorge Posada (95), Bob Johnson (86), Urban Shocker (80), Dick Redding (79), Phil Rizzuto (79), Wally Schang (79) and Sal Bando (68) rounded out the top 20.

2018 - (December 4 - December 18, 2017) - elect 4

Name               HOFm HOFs Yrs WAR  WAR7 JAWS
Chipper Jones       180  70   19 85.0 46.6 65.8
Jim Thome           156  57   22 72.9 41.5 57.2
Scott Rolen          99  40   17 70.0 43.5 56.8
Andruw Jones        109  34   17 62.8 46.4 54.6
Johan Santana        82  35   12 51.4 44.8 48.1
Johnny Damon         90  45   18 56.0 32.8 44.4
Jamie Moyer          56  39   25 50.4 33.2 41.8
Carlos Zambrano      30  23   12 44.6 39.0 41.8
Omar Vizquel        120  42   24 45.3 26.6 36.0
Chris Carpenter      70  26   15 34.5 29.6 32.0
Livan Hernandez      41  16   17 31.1 27.8 29.4
Orlando Hudson       20  18   11 30.9 27.2 29.1
Kevin Millwood       34  20   16 29.4 24.8 27.1
Kerry Wood           24  14   14 27.7 25.0 26.4
Carlos Lee           78  35   14 28.2 23.4 25.8
Ben Sheets           19  11   10 23.4 22.3 22.8
Jack Wilson          12  16   12 23.5 20.9 22.2
Hideki Matsui        36  21   10 21.3 21.2 21.3
Aubrey Huff          30  20   13 20.2 22.5 21.3
Adam Kennedy         12  16   14 21.0 20.4 20.7
Jeff Suppan          11   9   17 17.4 18.3 17.8
Carl Pavano          16   6   14 16.9 18.5 17.7
Francisco Cordero    77   9   14 17.2 14.6 15.9
Miguel Batista       10   3   18 12.7 15.9 14.3
Jason Isringhausen   71   7   16 13.2 12.2 12.7
Brian Fuentes        48   9   12 10.7 11.3 11.0
Brad Lidge           48  10   11  8.2 12.4 10.3
Scott Podsednik      15  15   11  6.9  7.8  7.4
Guillermo Mota       13   7   14  6.3  7.6  7.0
JoeD has the Imperial March Stuck in His Head Posted: December 19, 2016 at 09:12 PM | 415 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

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

Page 3 of 5 pages  < 1 2 3 4 5 > 
   201. Rob_Wood Posted: April 28, 2017 at 04:47 PM (#5444517)
There is no way you are going to get me to swallow that Tommy Bond (with 5 decent seasons) is more valuable than Barry Bonds because he was able to take the mound and not walk anyone.

My position on this matter is clear. I have voted in each and every Hall of Merit election. Our first election was 1898 and our latest was 2017, making a grand total of 120 separate elections. In each election a voter votes for his top 15 players. 120 * 15 = 1,800. So since this project began, I have listed 1,800 names on my collective body of Hall of Merit election ballots.

In none of my 1,800 ballot slots have I ever listed the name of Tommy Bond. And I will not be including the name of Tommy Bond on any subsequent future ballot regardless of the information I posted above.

I don't know how I can make it any clearer. WAR is a problematic stat for 19th century pitchers. I will repeat it a thousand times if need be. As CPASR is 100% hitched to WAR, CPASR is a problematic stat for 19th century pitchers. I included them in the above lists for completeness purposes only. If you'd like to do what I do, you can mentally cross out any 19th century pitcher you see on the list and pretend that you never saw those names or figures.

Please don't throw the CPASR stat into the dustbin simply because you believe that WAR for 19th century pitchers are silly.
   202. Rob_Wood Posted: April 30, 2017 at 03:10 PM (#5445149)
I have calculated the CPASR for the additional players that Bleed the Freak suggested.

Bobby Mathews       2.175 (19th century pitcher, ignore?)
Charlie Buffinton   1.410 (19th century pitcher, ignore?)
John Clapp          0.706
Dave Orr            0.630
Jack Rowe           0.622
Orator Shafer       0.589
Abner Dalrymple     0.488
Mickey Tettleton    0.428
Del Crandall        0.408
Darren Daulton      0.379
Brian McCann*       0.360
Tug McGraw          0.323
Doug Jones          0.320

   203. Bleed the Freak Posted: April 30, 2017 at 07:02 PM (#5445302)
Nice turnaround time Rob!, 11 others I have with solid ratings in Baseball-Reference WAR:
Davy Force
Paul Goldschmidt
Tip O'Neill
John Morrill
Fred Pfeffer
Whit Wyatt
Levi Meyerle
Gary Peters

3 others:
Tony Pena - catcher
Sparky Lyle - reliever
Manny Machado - active guy
   204. Kiko Sakata Posted: May 02, 2017 at 11:10 PM (#5447321)
Thanks for all of this, Rob! I really like the concept of a sliding replacement value. I definitely like the idea that in-season replacement level is lower than "offseason" (?) replacement level. I also like the concept of pennants added. I need to sit down with your formulas and see if I can figure out a way to make it work with my Player won-lost records. But very interesting food for thought.
   205. Rob_Wood Posted: May 02, 2017 at 11:43 PM (#5447338)
Thanks for the nice comment Kiko. It goes without saying that I am a big fan of all of your stuff. (Ditto for Bleed the Freak.)

One of the most rewarding aspects of the Hall of Merit has been all the great research associated with the project (coming from many great researchers). I don't want to name names for fear that I will overlook somebody, but there has been a heck of a lot of great research. Research on minor-league equivalents, negro-league players, 19th century, scarcity/replacement-levels, pennants-added, and many other key elements of baseball.

Anyway, here is the latest batch of new CPASR for the additional players Bleed the Freak listed just above.
Davy Force        0.667
Levi Meyerle      0.577
Tip O'Neill       0.552
John Morrill      0.492
Paul Goldschmidt* 0.479
Gary Peters       0.424
Whit Wyatt        0.419
Manny Machado*    0.405
Fred Pfeffer      0.367
Sparky Lyle       0.337
Tony Pena         0.295
Of course, any and all questions about the entire body of CPASR results or the methodology are more than welcome.

   206. Bleed the Freak Posted: May 03, 2017 at 04:20 PM (#5447882)
Thanks Rob, if you decide to expand into Kiko's W-L or Baseball Gauge DRA WAR, I have some additional suggested candidates (although they don't fair so well with Baseball Reference WAR:

Kiko: Mike Hampton, Bret Boone, Dick Donovan, Greg Luzinski, Vic Raschi, John Wetteland, Dave Righetti, Jayson Werth, Fritz Peterson, Bob Forsch, Shane Reynolds, Edwin Encarnacion.

Gauge: Mike Marshall, Bob Stanley, Jason Varitek, John Roseboro, Shin Soo-Choo, Kyle Seager, Nelson Cruz, Hank Sauer, Burt Shotton, Dutch Ruether, Happy Felsch, Giancarlo Stanton.
   207. Bleed the Freak Posted: May 05, 2017 at 03:37 PM (#5449481)
Hey Kiko, he's long been elected, but Willie Mays fairs quite poorly in your component 5 measure, hits versus outs on balls in play.

http://baseball.tomthress.com/StatTables/FieldingStats.php?id=maysw101
He shows as 19th worst in component 5 fielding wins: http://baseball.tomthress.com/Leaders/Leaders.php?y=&y1;=&y2;=&l=&a=c&n=99&s=c5
Is he really that bad, did he have less chances with the pitching staffs in his career, some mix or otherwise?

And it's a small thing, but an error is coming up when you click on component fielding links:
http://baseball.tomthress.com/StatTables/Glossary/Component5.html

The links are broke in the fielding article for the components:
http://baseball.tomthress.com/HallofMerit/Fielding.php
Incorrect link:
http://baseball.tomthress.com/Articles/Component6.php
Correct link:
http://baseball.tomthress.com/HallofMerit/Component6.php
   208. Bleed the Freak Posted: May 07, 2017 at 08:40 PM (#5450412)
Recent discussion on a fine Negro Leaguer with no thread: Rev Cannady:
http://www.seamheads.com/NegroLgs/player.php?playerID=canna01wal

http://www.baseball-fever.com/showthread.php?130322-HOF-cases-of-Negro-Leaguers-using-MLEs-of-ERA-and-OPS-from-Seamheads&p=2634871#post2634871

http://www.baseball-fever.com/showthread.php?130322-HOF-cases-of-Negro-Leaguers-using-MLEs-of-ERA-and-OPS-from-Seamheads&p=2634810#post2634810
   209. Bleed the Freak Posted: May 11, 2017 at 12:02 PM (#5453478)
84. Kiko Sakata Posted: January 09, 2017 at 11:29 PM (#5380804)
Finally, shameless self-promotion: coming in June from McFarland, Player Won-Lost Records in Baseball: Measuring Performance in Context


Hello Kiko, what website do you receive the most $ from, I see a number of different avenues to purchase this?

Thanks!
   210. Kiko Sakata Posted: May 11, 2017 at 08:27 PM (#5453868)
Hello Kiko, what website do you receive the most $ from, I see a number of different avenues to purchase this?

Thanks!


To be perfectly honest, I don't know. If I had to guess, I would guess going through McFarland, but I don't know. By all means, go with whatever makes the most sense for you (i.e., is cheapest?). I'm not really in it for the money: I'm fairly sure the market for this book won't allow me to retire and live off the royalties (now, my second book - kidding).
   211. Kiko Sakata Posted: May 11, 2017 at 08:42 PM (#5453880)
Hey Kiko, he's long been elected, but Willie Mays fairs quite poorly in your component 5 measure, hits versus outs on balls in play.


Comments #209-#210 segue nicely into my answer here. I actually wrote a bit about this very topic in the book. Short answer: it's entirely possible that I'm wrong and the honest answer is that I don't know exactly why Mays fares so poorly. Longer answer / attempt to defend my ratings: buy the book - okay, summary: Mays was very good at Component 5 when he was very young (through 1956, age 25); it's possible that he lost a step or two as he aged and people either didn't notice or didn't care because, you know, he's Willie Freaking Mays out there.
   212. Bleed the Freak Posted: July 02, 2017 at 11:28 AM (#5486884)
Kiko, looks like the publication date has been pushed back for your book, do you know when I can anticipate this being available?
Hoping to have for my birthday ! :)
   213. Kiko Sakata Posted: July 17, 2017 at 09:16 AM (#5494532)
Kiko, looks like the publication date has been pushed back for your book, do you know when I can anticipate this being available?
Hoping to have for my birthday ! :)


Sorry it's taken me two weeks to respond to this. I don't know. I talked to my editor and he said that he has several books that seem to be moving more slowly through production than he expected. I think he's still hoping for some time this summer. He told me I should expect proofs in a week or two. But (a) I don't know how likely that is if things are moving more slowly in general than normal, and (b) this is my first book, so I don't know how long it takes to go from proofs to an actual book. I'm definitely planning to post something - probably here, on Twitter, and on my website - when it's available.
   214. Kiko Sakata Posted: August 01, 2017 at 08:51 PM (#5504767)
Bleed and anybody else interested, I got proofs for my book today! It's scheduled to go to the printer "by August 31st". Here's the page for it on McFarland's website: Player Won-Lost Records in Baseball: Measuring Performance in Context
   215. Kiko Sakata Posted: September 13, 2017 at 11:00 AM (#5530724)
Follow-up to #214. My book has been published. Player Won-Lost Records in Baseball: Measuring Performance in Context, published by McFarland, should be available wherever fine books are sold.
   216. Bleed the Freak Posted: September 16, 2017 at 06:36 PM (#5533145)
Thanks kiko, i placed my order with McFarland.

