Tuesday, December 31, 2013

Both Trammell & Whitaker are in top 10 in career WAR at their positions

Top 10 in career WAR among guys who played 50%+ games at 2B (from BBRef). Seems like they both should be in the Hall of Fame

Rogers Hornsby 126.9
Eddie Collins 123.9
Nap Lajoie 107.2
Joe Morgan 100.3
Charlie Gehringer 80.8
Lou Whitaker 74.8
Bobby Grich 71
Frankie Frisch 70.1
Ryne Sandberg 67.7
Roberto Alomar 66.7
Now the same thing for SS

Honus Wagner 130.4
Cal Ripken 95.5
George Davis 84.9
Robin Yount 77
Ozzie Smith 76.5
Bill Dahlen 75.3
Luke Appling 74.4
Arky Vaughan 73
Derek Jeter 71.5
Alan Trammell 70.3

Highest Career WAR For Players Who Never Got Any MVP Votes

Chet Lemon never got any MVP votes despite 4 top 10 finishes in WAR including 1984 on a 1st place team. Dan Hirsch compiled these numbers. Here are the leaders among position players:

Chet Lemon 55.3
Jason Kendall 41.5
Ron Fairly 35.2
Frank White 34.7
Jeff Cirillo 34.4
Ray Durham 33.7
Mark Ellis 33.7
John Valentin 31.6 (Update Jan. 1-John Valentin had 0.15 MVP Shares-it was supposed to be Jose Valentin, probably my fault)

Pitcher Chuck Finley had 58.4 with no MVP votes.

This list may not be a big surprise. 3 are 2Bmen (White, Durham, and Ellis) and a catcher, Kendall, are here. Previous research shows these positions don't do well in MVP voting. See MVP Awards And Award Shares By Position

Saturday, December 28, 2013

Hall Of Fame Voting And Race

This is based on an earlier post called My Predictions For The Hall Of Fame Vote. You need to go there to understand the variables and the model.

Here I simply added a single dummy variable, Nonwhite (1 for non-whites, 0 for whites). This is a very simple way to do it. It does not take into account the slopes of the other variables regarding race. And it just lumps blacks and Hispanics into one group.

Anyway, the results were about the same in terms of r-squared and the coefficient values. Below are the new coefficient values. Notice that Nonwhite, although negative, has an extremely small value (it also was not statistically significant with a t-value of -.125). It means that being non-white would lower your predicted vote by .136 percentage points. So where a white guy would get 50.136%, a non-white would get 50%.



Variable            b Coeff.  
-------------------------------
    Constant        -0.01223  
   WSAS/1000         0.00112  
   GGAS/1000         0.06243  
         MVP         0.04821  
    3000 HIT         0.45176  
      500 HR         0.16757  
      ASSQ10         0.00216  
       GGSQ7        -0.00122  
       500SB         0.04949  
WSIMPSQ50/10        -0.01188  
     10000PA         0.09924  
    Nonwhite        -0.00136  

Thursday, December 26, 2013

What is the highest WAR ever for a guy who got no votes at all in the MVP voting, maybe since 1931?

I posted this question on Twitter and Dan Hirsch came up with a short list for position players (some pitchers had higher WAR). Here are the four he gave. I also have  each guy's rank in WAR, their team's record, position in standings and how many games out they finished:

W. Davis '64 8.4 (5) (80-82, 6T, 13)
Clemente '68 8.1 (2) (80-82, 6, 17)
E. Mathews '63 8  (7) (84-78, 6, 17)
Bonds '89 8 (3) (74-88, 5 in 6 team division, 19)

These guys were all well known players so it is not like the writers did not know who they were. I know their teams were not even close to being contenders, but no votes at all?

Dan wondered about the RBI totals for these guys.

W. Davis '64 77
Clemente '68 57
E. Mathews '63  84
Bonds '89 58

Friday, December 20, 2013

Mike Mussina And The Hall Of Fame

Jen Mac Ramos has a good article in support of him at "Beyond the Boxscore." See Mike Mussina and the Hall of Fame argument.

