#38: Eddie Mathews

of the 113 participants, 97 had Eddie Mathews on their ballot (85.8%). 2 had him at #38 (1.8%). 35 and 60 had him ranked higher and lower, respectively (31.0% and 53.1%).


Eddie Collins Rank

Of the 101 people who ranked Eddie Collins in the top 50, 86 had him 39 or lower, 2 had him at #40 (not including Poz) and the remaining 13 had him between 41 and 50. The chart below shows the cumulative histogram,


the mean is 28.1, the most common values, 20 (9 times), 24 (6 times) and 29 (6 times).

return of the top 100 list! who will be #40.

So Poz wrote in a post a few days ago that he was planning to resume the top 100 list. The next one will be number 40, here is the breakdown of which player the contestants have pegged in that slot. For the record, WoC (wisdom of crowds) has Cal Ripken.

9 Eddie Mathews
9 Carl Yastrzemski

8 Johnny Bench

7 Nap Lajoie
7 Warren Spahn

5 George Brett
5 Roberto Clemente
5 Pedro Martinez

4 Pete Rose

3 Yogi Berra
3 Joe DiMaggio
3 Bob Gibson
3 Randy Johnson
3 Mel Ott
3 Albert Pujols
3 Cal Ripken

2 Cap Anson
2 Eddie Collins
2 Bob Feller
2 Lefty Grove
2 Al Kaline
2 Sandy Koufax
2 Phil Niekro
2 Jackie Robinson
2 Alex Rodriguez
2 Tris Speaker

1 Pete Alexander
1 Roger Clemens
1 Carlton Fisk
1 Omar Linares
1 Satchel Paige
1 Mike Piazza
1 Mickey Mantle
1 Kid Nichols
1 Ivan Rodriguez
1 Don Sutton
1 Cristobal Torriente
1 Mike Trout
1 Cy Young

interactive MDS plot

I’ve used d3 to create an interactive MDS plot. Roughly speaking, MDS (multi-dimensional scaling), is an algorithm that tries to represent the dis-similarity, i.e. the “distance” between a set of points, in lower dimensions. So in this case, I compute a “distance” between each pair of contestants based on how much their rankings differ, and then run the MDS algorithm and plot the results in 2d. In any case, I can’t post javascript here, so I put it here,


Click the numbers on the left to change the plot. The numbers represent the minimum rank to consider, that is, if you click 42, the MDS will be computed using ranks 42-50. Since 1-41 haven’t been posted yet, if you click those you will see the “Poz” entry move further and further away from the other names. I will update it as the new rankings from Poz get posted.


Poz100 player-rank heatmap

I have created a heat map that shows frequency of player-rank combinations. I limited it to the top 60 or so players to try to make the labels reasonably readable.

Here it is, click on it to enlarge it,





under-the-radar pitchers

my under-the-radar pitcher is Dazzy Vance. In 1924 he struck out 7.6 per 9, in a league that averaged 2.8 per 9. His numbers in other years are great also, but thats just an illustration. In terms of performance relative to your peers, I think you can make a good argument that his 1924 is the 3rd most dominant pitcher season ever, only behind Pedro 1999-2000 (of course).