I was pondering bob lemon the other day, he is a maligned hall of meriter through the lens of baseball reference and baseball gauge, but he is rather impressive with your win loss records. His era is much better than his fip, is this a case where lemon is apportioned more credit for balls in play/inducing weak contact by your methods compared to others?
   217. Kiko Sakata Posted: September 17, 2017 at 12:23 AM (#5533279)
Thanks for ordering the book, Bleed!

Lemon has a few things going for him in my rankings. He was a great hitter - which other systems should include, although BB-Ref separates Batting WAR and Pitching WAR which can make it easy to fail to include it. He looks much better in context by my numbers than out of context (40.7 pWORL vs. 32.6 eWORL). And, yes, he was quite good at the non-FIP components of pitching. Fangraphs shows this most strikingly, I think. They actually calculate WAR two ways: based on FIP - which gives Lemon 32.3 WAR - and based on RA/9 - which gives Lemon 52.6 WAR - 20 more WAR! - which improves his ranking on Fangraphs among pitchers from 258th in FIP-based WAR (which is, of course, their preferred WAR) to 120th in RA-based WAR.

And then, raw career WAR understates Lemon's value, since he had a relatively short career, in part due to World War II (he served for all of the 1943, 1944, and 1945 seasons - he was in the minor leagues in 1942 (as a 3B)).

Without getting into World War II credit, using my preferred weighting of things (pWins vs. eWins; WOPA vs. WORL; etc.) he ends up in the same general vicinity as Roy Campanella (also with no extra credit), Roy Halladay, Early Wynn, and Rod Carew, to pick four players who have virtually nothing in common except that they're all fairly solid mid-tier Hall-of-Famers/Meriters (well, Halladay will be once he's eligible).
   218. Bleed the Freak Posted: September 18, 2017 at 08:05 PM (#5534026)
Thanks Kiko.

Regarding your ratings versus B-R and B-G, is it mostly attributed to the numbers he had in context?
Per your suggestion, I've been using 1/3 pWORL and 2/3 eWORL, Lemon vaults into the top 60 hurlers of all-time.
When I use B-R and B-G, he falls in the low 100s range.

If it's not just context, does he excel in some other ways you capture differently/better than the other metrics?


   219. Howie Menckel Posted: September 18, 2017 at 08:40 PM (#5534048)
I could have sworn during our HOM discussion on Lemon many years ago that he had an amazing 100 career OPS+. now it is listed at 82.

so I misremembered it - or is it something else?

in 419 AB from 1947-50, Lemon had 20 HR, 71 RBI, and OPS+s of 177, 119, 134, and 113. he also was 76-40 in that span for an average W-L of 19-10. 144, 133, 112, and 108 ERA+s and led the AL in IP twice.

not true math, but has a 119 ERA+ and 82 OPS+. I wonder how many other Ps have a combined figure over 200? Sandy Koufax ain't one.
   220. Kiko Sakata Posted: September 18, 2017 at 10:06 PM (#5534093)
Regarding your ratings versus B-R and B-G, is it mostly attributed to the numbers he had in context?


That's certainly part of it. Lemon was very good on balls in play - the evidence suggests that he was very good at inducing weak contact. Per BB-Ref, for his career, MLB had an average batting line of .259/.332/.370 with a BABIP of .276. Lemon allowed a line of .240/.320/.337 with a BABIP of .258. So his BABIP allowed was .018 lower and his ISO was .097 vs. a lg-ISO of .111. I'm not entirely sure why that wouldn't show up to his credit in BB-Ref, though, since they base their pitcher WAR on runs allowed.

I'm not as familiar with Fangraphs - but there, the story is the non-FIP stuff which I discussed above - but, in BB-Ref, it's really hard to get them to give you a set of WAA and/or WAR that include both pitching WAR and batting WAR. In the play index, if you look at WAR numbers for pitchers, it's pitching only. Which doesn't matter for most pitchers, but Bob Lemon is very much not most pitchers: his batting was worth 11.3 WAR (and/or 11.3 WAA) per BB-Ref. I have him as the second-best hitting pitcher for whom I've calculated Player won-lost records (i.e., basically since 1930) behind Wes Ferrell - so, best-hitting pitcher since World War II and/or since integration.

As far as my personal ratings, he's also being measured as more of a prime candidate. He was very good from 1948 - 1956. Controlling for context, he's about the 7th-best player in MLB over that time period, the 3rd-best pitcher, and the best American League pitcher, measured by either eWOPA or eWORL. I'm not sure how to do a similar leaderboard for BB-Ref because of the issue w/ pitcher batting (which would also affect Warren Spahn).
   221. Bleed the Freak Posted: September 19, 2017 at 11:53 PM (#5534885)
In BB-Ref, it's really hard to get them to give you a set of WAA and/or WAR that include both pitching WAR and batting WAR.


Maybe The Baseball Gauge (Seamheads) has a search tool you are looking for:
WAR leaderboards
WAR and WAA are available for Baseball Reference and Gauge, and you can mix and match other ways too.

A downloads section is also available for baseball gauge for all seasons:
Baseball Gauge WAR download

A polarizing candidate has been Wilbur Wood for me.
To the wayback machine, Wood was underwhelming by Baseball Prospectus WARP and Joe Dimino's PA.
Conversely, Wilbur at Baseball-Reference and Baseball Gauge show him as a mid-level or lower end HOF quality player.
His Fangraphs FIP are a bit shy of a large HOF, his WPA and Baseball Prospectus DRA are terrible.
When I look at Win-Loss records, I see a pitcher outside the top 200 all-time.

What are your thoughts on Wood?
   222. The Honorable Ardo Posted: September 22, 2017 at 12:46 AM (#5536550)
Wood's entire case is 1968-73. Because he had so much bulk, he effectively packs four typical seasons of value into three seasons (both as a relief ace and a front-line starter). Yet he adds so little value outside of those six seasons that I can't justify giving him a ballot spot.
   223. Bleed the Freak Posted: November 06, 2017 at 06:43 PM (#5571927)
Apologies, I posted to 2017 thread, belongs in 2018:

Passing along that Retrosheet is planning to have fielding splits available, probably not until mid 2018 though:

Retrosheet Fielding Splits
   224. djrelays Posted: November 07, 2017 at 10:12 AM (#5572149)
223:
Passing along that Retrosheet is planning to have fielding splits available, probably not until mid 2018 though:

Retrosheet Fielding Splits


Interesting news. Any chance they can create ground ball splits that separate grass from turf?
   225. Rob_Wood Posted: November 12, 2017 at 01:25 PM (#5574680)
please ignore



<quote>

Pitcher          CPASR    WVA    WVR
Roger Clemens    2.065    82.4   123.3
Randy Johnson    1.553    54.0    88.2
Greg Maddux      1.540    65.2   106.9
Pedro Martinez   1.333    51.6    74.3
Bert Blyleven    1.325    34.6    75.9
Phil Niekro      1.299    23.7    66.6
Steve Carlton    1.249    33.6    76.6
Mike Mussina     1.208    39.0    68.7
Gaylord Perry    1.193    34.7    77.6
Fergie Jenkins   1.145    34.4    70.7
Curt Schilling   1.129    38.4    64.1
Tom Glavine      1.060    37.9    74.7
Roy Halladay     1.010    34.2    56.6
Kevin Brown      0.997    32.1    59.0
Rick Reuschel    0.989    22.0    51.2
Jim Palmer       0.974    45.7    77.8
David Cone       0.971    27.2    50.8
John Smoltz      0.967    30.8    57.6
Luis Tiant       0.904    28.1    56.1
Nolan Ryan       0.877    33.1    77.5
Dave Stieb       0.870    22.8    43.3
Jim Bunning      0.845    20.3    50.5
Kevin Appier     0.839    26.8    48.2
Johan Santana    0.813    28.0    43.5
Tim Hudson       0.785    30.7    56.7
Orel Hershiser   0.784    21.2    46.7
Bret Saberhagen  0.778    28.3    49.0
Chuck Finley     0.778    23.5    49.1
Andy Pettitte    0.776    25.9    53.4
Roy Oswalt       0.765    28.1    46.5
Don Sutton       0.677    28.7    72.5
Tommy John       0.595    23.7    62.2
Catfish Hunter   0.533     8.0    36.3
Jack Morris      0.519    19.0    50.2


</quote>

xxx
   226. theorioleway Posted: November 13, 2017 at 07:59 PM (#5575525)
In case people haven't paid attention to other threads, the good Doctor is running new MLEs for Negro League players. Of importance for this election are the ones for Ben Taylor and Hilton Smith, both of whom look like they deserve serious considerations for being on the ballot.
   227. bachslunch Posted: November 15, 2017 at 10:42 AM (#5576529)
Second time voting, here's what I've got thus far.

Disclosures: am going with Seamheads for Negro Leaguers. Preference for BBRef WAR with some influence of OPS+ and ERA+ for the rest. Am valuing hitting prowess at C, SS, 2B, CF a bit extra. Being best available candidate at your position helps also. Still trying to sort out peak vs. longevity, but often favoring the latter. Fine with giving Negro League credit, unsure so far on War or Minor League credit. Not docking for WW2 play as of now. Do not give postseason credit. Not sure how to evaluate 19th century pitchers, but for now tending to discount AA, NA, and UA stats as possibly suspect, with one exception. Not taken with giving relievers a lot of emphasis. Will dock 1st year candidates who bet on games, threw games, impeded players of color, were caught using PEDs post-2005 (Manny, ARod), and likely used pre-2005 if it looks like they'll get an immediate free pass by BBWAA HoF voters (IRod, Ortiz, Pettitte).

1. Chipper Jones. Best WAR of anyone, and at a relatively premium position.
2. Scott Rolen. Third best WAR but not far off Thome, also at a relatively premium position.
3. Jim Thome. Second best WAR of everyone, though not at a premium position.
4. Jim McCormick. Best WAR for starters not in by a mile. Short career, but played in NL except for one UA season.
5. Luis Tiant. Best WAR for non-19th century starters.
6. Buddy Bell. Best WAR at 3B until this year. Currently inclined to trust the metric for him.
7. Andruw Jones. Best CF WAR. Close between him and Bell for me.
8. Vladimir Guerrero. Better WAR than I remember, thought he'd go lower.
9. Sammy Sosa. Again, better WAR than I remembered. Happy to give him some benefit of the doubt given his treatment by the BBWAA.
10. Bob Johnson. Best WAR among available LFs.
11. Wally Schang. Among best C WAR, also hit well.
12. Vic Willis. Good WAR.
13. Ben Taylor. Best NGL position player per Seamheads.
14. Dick Redding. Best NGL pitcher per Seamheads.
15. Vern Stephens. Among best SS WAR, also hit really well.

16-25: Kenny Lofton, Tommy John, Sal Bando, Urban Shocker, Thurman Munson, Tommy Bridges, John Olerud, Jim Fregosi, Bert Campaneris, Bobby Bonds.