One thing Mussina had going for him was a great strikeout-to-walk ratio. He had 3.58, while the league average was 1.80. So his relative score, according to Lee Sinins Complete Baseball Encyclopedia, was 199 (3.58/1.80 times 100). That was 9th best all-time for pitchers with 3000+ IP through 2012 (130 pitchers in that group).

Mussina had pretty good peak value. He had 7 top 5 finishes in WAR for pitchers at Baseball Reference. Click here to go to his page. 

Now maybe pitchers should be judged on what they largely control: HRs, strikeouts and walks. So I created a crude fielding independent ERA relative to the league average for all pitchers with 3000+ IP. It was based on a regression with a pitcher's relative ERA being the dependent variable and HRs, strikeouts and walks (all relative to the league average) being the independent variable.

The equation was

ERA = 47.38 + .167*BB + .253*SO + .17*HR

So, taking Mussina for example, his relative HRs was 113 (he gave up 13% fewer HRs than average), relative BBs was 174 and strikeouts 114 (struck out 14% more than average). Plugging those numbers into the equation we get 124.49 (almost exactly his actual relative ERA of 124). So he was about 24% better than average. Curt Schilling & Kevin Brown both do even better here and on the next list.

Here is the top 20


Rank
Pitcher
Pred Rel ERA
1
Lefty Grove
133.96
2
Walter Johnson
133.59
3
Cy Young
129.70
4
Roger Clemens
129.36
5
Tim Keefe
128.91
6
Amos Rusie
128.82
7
Nolan Ryan
127.90
8
Greg Maddux
127.55
9
Curt Schilling
127.18
10
Randy Johnson
127.05
11
Jim Whitney
126.83
12
Kevin Brown
126.82
13
Christy Mathewson
126.57
14
Eddie Plank
126.25
15
Eppa Rixey
124.69
16
Bob Feller
124.64
17
Mike Mussina
124.49
18
Paul Derringer
123.27
19
Three Finger Brown
122.87
20
Grover C Alexander
122.80

So he ranks pretty high (again, out of 130 pitchers-Jack Morris was 96th and Morris only had 261 more career IP). Here are the Hall of Famers he is ahead of that he also exceeds in IP:

Three Finger Brown
Dennis Eckersley
Chief Bender
Stan Coveleski
Don Drysdale
Whitey Ford
Herb Pennock
Catfish Hunter
Rube Marquard
Juan Marichal
Joe McGinnity

I might have missed one or two. None of the stats were park adjusted. Using the Bill James Handbooks, I came up with a rough estimate of his HR park rating for his career of about 108, meaning his parks allowed about 8% more HRs than average. So he got no help here.

The Sinins database does have a park adjusted stat for pitchers called RSAA or runs saved against average. I put that into a per 9 IP stat for this group of pitchers. Here is that top 20:


Rank
Pitcher
Per 9 IP
1
Lefty Grove
1.53
2
Roger Clemens
1.34
3
Kid Nichols
1.21
4
Randy Johnson
1.15
5
John Clarkson
1.01
6
Cy Young
0.99
7
Greg Maddux
0.99
8
Walter Johnson
0.98
9
Curt Schilling
0.95
10
Whitey Ford
0.91
11
Grover C Alexander
0.91
12
Carl Hubbell
0.89
13
Amos Rusie
0.88
14
Kevin Brown
0.84
15
Three Finger Brown
0.84
16
Mike Mussina
0.82
17
Stan Coveleski
0.82
18
Bob Gibson
0.81
19
John Smoltz
0.81
20
Christy Mathewson
0.76

Again, Mussina does very well, saving .82 runs per 9 IP compared to the average pitcher and it is park adjusted (Jack Morris was 104th). Now it is not fielding independent, but it is impressive how high he ranks in both cases. Here are the guys he is ahead of in this last one and IP, too, who are in the Hall:

Juan Marichal
Joe McGinnity
Don Drysdale
Dennis Eckersley
Herb Pennock
Rube Marquard
Catfish Hunter
Chief Bender
Stan Coveleski