Guessing Number One Was Easy

[NetShrine]

This is very enlightening. Using the Sinins' Saber Encylo, I ran a report of Baserunners per 9 IP, career, 1000 min IP, as compared to league average:

BASERUNNERS/9 IP DIFF
1 Pedro Martinez 3.41
2 Mike Mussina 2.67
3 Walter Johnson 2.33
4 Cy Young 2.21
5 Tiny Bonham 2.19
6 Bret Saberhagen 2.18
7 Curt Schilling 2.17
8 Roger Clemens 2.14
9 Carl Hubbell 2.14
10 Ed Walsh 2.13
11 Addie Joss 2.11
12 Greg Maddux 2.10
13 Lefty Grove 2.04
14 Christy Mathewson 2.03
15 Dick Hall 2.03
16 Babe Adams 2.01
17 Kid Nichols 2.00
18 Spud Chandler 1.98
19 Three Finger Brown 1.97
20 Sandy Koufax 1.95

Mussina @ #2 all-time. Did you know that? Better than Johnson, Clemens, Maddux, and Koufax! Nice company Moose. Way to go!

[nyy26wc]

I love that list. I couldn't wait to show it to a friend when I first discovered it right after the Yankees signed Mussina.

[mainsr]

LET'S SEE HOW THEY END THEIR CAREERS. Pitchers have a way of polluting their numbers during the down years.

The most interesting number to me is Hubbell better than Grove. If, as I assume, these numbers are weighted by innings, it may suggest that Grove may not be the best lefty of all time, much less the best pitcher.

[NetShrine]

It's the wins thing - - people see wins and think "better" or "best" - - the yardstick should be control of the zone BB/K, power K/IP, and effectiveness H+BB/IP. Wins are luck.

[nyy26wc]

Originally posted by mainsr
The most interesting number to me is Hubbell better than Grove. If, as I assume, these numbers are weighted by innings, it may suggest that Grove may not be the best lefty of all time, much less the best pitcher.

Those numbers are not weighted by innings, but there is a big degree of weighting going on.

Here's what I mean--

To create that chart, each pitcher was compared to his league average. The league average is perfectly tailored to that particular pitcher.

For example, Hubbell had 124 IP in the NL in 1928, which was 3.45% of his career total. So, the 1928 NL average counts towards 3.45% of Hubbell's league average.

That particular chart is base runners per 9 innings. Whenever you compute a stat on a per 9 IP basis, it's not weighted towards pitchers who pitch more innings.

No problem.

All we have to do is rerun the chart by comparing each pitcher's total career number of baserunners allowed to his league average.

Here are those leaders--

1 Cy Young 1808
2 Walter Johnson 1532
3 Kid Nichols 1122
4 Grover C Alexander 1101
5 Christy Mathewson 1079
6 Don Sutton 1014
7 Tom Seaver 1002
8 Warren Spahn 944
9 Robin Roberts 930
10 Lefty Grove 891
11 Roger Clemens 873
12 Carl Hubbell 855
13 Ferguson Jenkins 825
14 Tim Keefe 794
15 Greg Maddux 774
16 Gaylord Perry 733
17 Juan Marichal 713
18 Ed Walsh 703
19 John Clarkson 684
20 Bert Blyleven 681

Other active players in the top 100--

27 Bret Saberhagen 618 (in case he ever returns)
29 Pedro Martinez 597
30 Mike Mussina 597 (Pedro's 597 is a little higher when decimal points get involved)
45 Randy Johnson 488
48 Kevin Brown 470
49 Curt Schilling 459
54 David Wells 445
67 John Smoltz 420
76 David Cone 384 (assuming he's not done)

Meanwhile, when we look at the chart in reverse, Bobby Witt is the worst pitcher ever, by this particular standard. Coming into this season, Witt allowed 354 more baserunners than his league average pitcher--3843 vs. 3489. That puts him a confortable distance ahead of Chick Fraser, who allowed a "mere" 315 more baserunners than his average.