Required comments for those outside my 25. Fred McGriff and Bob Elliott drop off last year's top 25, too much competition. Johnny Damon has much better WAR than I remember, but is helped a lot by a long career, also has OPS+ just above league average; can't really rank him above Bonds. Johan Santana has decent war, but is all peak, and tend to prefer career for pitchers; might reconsider him, though. Gavvy Cravath has an amazing OPS+ but less taken with his career brevity, might be swayed to move him up. Phil Rizzuto doesn't do much for me (low OPS+, low WAR, short career). Bucky Walters also doesn't sufficiently impress me (good but not top of the line WAR or career ERA+). Bad defense and play calling keeps Jorge Posada out of top 25 for me, though has good WAR numbers relative to position.
   228. bachslunch Posted: November 15, 2017 at 02:16 PM (#5576674)
No idea why, but I left Jeff Kent off last year and this time. Need to fix that.

Change #15 to Kent and drop Vern Stephens to #16. Kent has more WAR and a better OPS+ at a premium position and also more PAs than Stephens. That puts Bonds off my top 25.

My bad.
   229. Dr. Chaleeko Posted: November 16, 2017 at 10:29 AM (#5577179)
Bach’s ballot brings up a challenge to the electorate from me. This is an intentionally provocative post, forgive me if I go too far in that direction. Not trying to be a dick.

How many pitchers do we really need from the period of roughly 1880-1892? Currently we have Keefe, Caruthers, Hoss, Galvin, and Clarkson who are square in that zone. Plus Young, Nichols, and Rusie with multiple full seasons prior to 1893. IIRC (and I’m in a STL hotel room and have 15 minutes before my next mtg) at the beginning of the 1880s, teams were effectively using 1.5 man “roatations.” By the end of this era they were at 2.0 to 2.5 man roataions. That suggests that there’s approximately 400 annual rotation slots across all teams during that era. Our honorees occupy approximately 75-100 of those slots. Do we really need to honor more 1880-1892 pitchers than that? Especially since we have way too many 1800s players to begin with?

Welch, McCormick, Mullane, and Buffinton continue to receive some support, and I would suggest that their supporters ought to provide substantial evidence that we need to honor a significantly higher percentage of 1880-1892 pitchers than we should from other eras. IMO, “my system really likes him” (pls forgive that absurdly reductive straw man like paraquote) and can’t be strong enough reasoning if we are trying to be fair to players from all eras and all positions. There are many deserving players from other positions and eras who could use the dozen or so votes going to pre-1892 pitchers.
   230. DL from MN Posted: November 16, 2017 at 03:50 PM (#5577572)
I agree that the 1880s and 1890s have been covered, but so have the 20s and 30s and your NGL data is bumping up even more from that era. I haven't seen anyone revised down with the new NGL MLEs which makes me suspicious that there is something biasing the numbers too high.

Honestly, we should be mostly electing players from 1970 onward with an occasional backlog person every decade.
   231. Dr. Chaleeko Posted: November 16, 2017 at 04:19 PM (#5577596)
There are two items to respond to here:
1) Inflated MLEs. There are four reasons I think that describing them as inflated would be potentially inaccurate:
A) It suggests the previous versions were definitely accurate, which we don’t know for certain.
B) We have MUCH more and better data than we did 10 years ago. We didn’t have leaguewide OBP or walks data ten years ago, for example.
C) Previous MLEs were not denominated in WAR. These are. WAR was barely a thing in 2007. I tried to include it in my last couple MLEs back then, but it wasn’t a full implementation by any means. And comparing WS to WAR is something of an apples-to-pears situation.
D) I’m so far only rolling out MLEs for the best of the best among Negro Leaguers. When tertiary or quaternary candidates’ MLEs are rolled out, you’ll see that the MLEs for many of them are not so rosy.

In addition, I’m not sure that we can compare my new MLEs to my old ones (nor Brent’s or Chris’) because we were all attempting to translate component stats which a really different process with different assumptions.

I’m far more confident in my current ones than my old ones. But I’m not so confident that I wouldn’t be glad to hear feedback on the method, which you can find in the linked articles.

2) 1920s and 1930s: I am not advocating specifically for this time period, and I’m not sure what in my post about 1880s pitchers would suggest I do. In fact, the question of the MLEs seems pretty unrelated to me. But if I’m sorry if that was what I communicated. I’m on record in other threads that catchers, third basemen, deadballers, and latter day guys are on the list of underserved players. Specifically I have mentioned Wally Schang and Thurman Munson and Buddy Bell as good candidates for this reason. Actually, 2018 should help at 3B. I have a hard time believing we won’t elect Chipper Jones, and Scott Rolen could be elected too.
   232. bachslunch Posted: November 16, 2017 at 04:52 PM (#5577624)
@229: fair point to bring up. My feeling is that there are going to be periods with an overload of excellent players at certain positions and lacking good options at other positions in other eras. There’s the glut of great 3B from the 50s-80s and the lack of good catcher choices from 1900-1920, for example. I’m choosing to take the numbers literally, at least to a fair extent.
   233. Dr. Chaleeko Posted: November 16, 2017 at 07:49 PM (#5577723)
@232

So I went and did the math. From 1880-1892, HOM SP accounted for 11.9% of all the league’s starts annually with a high of 19.9% in 1881. Actually the number would be higher overall for this epoch but i included the UA.

Remember that’s 11.9% annually.

I looked at 1980 because so many of those long-career 300 winners were around. The 12 HOM SP in the league (including Eck) totaled 9.4% of all starts. The leader in GS in the league had 38. In order to match 11.9%, in 1980, 13 HOM SP would each have needed to start 38 games.

I repeated this for 1998. In that season, 11 HOM SP combined for 7.5% of all MLB GS. The leader had 35 GS, and the league would have needed 16 pitchers making 35 starts to match 11.9%.

Finally, I looked at the deadball era. I did get 12.1% for 1904, but 1916 was 8.7%. 1907 or 1908 might also be higher yield years but I chose 1904 to maximize the number of starts in the league. (I’m on a phone in a hotel room.)

Overall, it looks to me like the 1880s are already the most HOM-rsaturated period for SP because even when I tried to stack the deck, I topped out at the AVERAGE of the 1880s. And it would be difficult in my mind to see a way in which adding more SP would make sense even from the perspective othat some eras just have more guys at certain positions. Which is a theory I expressly agree on, BTW, but which I don’t think fits the evidence in this instance.

   234. Dr. Chaleeko Posted: November 16, 2017 at 08:12 PM (#5577729)
Forgot to say that racking on one of Mullane, Welch, Buffinton, or McCormick brings the percentage to 13.7% of all games started in MLB from 1880-1892.
   235. DL from MN Posted: November 16, 2017 at 09:24 PM (#5577773)
When tertiary or quaternary candidates’ MLEs are rolled out, you’ll see that the MLEs for many of them are not so rosy.


This suggests we need to adjust for standard deviations.
   236. Howie Menckel Posted: November 16, 2017 at 10:50 PM (#5577806)
old school point, and HOM voters may know that I have no pet projects left - and I think rarely did - but is it possible that a SP who pitches a huge proportion of innings is more valuable than one who doesn't?

not my hill to die on, but the angle should probably be dismantled at least. kind of a thought exercise.
   237. Dr. Chaleeko Posted: November 17, 2017 at 12:14 AM (#5577844)
@235, STDEV adjustments are baked into the protocol. Like I said, so far you are only seeing the best players’ MLEs.

@236, this is an interesting and important question to me too. Basically, I reckon that if everyone is doing it, it ain’t that special. The reason we see it as special is because we have trouble comprehending those superhuman innings totals. But the game, with one brief exception in the 1960s/1970s, has constantly evolved away from the model of the 1880s. Why? Because a) the longer mound distance of 1893 made it more physically difficult to pitch b) schedules got longer, necessitating more arms c) a team’s tactical options were increased by more frequent substitution for pitchers, especially the lesser starters.

Whatever sense we have that pitchers were more valuable at the time might be tempered by the fact that the mound was closer to the plate, making it easier to pitch effectively on a physical basis. (And weren’t pitchers allowed to bound toward the plate as they delivered the pitch? (i.e.: there was no rubber they had to remain in contact with but rather a box that they had to remain in.) And they threw underhand/sidearm, which is easier on the arm. If pitching hadn’t been less physically demanding, the 1880s guys couldn’t have lasted as long as they did! Both within a season and across seasons.

So my line of thinking goes that:
1) everyone did it in that time
2) it was easier physically to pitch
3) the game has near constantly moved away from the 1880s model, so how valuable was it really?
4) then why didn’t we reduce the number of hitters from that period commensurately?




   238. DL from MN Posted: November 17, 2017 at 02:08 PM (#5578114)
A ranking of NGL players by your new MLEs would be helpful to me to see players in the context of their peers.
   239. Rob_Wood Posted: November 17, 2017 at 11:33 PM (#5578306)
I am starting to compile my preliminary thoughts on this ballot. First, here are my thoughts on the first-timers on the 2018 HOM ballot.

To supplement the data that JoeD posted at the head of this thread, here is the value of my new CPASR stat for each of the newbies on the ballot worthy of consideration. (CPASR, which stands for Career Pennants Added using a Sliding Replacement level, was introduced around post #180 in this very thread. It is based upon and supplemental to JoeD's own pennant-added stat. Below, in addition to CPASR, I show the WAA and WAR figures from Baseball-Reference.com on which it is based.)
PLAYER           WAA  WAR  CPASR
Chipper Jones     53   85   1.17
Scott Rolen       44   70   1.00
Jim Thome         37   73   0.96
Andruw Jones      36   63   0.88
Johan Santana     33   51   0.81
Carlos Zambrano   29   44   0.69
Johnny Damon      19   56   0.67
Chris Carpenter   15   34   0.47
Orlando Hudson    13   31   0.41
Jamie Moyer       13   50   0.41
Kerry Wood        17   28   0.40
Omar Vizquel       5   45   0.37
Livan Hernandez    6   31   0.32
Kevin Millwood     6   29   0.31

Chipper Jones is a no-brainer. Electing Chipper will help fill-in our deficit at third base.

Scott Rolen is also sure to be high on my ballot (probably second). Very solid offense with outstanding defense at an important defensive position. I don't want to get carried away, but electing Rolen will help fill-in our deficit at third base.

Jim Thome will also be high on my ballot. Everybody knows about his power. Lifetime on-base percentage of .402 and OPS+ of 147.

Andruw Jones's case rests squarely on his other-worldly defensive statistics early in his career. I will undoubtedly have him on my ballot, but I am not sure how high.

Johan Santana is difficult. As many know, I have a parallel stat to WAA based upon a game-by-game evaluation of every start in Santana's career called Win Values Average (WVA). Santana's WVA is 28, below his WAA of 33. It is not unusual for a pitcher's WVA to be less than his WAA. But this tells me that a game-by-game inspection of Santana's career does not uncover significant hidden value that could boost his place on my ballot. Also, for what it is worth, Santana's Win Values Replacement (WVR) is only 44 reflecting, of course, his short career.

Carlos Zambrano will not be on my ballot.

Johnny Damon will not be on my ballot.

Chris Carpenter will not be on my ballot.

Orlando Hudson will not be on my ballot.

Jamie Moyer will not be on my ballot.

Kerry Wood will not be on my ballot.

Omar Vizquel will not be on my ballot. (People in the other BTF thread claim that Vizquel will do well on the BBWAA ballot.)

Livan Hernandez will not be on my ballot.

Kevin Millwood will not be on my ballot.



   240. The Honorable Ardo Posted: November 18, 2017 at 05:49 PM (#5578420)
I certainly don't see any need to induct more 19th century pitchers and haven't voted for one in 50+ ballots. Dolf Luque and Hilton Smith are the only pitchers I feel strongly about; my ballot this year will feature them, Johan Santana, and twelve position players.
   241. JoeD has the Imperial March Stuck in His Head Posted: November 21, 2017 at 10:21 AM (#5579302)
Hey guys!

What do we think of starting the voting December 4, and running it through December 18?
   242. JoeD has the Imperial March Stuck in His Head Posted: November 21, 2017 at 10:24 AM (#5579304)
In case people haven't paid attention to other threads, the good Doctor is running new MLEs for Negro League players. Of importance for this election are the ones for Ben Taylor and Hilton Smith, both of whom look like they deserve serious considerations for being on the ballot.


That is great to hear! That he's doing the work, not about the players specifically. Although I have always liked Ben Taylor as a candidate, and it will make me look at Smith again as well.
   243. JoeD has the Imperial March Stuck in His Head Posted: November 21, 2017 at 10:25 AM (#5579307)
Follow-up to #214. My book has been published. Player Won-Lost Records in Baseball: Measuring Performance in Context, published by McFarland, should be available wherever fine books are sold.


That is great, congrats!
   244. JoeD has the Imperial March Stuck in His Head Posted: November 21, 2017 at 10:31 AM (#5579311)
Adding to the 1880s pitchers probably being done commentary ... Am I wrong to assume that what was credited to pitchers in the 1880s was probably much more likely the contribution of the defense? As in pitchers had less impact on the outcome of a given AB than the hitters and fielders, relative to modern day baseball?
   245. JoeD has the Imperial March Stuck in His Head Posted: November 21, 2017 at 10:51 AM (#5579328)
Thanks for all of this, Rob! I really like the concept of a sliding replacement value. I definitely like the idea that in-season replacement level is lower than "offseason" (?) replacement level. I also like the concept of pennants added. I need to sit down with your formulas and see if I can figure out a way to make it work with my Player won-lost records. But very interesting food for thought.


Interesting system Rob ... thanks for sharing. Apologies for being late to the party.

I have heard the sliding replacement level argument for a long time. My problem with it, is that I think it assumes that these average league replacements over time are free. That you don't have to use your limited resources (dollars, draft picks, etc.) to add these players to your team. Having Don Kessigner there for 10 years as an above (annual, for the sake of argument) replacement level player, allows you to spend those resources to go find a better 2B, 3rd starting pitcher or left fielder. I don't see how sliding replacement level accounts for this.

I'd also note that in historical leagues I've played in where you can keep players forever (basically reserve clause type leagues), managers who take this approach, over-drafting (IMO) higher peak, shorter career players, especially in the middle rounds, after the stars are gone, tend to do pretty poorly. But admittedly, that could just be selective memory.
   246. Rob_Wood Posted: November 22, 2017 at 12:58 AM (#5579966)
Yes, player valuation inevitably relies upon the distribution and cost of alternative talent. The less costly and more available are alternatives with significant probability of eventually being better than "replacement" level (and even higher than league average), the more critical it is to have "high peak" players on your roster rather than "long career" players.

My new CPASR stat is based upon my attempt at analyzing the empirical flow of talent into the major leagues over time. It is apparent that the players that this system "values" is more than occasionally different than the players that I have been including on my HOM ballots.

I hope to integrate the CPASR figures (information) into my HOM ballots going forward.

   247. Carl Goetz Posted: November 22, 2017 at 10:17 AM (#5580036)
Rob, I just read through your earlier posts from April regarding CPASR as well as your more recent posts. I tend to agree with a preference for higher peak players over longer career guys. My questions in this process are in regards to the sliding replacement level. I understand what your are saying regarding teams fielding an average player at a given position over a 20 year period, but am wondering whether that tells the whole story. I'm going to make a few statements/questions and I'd be interested in your reactions/answers. This may be that I haven't put as much thought into this as you, so I'm making these statements with full respect to the amount of time and energy you've clearly put into this project.
1) Does this ignore individual circumstances? Theo Epstein is pretty good at drafting from what I can tell. It makes sense for him to not tie down a 'proven veteran' to a long term contract where a worse drafter might be better off with the veteran.
2) Most teams have a backup plan for even their stars in the minors. Its not like the Braves didn't draft 3B replacements during Chipper's career. I remember at one point, he was blocking Wes Helms. So the Braves adjusted and moved Chipper to LF. This had value to the Braves since they needed a LF and most of Chipper's value was offensive anyway. Of course Wes Helms turned out to be, well, Wes Helms (no offense Wes Helms, I would love to have had your major league career and the paychecks that went with it), so Chipper moved back.
3) Even if the veteran is truly blocking a good replacement, most teams don't waste that resource. They trade him for a useful piece right now or a comparable prospect at a different position. If we're saying the average value of the 'blocked' player over time will approach an average player, aren't we assuming the team will derive no value from that player? ie they rot in the minors while the veteran plays in the majors.
4) Realistically speaking, on average, the average 'blocked' player should yield in a trade an average player at a different position. Few teams (even great ones) have better than average players at every position. Therefore, acquiring an average player at a position of need is helpful to most teams. Shouldn't this mitigate some of the blockage?
5) I feel that as CPASR is currently constructed, there is an assumption that teams can't trade value for value and therefore utilize a 'blocked' resource. At the very least, it assumes we are looking at that player's position on the team in a vacuum without considering the other positions on that team.

I think my initial reaction is that I'm with you on the idea of a sliding replacement level over time. It should definitely move higher than the traditional AAA replacement player. But, for the reasons above, I question whether it should get all the way to a league average player.

I'm definitely very interested to hear you thoughts on anything I've written or assumed here.

Thanks for all of your work here as well. Even if I don't agree with you 100%, you have made me think about replacement level in a different way than I have in the past. I certainly appreciate that.
   248. JoeD has the Imperial March Stuck in His Head Posted: November 22, 2017 at 10:49 AM (#5580069)
Rob, I don’t really think that answers the question I asked. It seems like this ignores the fact that when you have a long career player 1) there is value in not having to expend resources to replace that player, and 2) there is value in having one less position to worry about.

The theory that replacement level becomes average over time and specifically CPASR says that if you have a Mickey Mantle I who plays from 1951-1959; he dies and is replaced by Mickey Mantle II from 1960-1968 - somehow the two players are more valuable than the real Mickey Mantle. Despite the fact that in the hypothetical, you have to expend the resources to sign, draft and develop two Mickey Mantles. If Mickey Mantle I had lived you could have traded Mickey Mantle II for Ron Santo.

Am I missing something? I have never thought this theory held any water, I still remain unconvinced.
   249. JoeD has the Imperial March Stuck in His Head Posted: November 22, 2017 at 10:51 AM (#5580072)
I mean teams don’t just field average players over 20 years. They field average players, at every position on average every year.
   250. Mike Emeigh Posted: November 22, 2017 at 11:52 AM (#5580164)
Keep in mind that Retrosheet fielding data is incomplete prior to 1989, and very incomplete prior to 1973 as many of the sources don't have fielding credits on hits and (before 1973) on many outs.

-- MWE
   251. Carl Goetz Posted: November 26, 2017 at 09:35 PM (#5581267)
Ok, I've finally got a preliminary ballot together.
I use Baseball Gauge custom with Baseruns for offensive WAR and DRA as defensive WAR up through 2002 and DRS from 2003 on. I use the average of runs based and FIP based WAR for pitchers. I look at prime (best 8 consecutive years) as my starting point, then productive career (career minus negative WAR seasons at start and finish of career) and peak (best 3 years not necessarily consecutive) bump players up and down the list. I also give white MLB players from pre-integration a knock down for not playing the best black players of their day. I don't give the black players of that era the same knock down because I feel their numbers are already conservative.

Prime 8 Career Best 3
Bell, Buddy 47.1 71.0 21.8
White, Roy 50.9 64.5 24.2
Jones, Andruw 50.2 67.2 23
Rolen, Scott 44.5 69.9 21.7
Redding, Dick 43.0 55.0 24.6
Jones, Chipper 37.4 70.5 19.1
Munson, Thurman 42.6 48.9 18.6
Leach, Tommy 51.6 77 24.7
Thome, Jim 42.1 73.1 19.6
Chance, Frank 48.1 57.0 24.5
Cicotte, Eddie 45.3 56.8 24.0
Dandridge, Ray 41.9 70.2 18.2
Bonds, Bobby 46.8 64.2 20.7
Bancroft, Dave 46.0 62.1 22.7
Santana, Johan 42.6 46.6 21.8
Shocker, Urban 45.5 54.2 21.0
Appier, Kevin 44.0 53.2 20.0
Rice, Sam 41.8 66.6 19.4
Smith, Hilton 41.6 67.9 19.8
Taylor, Ben 39.4 72.7 17.1
Gooden, Dwight 40.4 52.0 22.2
Campaneris, Bert41.7 63.3 20.4
Newcombe, Don 38.1 53.2 18.5
Schang, Wally 32 52.6 14

Some notes.
1) I feel Top 12 are all HoMers so I'm not down on any of these players. I understand I'm lower on Chipper than many here and I agree that he does particularly poorly under my way of doing things. I'm considering moving him up a few spots for the final ballot, but am also high on the players above him so not sure where that will ultimately leave him. Also not sure it matters since I suspect he will be elected this year without me.
2) After my top 5 outfielders, I have a huge blob of very very good OFs that I am having trouble separating with my system. These include Vlad, Sammy, and Lofton, as well as: Jose Cruz, Hugh Duffy, Bob Johnson(PCL Credit), Kiki Cuyler, Bernie Williams, Tony Oliva, Gavvy Cravath(PCL Credit), Mike Tiernan, Jimmy Ryan, George Van Haltren and Dale Murphy. I am currently considering moving Sosa ahead of Bonds and Rice on this list due to his better 3 year peak. Not seeing the Lofton and Vlad as being significantly better than others in the top 30-40 overall players.
3) Tiant is a career guy and that isn't my first choice for separating players. As career guys, I like Tommy John and Andy Cooper better and they aren't super close to cracking my top 15 either.
4) Jeff Kent is in the next group of 2B/3B/SS after the above listed. Also in that group: Evers, Cey, Lazzeri, Rizzuto, Pesky, Stephens, Bando, Concepcion, Garciaparra, Clarkson. I'm a fan of all, but feel all are just on the outside looking in as far as the HoM is concerned.
5) Vic Willis is definitely a top 25 pitcher, but like OFs, there is a large blob of pitchers I'm having trouble differentiating. I like Gooden, Santana, Lolich, Cooper, and Ol Diz better right now.

Definitely interested in feedback. There is a lot that is very close here and a small change in my line of thinking could change this ballot significantly.
   252. DL from MN Posted: November 27, 2017 at 10:26 AM (#5581375)
Carl, is that your actual order or do I have to re-order according to points in a different column? I don't get Chipper behind Rolen and Buddy Bell.
   253. Carl Goetz Posted: November 27, 2017 at 10:59 AM (#5581408)
That is the order. Bell reached 47.1 WAR in his 8 year prime and Chipper only 37.4 which is the primary area of difference. Their career totals were similar and Bell's 3 best were slightly better than Chipper's 3 best. Chipper is actually already moved higher on this list than the above numbers suggest due mostly to his postseason performance and due to his not fitting neatly into the somewhat arbitrary nature of my Peak/Prime measures. That said, I don't love my measures mostly due to that arbitrary nature. In short, I'm a peak/prime voter, but I don't like my methodology in determining those measures. I'm trying to be as objective as possible in my ranking, but certain things about my list above (including Chipper's placement) don't "feel" right to me. I feel like as a peak/prime guy, I'm left choosing between a total WAR from an arbitrary number of a players best years or a number of years in which a player exceeded and arbitrary WAR threshold. I'm currently looking at using a Pennants Added calculation and working on a project to select my own All-Star teams for each season (I doubt this will be completed in time to be useful for the 2018 ballot). If I'm understanding PA correctly, it seems like a good way to measure peak while avoiding any arbitrary decisions. As for my AS teams, I guess I'll be selecting an arbitrary number of top players each season, but it still seems like a worthwhile project. Granted, if I (for example) select the top 3 2B each year, there's always going to be a worthwhile candidate who was 4th best a bunch of times and this stat would therefore miss. Does anyone have other ideas for measuring peak/prime in a more objective, less arbitrary manner?
   254. Rob_Wood Posted: November 27, 2017 at 01:45 PM (#5581511)
Here are the results of the tens of millions of simulations (PA stands for Pennants Added):
WAR PA
0 .0000
1 .0142
2 .0296
3 .0465
4 .0630
5 .0809
6 .1002
7 .1197
8 .1398
9 .1615
10 .1829
11 .2055
12 .2292
13 .2530
14 .2758
15 .3011

A few observations are in order. First, note that the relationship between Pennants Added and WAR is non-linear. Doubling WAR more than doubles Pennants Added, especially as the WAR gets larger. Second, I replicated the simulations with different number of games in the season and have found that the relationship is scalable. For example, the Pennants Added by a 2-WAR player in a 81 game season is virtually identical to the Pennants Added by a 4-WAR player in a 162 game season.

To capture the non-linearity, I posited that there was a "kink" in the relationship and that the difference in a league-average performance and a league-replacement level performance was 2.5 wins. I find that the following equation is the best fit for the Pennants Added relationship under those assumptions:

PA = (.0144)*F*[WAR-(WAA-1)] + (.0144)*ONEWAA*[F*(WAA-1)]^1.1310

where F = 162/L is the scale factor and reflects the length of the season (L), and ONEWAA is an indicator function whether or not WAA exceeds 1. It actually is a much easier formula than it looks. Essentially it mirrors the numbers shown in the WAR-PA relationship enumerated above.


Above is from my post #180 where I give two simple ways to calculate Pennants Added figures.

Other feedback:

(1) Some of Carl's figures look strange to me. For example, take a look at Tommy Leach. Those figures make him a superstar. BB-Ref WAR figures are nowhere near those figures.

(2) Reliance on consecutive season performance has always been very very strange to me. It doesn't make any sense (to me) to believe that between two players with similar seasonal WAR figures, the one with the highest consecutive WAR total over 3 years, say, would be significantly more "valuable" than the other. Maybe I just have a blind spot on this issue.

(3) In any event, calculating something like Pennants Added should largely obviate the need to calculate "prime", "peak", "consecutive peak", etc.
   255. Carl Goetz Posted: November 27, 2017 at 02:08 PM (#5581540)
1) I use Baseball Gauge's custom metric. Leach does very well under DRA for defense. I don't completely trust that which is why he's lower than those numbers would suggest and I'm looking at even lower.
2) The thought is that a consistently good player year in and year out is easier to plan the rest of the team around. My main issue is that the number of years I chose is completely arbitrary. 8 years hurts a guy like Chipper. He would look great if I chose 10 or 12. Someone else might look bad by that measure. I am going to work on adding PA to my methodology.
3) I am coming around to agreeing with you on this.

Thanks for posting this formula. I appreciate that.
   256. Carl Goetz Posted: November 27, 2017 at 02:46 PM (#5581567)
WAR PA
0 .0000
1 .0142
2 .0296
3 .0465
4 .0630
5 .0809
6 .1002
7 .1197
8 .1398
9 .1615
10 .1829
11 .2055
12 .2292
13 .2530
14 .2758
15 .3011

Quick question. Does this WAR to PA chart assume 162 game season?
   257. Carl Goetz Posted: November 27, 2017 at 03:31 PM (#5581602)
Great article on Leach from the Hall of Miller and Eric website.
https://homemlb.wordpress.com/2013/12/27/why-we-elected-tommy-leach/

I will probably rerun the numbers Leach on a similar 2/3 DRA 1/3 TZ basis for defense. After that, I'll run PA for my player set and hopefully get an updated ballot out by weekend.
   258. Rob_Wood Posted: November 27, 2017 at 03:58 PM (#5581634)
The WAR to PA chart technically is based upon 161 game season, but clearly the difference between 161 games and 162 games is insignificant.
   259. Carl Goetz Posted: November 28, 2017 at 10:03 AM (#5581899)
Did a few players' PA last night and did notice that a player with higher peak but a couple years shorter career came in a little behind 2 lower peak longer career guys. Is there any statistical validity to taking a PA average (I'd probably do PA/20 Seasons to make the numbers more easy to work with) to highlight who got there with peak and who got there with career?
   260. Rob_Wood Posted: November 29, 2017 at 02:49 PM (#5582828)
I performed an old-style Bill James paired-player study to look into how much additional "pennant-added value" a player with a higher total WAR in his best consecutive three-season stretch provides his team over a similar player with a lower best consecutive three-season stretch.

In putting together my new-fangled CPASR stat, I created a database of over 1,000 players. The database includes the player's season WAR figures over his entire career. I will attempt to utilize this database to answer the question posed above.

My study uses the following data:
- hitters only (approx 600 in database)
- seasonal WAR figures adjusted for season length (WAR scaled up to 162-game seasons)
- top three adjusted seasonal WAR figures for each player

I then searched for pairs of players for the study having the following characteristics:
- one player has his best three adjusted seasonal WAR figures consecutively, other player has his best three adjusted seasonal WAR figures non-consecutively
- each of the top three adjusted seasonal WAR figures are within 0.3 across the two players
- sum of the top three adjusted seasonal WAR figures are within 0.5 across the two players
- sum of the players' best consecutive three adjusted seasonal WAR figures are at least 2.0 apart across the two players

I have identified 67 such player pairs.

I then tracked for each player pair the following:
- team win total in each of the three seasons in question (scaled up to 162 game season length)
- team place finish in regular season standings (scaled down to 8-team league if necessary)

Here are the results:

- players with best three seasons being consecutive average a total of 18.2 WAR in their best three seasons vs. players with best three seasons being non-consecutive average a total of 18.2 WAR in their best three seasons (this similarity was baked into the study)

- players with best three seasons being consecutive average a total of 18.2 WAR in their best three consecutive seasons vs. players with best three seasons being non-consecutive average a total of 15.5 WAR in their best three consecutive seasons (this disparity was baked into the study)

- teams with players with best three seasons being consecutive average 88.9 wins per season vs. teams with players with best three seasons being non-consecutive average 87.3 wins per season

- teams with players with best three seasons being consecutive had an average place finish of 3.0 (third-place) vs. teams with players with best three seasons being non-consecutive had an average place finish of 3.3 (between third-place and fourth-place)

- teams with players with best three seasons being consecutive had more team wins in 36 of the 67 player pairs

- teams with players with best three seasons being consecutive had higher team place finish in 37 of the 67 player pairs

- teams with players with best three seasons being consecutive won the regular season pennant in 48 seasons (of the 67*3=201 possible seasons) vs. teams with players with best three seasons being non-consecutive won the regular season pennant in 42 seasons (of the 67*3=201 possible seasons)

I will hold off on drawing any sweeping conclusions based upon a 67-sample study, but the study does suggest that empirically consecutive seasons may provide more "pennant-added" than identical seasons that are not consecutive.

Comments welcome.










   261. Dr. Chaleeko Posted: December 03, 2017 at 01:48 PM (#5585129)
I've updated my position counts for the HOM. It now includes all appearances data from the NLDB, as well as some from select players' minor leagues time (Campy, Jackie, Irvin, Grant, Minoso, and Trouppe). I've seen a couple different ways that people prefer to assess position, so here's a few ways we could look at it. The totals below are all rounded.

PRIMARY = The position the guy played the most (i.e.: Ernie Banks is at 1B)
SUM ALL = The sum across all HOMers at every position of the percentage of appearances (i.e.: Ernie Banks counts 51% toward 1B, 45% toward SS, 3% toward 3B, and 1% toward LF)
SUM 33%: Similar to SUM ALL but only counting positions where a player made at least 33% of his appearances (i.e. Banks only counts 51% toward 1B, 45% toward SS, and everything else is ignored)
SUM 25%: Ditto but with a 25% threshold

          C   1B  2B  3B  SS  LF  CF  RF  DH  SP  RP
====================================================
PRIMARY   20  29  22  17  26  24  25  24   3  69   4 
SUM ALL   18  29  22  20  23  27  26  24   5  58  12
SUM 33%   17  23  20  15  21  20  21  18   2  57   4
SUM 25%   18  24  20  16  22  22  21  18   2  57   6
----------------------------------------------------
AVERAGE   18  26  21  17  23  23  23  21   3  60   6


Based on any of these means of reckoning position, we remain well short at catcher and third base, and mildly so at second base. If we were to elect Chipper, Rolen, and Thome as three of our four, which seems pretty likely, it would have a very positive effect at 3B, of course. Here's how it would play out by each of the four methods above:
-PRIMARY: Adds 2 3Bs, total of 19 3Bs
-SUM ALL: Adds 2.02, total of 22
-SUM 33%: Adds 1.82, total of 16 (rounding error, goes from 14.6 to 16.4)
-SUM 25%: Adds 1.82, total of 18

My personal system recommends looking at these candidates the most closely at these three positions, and, of course, YMMV:

CATCHERS
Thurman Munson
Wally Schang

THIRD BASE
Sal Bando
Buddy Bell
Bus Clarkson (if you see him as a 3B)
Chipper Jones
Tommy Leach
Carlos Moran
Scott Rolen

SECOND BASE
Jeff Kent
Tony Phillips
Marv Williams
   262. theorioleway Posted: December 03, 2017 at 02:33 PM (#5585151)
Hey Doc, does Ray Dandridge make your consideration set for 3B? I know you were skeptical of the numbers you came up for him, but it seems like he should still be in at least consideration for a ballot spot. Did your new way of coming up with defensive numbers hurt him?
   263. Dr. Chaleeko Posted: December 03, 2017 at 02:54 PM (#5585160)
Yes! I forgot to include him. Sorry about that oversight. On the other hand, I'm much less certain about his defensive value than anyone else on that list, and his case strongly depends upon it. I'm personally going to wait before I push his candidacy.
   264. Dr. Chaleeko Posted: December 03, 2017 at 03:47 PM (#5585175)
I've also now updated my figures for comparing eras to one another. What I do is for every season, I determine what percentage of players' careers fell into that season. Then, I compare that to a theoretical "ideally balanced" HOM in which ease season has representation in proportion to the number of team-seasons played that year vs the number of team-seasons in all of history. But I also back out some career percentages from this theoretically balanced figure from recent seasons to account for the fact that some HOMers are still playing now. I base that on active players who already have a very strong chance (IMO). Finally, I subtract the actual career percentages that we've elected from the theoretically balanced figure to determine whether a given season are over or under the theoretical balance. Once that's done, I gang together groups of seasons that have a representation trend. Here's what those look like.

1871-1875: -0.90 vs theoretical balance, 83% of theoretical balance
1876-1890: +0.98, 105%
1891-1908: +6.04, 123%
1909-1921: -2.64, 89%
1922-1942: +9.46, 122%
1943-1945: -2.39, 60%
1946-1960: +3.54, 114%
1961-1968: -0.61, 96%
1969-1990: -9.28, 84%
===includes allowances for not-yet-eligible players===
1991-1998: -1.59, 92%
1999-2008: -1.02, 92%
2009-2011: -1.28, 24%

What I see here is that 1969-1990 and 1909-1921 are the paces we should probably revisit most closely. (The last three will continue to add more players in the next few years, so no worries about them.) Here are players from 1969-1990 and 1909-1921 that my personal system sees as the strongest candidates.

1909-1921
C: Wally Schang
1B: Ben Taylor, Bill Pettus
3B: Carlos Moran, Tommy Leach
SS: Art Fletcher, Dave Bancroft, Joe Tinker
LF: Bobby Veach
RF: Harry Hooper, Sam Rice

1969-1990
C: Thurman Munson, Gene Tenace, Jim Sundberg
2B: Tony Phillips
3B: Buddy Bell, Sal Bando
CF: Willie Davis (sorta), Cesar Cedeno
LF: Roy White, Jose Cruz
RF: Bobby Bonds

Now, the smarty pantses among you know where I'm headed. Let's cross the three positions we're light on (C, 3B, 2B) with the two eras we're light on (1909-1921 and 1969-1990). Here's where they fall for me. My CHEWS+ system is a hybrid between JAWS and Adam Darowski's Hall Score. 100 is the borderline. But that's based on the 220 the Hall have elected, so there's room for interpretation. With 263 players, assuming a 70/30 pitching split, the HOM's borderline is the 23rd player at each position. Rankings include Negro Leaguers but not active players.

CATCHERS
Munson: 106.1 CHEWS+, ranked 18th at position
Schang: 102.3 CHEWS+, ranked 19th
Tenace: 95.6 CHEWS+, ranked 22nd
Sundberg: 94.7 CHEWS+, ranked 23rd

3B (includes Molitor and Edgar as 3Bs)
Bell: 113.7 CHEWS+, ranked 17th
Moran: 110.1 CHEWS+, ranked 21st
Leach: 108.7 CHEWS+, ranked 22nd
Bando: 97.5 CHEWS+, ranked 27th

2B
Phillips: 101.3 CHEWS+, ranked 23rd

So, to my mind, Bell, Munson, and Schang are our best bets if we are interested in balance by position and by chronology.
   265. Bleed the Freak Posted: December 03, 2017 at 06:32 PM (#5585273)
Moran: 110.1 CHEWS+, ranked 21st


Greatly appreciate the time era electees versus theoretical balance.

Do you have anywhere online posted with CHEWS+ scores for guys not elected to the HoME and Negro Leaguers?

Moran at 110.1 is a very impressive total, is he a candidate you are fairly confident in the latest MLEs?

Some guys I'm curious on CHEWS+ scores from the 1969-1990 time frame who are in my consideration set not mentioned in your post:
Bert Campaneris
Ron Cey
David Concepcion
George Foster
Toby Harrah
Dale Murphy
Darrell Porter
   266. Rob_Wood Posted: December 03, 2017 at 08:48 PM (#5585365)
While I appreciate the information posted just above, the HOM project is not presently seeking, nor ever was seeking, equal representation by either era or position.

The mandate baked into the rules and to which each voter must abide is to be fair to all eras and all positions. These are similar, but not identical, ideals.

Having said that, I join with those wise posters above in encouraging all voters to give another deep-dive look at the players listed above and elsewhere in this thread.


   267. Dr. Chaleeko Posted: December 04, 2017 at 08:02 AM (#5585463)
Is it be fair to all eras to prioritize Phil Rizzuto or Bob Johnson or Tommy Bridges ahead of someone from the two underpopulated eras I’ve identified?

Is it fair to prioritize backlog players at other positions when catcher and third have an obvious need? (Second base, which I mentioned, is light but not a big deal. But the other two are noticeably lacking.)

Do we, as an electorate, have priorities? Or is it every voter for themselves?
   268. DL from MN Posted: December 04, 2017 at 09:20 AM (#5585496)
Catcher and 3B have lower representation but I still think we're shorter on pitchers than any other position
   269. Rob_Wood Posted: December 04, 2017 at 11:04 AM (#5585609)
In case Dr.C's questions were not purely rhetorical, every voter is committed to voting for who they judge to be the best players on the ballot. Nothing more, nothing less.

We have had well over 100 elections over more than ten years. I don't think we have ever had "priorities" and I don't think we should start now.
   270. bachslunch Posted: December 04, 2017 at 02:08 PM (#5585809)
Anybody know when the thread for posting official ballots will go up?
   271. Carl Goetz Posted: December 04, 2017 at 04:53 PM (#5585991)
I'm doing a deeper dive into Chipper. There's clearly 2 issues that are causing me to underrate him (at least as compared to the consensus here).
1) Offensive WAR: I'm using BaseRuns on the Baseball Gauge custom metric. I'm getting 81.7 vs BBRef 87.4 for career offensive WAR. Almost 6 WAR seems like alot over his career (though only 1 WAR every 3 seasons. I was under the impression that the offensive component of the various WARs weren't that different.
2) Defensive WAR: I'm using DRA 2002 and prior and DRS 2003 and later. DRA is not kind to the early part of Chipper's career. Career DRA is -13.9 DefWAR and DRS/TZ is -2.4 DefWAR. My combination is -11.5 so DRA is even kinder at the end as well. Is there some reason to believe that DRA is unfair to Chipper from 1995-2002 (the period where I am using that metric)?

Even if I'm ultimately comfortable with my system, he's in a tight cloud at 3B that I'm currently ranking: Bell, Chipper, Rolen, Dandridge. I believe all are Hall-worthy, but I'd like to get my ballot right even if Chipper is a shoo-in without me this year. Even a small adjustment on defense for 95-02 moves Chipper to the front of the pack. Even splitting the difference between DRA and DRS/TZ those years would make him an easy #1 on my ballot.
   272. Dr. Chaleeko Posted: December 04, 2017 at 05:41 PM (#5586046)
Definitely not rhetorical. Here's the thing I'm seeing. The backlog is pretty long at this point. We been sorting through these guys for more than a real-time decade now. If they haven't been elected by now, then they are very close, one way or the other, to the borderline and to one another.

Which means that they are very likely within the margin of error that any system will have, no matter how well crafted it is. (Mine, yours, anyone's.)

Is the argument that "Well, Bob Goodbat from the 1930s is my system's preferred candidate" a sufficient answer to the question "Is it more important to elect qualified players from underrepresented positions and eras or to elect more firstbase/shortstop/outfielders from well represented or overrepresented eras?" Especially in light of the strong likelihood that the majority of the backlog are extremely close to one another in quality?

In light of this line of questioning and the diffuseness of the backlog, why shouldn't the electorate have priorities? Especially since the classes of 2020 and beyond aren't bursting with excellent candidates like the last several slates of newbies have been. After 2019, we're right back into the backlog, and these questions will take on renewed significance.

Hey, sorry if this is coming off kind of high-toned. Not trying to be a turkey. I just get excited around election time.... ;)
   273. Kiko Sakata Posted: December 04, 2017 at 06:25 PM (#5586078)
Is there some reason to believe that DRA is unfair to Chipper from 1995-2002 (the period where I am using that metric)?


Yes. This very old article asserts the problem with Chipper's fielding ratings without necessarily proving it (I remembered reading very strong arguments from Chris Dial that Chipper Jones was badly underrated as a fielder, so I Googled Chipper Jones fielding Chris Dial and this popped up). Basically, the argument is that the batted ball distribution of Braves' opponents was skewed heavily against ground balls toward third base. For example, Tom Glavine was a left-handed pitcher with a high ground-ball propensity, which you'd expect to translate to a lot of right-handed hitters pulling the ball on the ground toward 3B. But Glavine lived on the outside corner (or two inches off the outside corner if the umpire would give it to him), so right-handed hitters tended not to pull the ball against him.

My player won-lost records model wins and losses, so you can compare total decisions across players to see who seems to have faced more chances. Here's a comparison of the fielding records of Chipper Jones and Scott Rolen. Jones was in LF in 2002-03, so try to ignore those seasons. It's a little hard to compare some other seasons because Rolen tended to play fewer games per season. For his career, I have Jones with 111.7 fielding decisions at 3B in 17,102.2 innings and Scott Rolen with 122.0 fielding decisions in 17,479.1 innings at 3B. That works out to 0.98 decisions per 150 innings for Jones vs. 1.05 decisions per 150 innings for Rolen - or about 7% more decisions for Rolen.

Rolen was certainly the better fielder of the two but Jones rates basically average (-1.7 net fielding wins at 3B for his career) by my Player won-lost records.
   274. Kiko Sakata Posted: December 04, 2017 at 06:35 PM (#5586084)
Is the argument that "Well, Bob Goodbat from the 1930s is my system's preferred candidate" a sufficient answer to the question "Is it more important to elect qualified players from underrepresented positions and eras or to elect more firstbase/shortstop/outfielders from well represented or overrepresented eras?" Especially in light of the strong likelihood that the majority of the backlog are extremely close to one another in quality?


I think this is an issue that deserves some thought. To put some names on it, if I had been voting 60 or 70 "years" ago, I think I would have been a strong supporter of Bob Johnson and fairly meh on Earl Averill. In the present day, Vlad Guerrero looks similar in my system to Bob Johnson. If Johnson comes out slightly ahead of Vlad, does that mean I have to / should support Johnson in order to rectify what my system views as a past wrong? But one can only rectify it halfway - we could elect Johnson, but we're not going to unelect Averill. Or does it make sense for me to view Averill as essentially having taken Johnson's slot such that I should push for Vlad even if I think Bob Johnson was slightly better (presuming that I think both of them are HOM/ballot-worthy). I kind of do think that, at a minimum, our bias should be toward the newer candidate(s) who haven't had as many opportunities for election yet.

That said, I'm reasonably sure that Tommy John, Vern Stephens, Wally Schang, and probably Ben Taylor are going to be on my ballot this year. So, I'm not saying "if you retired before 1990, you had your chance; make way for the new kids" - which, as I understand it, would be a clear violation of the HOM's "fair to all eras" requirement.
   275. Rob_Wood Posted: December 04, 2017 at 06:43 PM (#5586088)
Definitely not rhetorical. Here's the thing I'm seeing. The backlog is pretty long at this point. We been sorting through these guys for more than a real-time decade now. If they haven't been elected by now, then they are very close, one way or the other, to the borderline and to one another.

Which means that they are very likely within the margin of error that any system will have, no matter how well crafted it is. (Mine, yours, anyone's.)

Is the argument that "Well, Bob Goodbat from the 1930s is my system's preferred candidate" a sufficient answer to the question "Is it more important to elect qualified players from underrepresented positions and eras or to elect more firstbase/shortstop/outfielders from well represented or overrepresented eras?" Especially in light of the strong likelihood that the majority of the backlog are extremely close to one another in quality?

In light of this line of questioning and the diffuseness of the backlog, why shouldn't the electorate have priorities? Especially since the classes of 2020 and beyond aren't bursting with excellent candidates like the last several slates of newbies have been. After 2019, we're right back into the backlog, and these questions will take on renewed significance.

Hey, sorry if this is coming off kind of high-toned. Not trying to be a turkey. I just get excited around election time.... ;)



I haven't read the Constitution in a while, but I am quite sure that submitting a ballot with Player X higher than Player Y despite the fact that the voter actually thinks that Player Y was a "better" player than Player X is against the rules. It would fall under the "strategic voting" umbrella which is expressly forbidden in Hall of Merit voting. We have all agreed to submit individual ballots which accurately reflect our personal judgments as to the "worthiness" of each of the candidates. We also agree to live with the results of the group voting.

As I have said more than once, and something I personally take to heart, voters should be encouraged to re-visit all new players and all players in our backlog. And if you say that voters should pay extra attention to players in "under-represented" eras or positions, I would agree. In fact, I encourage voters (myself included) to advocate for specific players that they feel may be receiving less voting support than they merit. That is part and parcel of this process and project.

But my view is that voters should not as a rule give "extra credit" to any players in "under-represented" eras or positions due simply to their "under-representativeness". I really don't think this point is at all debatable.
   276. Carl Goetz Posted: December 04, 2017 at 08:58 PM (#5586117)
"Basically, the argument is that the batted ball distribution of Braves' opponents was skewed heavily against ground balls toward third base. "
So, is TZ then accounting for his lack of chances at 3B while DRA is not?
   277. Kiko Sakata Posted: December 04, 2017 at 09:48 PM (#5586122)
So, is TZ then accounting for his lack of chances at 3B while DRA is not?


I think so. TZ is zone based, so it looks at Jones's plays made relative to plays in his zones, I believe. Whereas DRA is basically just based on Jones's raw putouts and assists and the overall balls-in-play for the Braves as a team. A zone-based analysis should notice if Jones simply isn't getting many balls hit his way, whereas DRA isn't necessarily going to catch such a thing (in theory, DRA would presumably be over-valuing Jones's teammates - most likely the middle infielders - who are getting more opportunities than expected).
   278. theorioleway Posted: December 04, 2017 at 09:58 PM (#5586125)
I think it's fair to argue that the HOM has had priorities throughout its history. Most of them I think are of the noble kind - making sure players from the Negro Leagues and the 1800s were fairly represented, expecting voters to behave in a rational and systematic manner, etc. But we also decided in 2012 that Phil Rizzuto shouldn't get elected after it was remembered that it was supposed to be an elect-4 year instead of an elect-3 year, in large part because people were more comfortable with more current players making the HOM instead of a borderline old candidate in Rizzuto.

And to echo the good Doctor's point, thinking through these things are important (not that I'm accusing anyone of not having done that). After all, who knows if Rizzuto will ever end up getting elected to the HOM.
   279. Carl Goetz Posted: December 04, 2017 at 10:01 PM (#5586126)
Actually, was leafing through my copy of Wizardry and Humphreys does address bad Chipper's DRA showing for about 2 pages early in the Third Base section. He references a study Tom Tango did comparing batted ball outs by Chipper vs other 3B for each pitcher (throughout those pitchers careers). Tango estimates that Chipper made 261 plays below expectation throughout his career (2008 was the time of the study) which could translate to close to minus 200 defensive runs. Humphreys then notes the study includes infield flies and other discretionary chances that should have been ignored. Humphreys' conclusion: "Given this mixed evidence, I'm currently inclined to think that Chipper was indeed well below average as a fielder, though perhaps somewhat better than the DRA estimate." Given that the creator of DRA believes that DRA overestimates how bad Chipper was at 3B, I'm inclined to give him some credit back. I'm inclined to be conservative about this, but it won't take much at all to move him to the top of the 3B list and probably the ballot.
   280. Devin has a deep burning passion for fuzzy socks Posted: December 05, 2017 at 10:40 AM (#5586307)
Just talked to Joe on Twitter, and he'll get the 2018 Ballot thread up at some point today. Might not be until the evening, though.
   281. Carl Goetz Posted: December 05, 2017 at 12:20 PM (#5586437)
Ok, I'm comfortable with Chipper at the top of my ballot. Now on to a few other players I seem to rate differently than the electorate.

Vlad Guerrero: Is this a Win shares vs WAR argument. He looks good with WS (from BBGauge) but the Custom WAR I use (which comes out similar to BBRefWAR and gWAR) has him around 15-19ish on my OF list. What is WS catching that WAR is missing?

Jeff Kent: He's somewhat down on my overall infielders list though is the top 2B (narrowly edging out Johnny Evers) on the list. Are people giving him a bump for being the best available at an underrepresented (slightly) position, or is there something WAR is missing here?

Roy White and Jose Cruz Sr.: Both rate higher on my list than the general electorate. I'm certain its because DRA rates both as significantly better defenders than TZ (particularly White). Humphreys addresses some park factor issues that may affect DRA for both, but suggests White's adjustments for his old Yankee Stadium years should be minimal. A) Do 2017 DRA WAR numbers already make park factor adjustments? B) Whether A is yes or no, are further adjustments warranted in either case?

Hilton Smith and Andy Cooper: Based on MLEs from Hall of Miller & Eric's website https://homemlb.wordpress.com/2017/10/18/evaluating-negro-leagues-pitcher-part-iii-rogan-smith-and-williams/ and https://homemlb.wordpress.com/2017/10/04/evaluating-negro-leagues-pitchers-part-i/, I have them ranked 1 & 2 on my Pitchers list. Any reason to believe these aren't conservative enough?

Luis Tiant: He moved up my list a little when I looked at Pennants Added but I still feel he's pure career. I get a career guy would (and should) have El Tiante at or near the top of the list, but I just can't get there even as a guy who loves rewatching the 1975 series just to watch him pitch.

Vic Willis: I don't have him that far down. He's 7th on my pitcher list. He's in a group with Cicotte, Ol Diz, and Kevin Appier that I am having trouble separating. Right now I have Dizzy ahead of him but considering flip flopping. I have Appier in front of the group on the tiebreaker of his having to play better competition (ie the best black players of his day). Cicotte and Willis are especially tough to separate in my mind. Any thoughts on that one in particular. The only other players in front of Willis are Smith, Cooper and Redding (from NegL).

Tommy Bridges: I have trouble separating him from Bucky Walters (though I do give Bridges the edge), but they are both roughly 20th on my Pitcher list. There are several pitchers with more career WAR and more Pennants Added than these 2. Looking at his WS (and other versions of WAR), I'm not seeing a huge case there either. Can a Bridges supporter (or more than 1) run down the arguments in his favor?

   282. JoeD has the Imperial March Stuck in His Head Posted: December 05, 2017 at 12:45 PM (#5586462)
Hey guys, I'm going to get the thread posted now. Sorry for the delay, I'm on site for customer work this week, and it just completely slipped my mind.
   283. shoewizard Posted: December 05, 2017 at 03:16 PM (#5586641)
I know a lot of people bake in some WPA or leverage stats so I thought I'd share this here. I just discovered something interesting, and sent the following to Fangraphs. If anyone here has any insight to this, please let me know

Hi
I was comparing high leverage numbers for a couple of players today between FG and Baseball Reference, and I noticed that there are always a lot more High Leverage PA total at BR compared to FG. So I went to the league splits and found the following

site High  Med     Low unknown ttl lvgPA ttl PA
BR
:: 33780 68813  82478 224     185295    185295
FG
:: 16802 66771  93501   0     177074    185295 


Not sure how that will format, but the gist is this:

The total PA in all of MLB last year was 185295, which both sites agree on. BR accounts for all 185295 in their leverage PA totals, (putting 224 into unknown, which balances the books)

However FG is missing 8221 PA in total in their league leverage splits, and clearly there is a pretty big difference in the way High/low Leverage is defined and why there is such a large discrepancy between the two sites on this.
   284. Kiko Sakata Posted: December 05, 2017 at 03:19 PM (#5586647)
Vlad Guerrero: Is this a Win shares vs WAR argument. He looks good with WS (from BBGauge) but the Custom WAR I use (which comes out similar to BBRefWAR and gWAR) has him around 15-19ish on my OF list. What is WS catching that WAR is missing?


I would guess it's the lower replacement level in Win Shares. Vlad was very durable in-season through his prime (9 seasons of 150 or more games and 640 or more plate appearances). I think it's a question of how much value you give him for just showing up. (Which, to be clear, I think is a perfectly valid question to ask for which one can come to a multitude of reasonable answers.)

On Kent, White, and Cruz, for me personally, my system weights fielding less than WAR does (correctly so, I believe), which bumps Kent up (his bad fielding hurts less) and the reverse for White and Cruz.
   285. shoewizard Posted: December 05, 2017 at 03:47 PM (#5586681)
Edit to number 283 above: I made a mistake and is read AB for PA in the medium leverage info, should have been 74992.

So FG not missing those PA...but still major discrepancy in the number of high levg PA between the two sites. Wouldn't that heavily impact how they calculate WPA ?

Edit

received feedback already from FG

We define high leverage as any play with a LI greater or equal to 2.0, low leverage is less than or equal to 0.8. I think baseball-reference has something slightly different. The idea is to keep high leverage plays around 10% of total plays. I guess we're at more around 9% this year.
   286. DL from MN Posted: December 05, 2017 at 05:10 PM (#5586785)
I went through and figured out when backlog players would have made my PHoM.

1927 - Gavy Cravath
1958 - Tommy Bridges
1967 - Phil Rizzuto, Bus Clarkson
1968 - Urban Shocker
1972 - Bucky Walters
1973 - Ben Taylor
1976 - Dave Bancroft
1986 - Bob Johnson
1987 - Hilton Smith, Wally Schang
1991 - Bert Campaneris, Luis Tiant
1995 - Tommy John
1997 - Norm Cash
2004 - Don Newcombe
2005 - Johnny Pesky
2008 - Dick Redding
2009 - Kevin Appier, Tommy Leach
   287. Bleed the Freak Posted: December 05, 2017 at 05:30 PM (#5586805)
Roy White and Jose Cruz Sr.: Both rate higher on my list than the general electorate. I'm certain its because DRA rates both as significantly better defenders than TZ (particularly White). Humphreys addresses some park factor issues that may affect DRA for both, but suggests White's adjustments for his old Yankee Stadium years should be minimal. A) Do 2017 DRA WAR numbers already make park factor adjustments? B) Whether A is yes or no, are further adjustments warranted in either case?


Per the discussions with Doc, DRA isn't reliable in measuring outfielder arm value, it's best to substitute Baseball Reference R-Of when available.
This takes away ~2 WAR from each player's case.

Hilton Smith and Andy Cooper: Based on MLEs from Hall of Miller & Eric's website https://homemlb.wordpress.com/2017/10/18/evaluating-negro-leagues-pitcher-part-iii-rogan-smith-and-williams/ and https://homemlb.wordpress.com/2017/10/04/evaluating-negro-leagues-pitchers-part-i/, I have them ranked 1 & 2 on my Pitchers list. Any reason to believe these aren't conservative enough?


The Hilton Smith is confirmation of the MLE's Alex King did in the past, his support at Baseball-Fever, his general reputation from the literature, and his election to the HOF.

For Andy Cooper, Doc notes that "Cooper might look better in this MLE than he does on the Negro Leagues database. Part of the reason could be that our protocol substitutes the pitcher’s MLE-career average performance in instances where we have no data available to us. Much of the heart of Cooper’s career is missing, so he either looks worse on the Negro Leagues Database than he ought to, or he looks better in our MLE. We won’t know until we know. So the combination of lopping off a couple of lesser end-of-career seasons, giving him MLE-career average performance in missing seasons, and giving him more innings in his salad days makes him come forward a bit." The current MLEs give him bubble status for me.

Tommy Bridges: I have trouble separating him from Bucky Walters (though I do give Bridges the edge), but they are both roughly 20th on my Pitcher list. There are several pitchers with more career WAR and more Pennants Added than these 2. Looking at his WS (and other versions of WAR), I'm not seeing a huge case there either. Can a Bridges supporter (or more than 1) run down the arguments in his favor?


Bridges ranges from worthy by Baseball Gauge and Joe D's classic PA, somewhat close but short by Baseball Reference and Fangraphs, and well short by Kiko's W-L.
   288. Dr. Chaleeko Posted: December 05, 2017 at 07:02 PM (#5586866)
Yeah, so I wouldn't yet trust the Andy Cooper MLE. It's a work in progress. Once Gary A. and KJ and the gang have more of his career in the NLDB, I'll look more closely. For now, eh. Same for Hilton Smith in some regard as well. We're missing the latter half of the 1950s, so I'm loathe to put too much stock in his numbers. I would feel differently about someone like Monte Irvin for whom we have MLB and MiL stats that we can work with. But in the absence of information, I get a little antsy.
   289. Kiko Sakata Posted: December 05, 2017 at 08:33 PM (#5586909)
Public Service Announcement for anybody who uses by Player won-lost records.(*) Retrosheet did their latest release the Sunday after Thanksgiving. I am in the process of updating my Player won-lost records to incorporate this latest release and hope to have everything updated this weekend. I'm going to wait until next week to vote to see if anything noteworthy pops up. Of particular interest, Retrosheet now has complete games (including deduced games) from 1941 - 2017 (side note: with the unfortunate passing of Bobby Doerr recently, the release of 1941 deduced games means that Retrosheet has play-by-play for every game played by every living current or former major-leaguer) and partial games for every season from 1925 - 2017 - note: this increases the set of consecutive seasons by 5 years.

Looking ahead to next year, Dave Smith tells me that he hopes to release consecutive seasons back to 1921 this summer - possibly even 1920 - and we should have 1940 deduced by the summer release as well. There are a number of players whose candidacies may well be impacted by what could essentially be an additional decade of data by later this year.

I apologize for updating my data in the middle of the ballot period. Just bad timing with Retrosheet's data release and Hall-of-Merit voting right on top of each other.

(*) - and for those of you not using my Player won-lost records - why the hell not!? They're awesome!
   290. Bleed the Freak Posted: December 05, 2017 at 08:41 PM (#5586917)
Thanks Kiko, I will hold off on voting until you have the updates ready.
The result could impact the 4th electee, Chipper Jones, Jim Thome, and Scott Rolen are likely to be 1st ballot inductees, with Johan Santana and Andruw Jones falling in with the strong hold overs.

From post 221, I had a note about Wilbur Wood, I was wondering if you could help illuminate why your system is so low on him compared to some others.
   291. Rob_Wood Posted: December 05, 2017 at 09:26 PM (#5586931)
Kiko, Just wanted to extend my heartfelt gratitude to you and all the other Retrosheet people for all you do in making complete and accurate data available. Thanks much.
   292. DL from MN Posted: December 05, 2017 at 10:00 PM (#5586952)
Can a Bridges supporter (or more than 1) run down the arguments in his favor?


Sure. First, you have to give him WWII credit to make him a candidate but he deserves two full seasons at a high level (8+ WAR). Second, you have to value performance above average. Bridges pitched fewer innings than his contemporaries but at a better rate than almost everyone.
   293. Kiko Sakata Posted: December 05, 2017 at 10:01 PM (#5586953)
Oh, yeah, I remember you asking about Wilbur Wood. Sorry I didn't answer that earlier.

I've found that my system tends to be a bit lower on knuckleball pitchers than BB-Ref. I think it's because BB-Ref isn't dinging knuckleballers enough for their contribution to unearned runs. As I understand BB-Ref's systems, passed balls and perhaps even wild pitches are charged entirely to catchers in their defensive ratings. Pitchers' WAR are then adjusted for their team's defense. So, basically, the passed balls that knuckleball pitchers allow are blamed on their catchers. I believe that BB-Ref also has too wide a range on its fielding numbers, which over-corrects for pitchers with extreme defenses behind them - this, of course, is not unique to knuckleballers, but I think does affect Wood, who pitched in front of some fairly sketchy defenses in Chicago.

If you go to Wood's Baseball-Reference page, his "RA9def" - their defense adjustment - are consistently large and negative (meaning he pitched behind below-average defenses, if I understand BB-Ref's definition(s)) through his big seasons.

If you look at Wood on Fangraphs, he scores much lower than on BB-Ref (52.2 career bWAR). Fangraphs gives him 36.9 fWAR based on FIP and 37.9 based on RA-9. Converting my numbers to something on the same scale as WAR (see here), I get a value for Wood of about 31.9 (using eWins - Wood looks a tick worse in pWins, although that's not a big part of the explanation here). So, I'm not THAT far from Fangraphs (about 5 - 6 wins). Some of that might be batting - Wood was a lousy hitter (-1 WOPA over his career - the introduction of the DH probably saved him another win).

To be honest, I've started to be increasingly skeptical of BB-Ref's pitcher WAR. Their adjustments are very inscrutable and in many cases do a LOT of the work (compare Tommy John on BB-Ref vs. Fangraphs, for example - and keep in mind, Fangraphs shows two numbers, one based on FIP but also one based on RA-9).

I'm not really sure what else to say here. Looking at his career, Wood was good to great for about four seasons - over which he pitched about six seasons worth of innings - which is pretty thin for a peak when that's all there really is.
   294. Mike Webber Posted: December 05, 2017 at 10:53 PM (#5586975)
Can a Bridges supporter (or more than 1) run down the arguments in his favor?


Sure. First, you have to give him WWII credit to make him a candidate but he deserves two full seasons at a high level (8+ WAR). Second, you have to value performance above average. Bridges pitched fewer innings than his contemporaries but at a better rate than almost everyone.


Another interpretation might be he was a very good pitcher who at ages 35 and 36 had really strong seasons in 1942 and 1943 before his career came to an end due to WW2.

The 6 pitchers ahead of him In BB Ref WAR that pitched at the same time as he did are in the HOM - Grove, Hubbell, Lyons, Feller, Newhouser, Ruffing. All those guys had 10 more WAR than Bridges except Ruffing. Ruffing had 2 more WAR than Bridges, was one year older, and left MLB after the 1942 season, but came back in 1945 and pitched effectively but sparsely for 3 years.

Both Ruffing and Bridges were effective World Series pitchers.

The guys just behind him are mostly out - Bobo Newsome, Wes Farrell, Mel Harder, Dutch Leonard, Bucky Walters, Dizzy Trout, Larry French, Lefty Gomez, Claude Passeau, Dizzy Dean. Those guys are all within 10 career WAR of Bridges, the ones that are in had the Big Season/Peak arguments.

Basically you have to decide what kind of WW2 credit he deserves, and if he belongs with that first group with Ruffing, or the second group with Passeau and Trout and French.
   295. Carl Goetz Posted: December 05, 2017 at 11:19 PM (#5586997)
"I would guess it's the lower replacement level in Win Shares."
Thanks for the input. I think ultimately, I need a little more peak from him.

"Per the discussions with Doc, DRA isn't reliable in measuring outfielder arm value, it's best to substitute Baseball Reference R-Of when available.
This takes away ~2 WAR from each player's case."
Ok. Cruz wasn't in my top 15 anyway. I'll have to think about how this affects White.

"Yeah, so I wouldn't yet trust the Andy Cooper MLE. It's a work in progress."
" Same for Hilton Smith in some regard as well."
Ok, I'll scale them back for now. I'd rather wait and get them right.

"Sure. First, you have to give him WWII credit to make him a candidate but he deserves two full seasons at a high level (8+ WAR)"
This seems aggressive to me given that he was 37 in 1944 and 38 in 1945 and 1943 was a step up for him over 1941 & 1942. I know 1944 & 1945 were the worst years in terms of league quality in MLB, but where does 1943 fall? It had to be lower quality than 1942, right? My thought was to bump him back in 1943 to his 1941/1942 performance level (prorated for his extra innings in 43) and then give him 1944 & 1945 credit at that level as well. Not sure I'm buying Minor league credit in the post war years though.

   296. Carl Goetz Posted: December 05, 2017 at 11:23 PM (#5587001)
Thanks for your input as well Mike. You posted while I was working on my reply to the others. I am considering giving Bridges a little extra bump for his WS play. He was effective in 3 of the 4 WS and was best in the one that they won.
"Basically you have to decide what kind of WW2 credit he deserves, and if he belongs with that first group with Ruffing, or the second group with Passeau and Trout and French."
I'm feeling like he's right between. I also like Dean, but I'm struggling with that short career.
   297. bachslunch Posted: December 06, 2017 at 07:08 AM (#5587033)
Is there any reason why I can post to this thread but not to the ballot thread? Any help that can be provided will be greatly appreciated.

Edit: seems to be working now. Still not sure what happened, though.
   298. Dr. Chaleeko Posted: December 06, 2017 at 09:27 AM (#5587088)
@297

Yeah, there's an easy reason: BTF. ;)

IOW, happens all the time, at least to me. Similar to how the Remember Me tick box on the login page doesn't actually remember me. Or how the search function on BTF doesn't usually return very useful results. It's all part of the charm. Actually, I kinda do find it charming in a wistful 2005 kind of way.
   299. DL from MN Posted: December 06, 2017 at 12:00 PM (#5587247)
I am pretty generous with WWII credit for Bridges because he played in AAA for four seasons (two of them rather well) after he was done in the majors and he pitched during the war. His bio is pretty interesting. Small guy described as 150 pounds of pure guts. Fantastic curveball.
   300. bachslunch Posted: December 06, 2017 at 12:54 PM (#5587327)
Flip
Page 3 of 5 pages  < 1 2 3 4 5 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
JPWF13
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.4763 seconds
42 querie(s) executed