Many thanks to Steve Smith for the posting on Game Scores.
As to whether it values the strikeout too much, I decided to look at games
from Monday, which featured most staff aces, and most of the game's big
Herewith, each starting pitcher's game score, his strikeout total in
parentheses, and the outcome of the game for his team (not just W or L awarded):
McDowell, ChiA (10) 84 (!) W
DMartinez, Mon (5) 78 W
Clemens, Bos (6) 72 W
Gooden, NYN (7) 72 W
Browning, Cin (4) 68 W
Mulholland, Phi (3) 63 L
Saberhagen, KC (5) 62 W
Leary, NYA (9) 52 L
Knudson, Mil (3) 49 W
Ryan, Tex (9) 46 L
Swindell, Cle (2) 43 L
Drabek, Pit (3) 40 L
Stieb, Tor (2) 39 L
Tanana, Det (2) 35 W
Ballard, Bal (3) 34 L
Scott, Hou (1) 32 L
I thought this would be an interesting sample for at least two reasons:
1) We're all pretty familiar with these pitchers
2) We all probably know something about how they pitched this particular
day, as (I assume) we all savor the feeling of finally looking at fresh
game reports/box scores after the long, dark winter.
My first impression was, "Hey, great stat, everybody's about right, from
incredible (McDowell) to stinkin' up the joint (Scott)." Then I started to
wonder a bit, as the highest game score went to the pitcher with the most
strikeouts and the lowest went to the pitcher with the least. Distribution
looks to be close to K/ER, so that the list goes "pitchers who won with lots
of strikeouts, pitchers who won with few strikeouts, pitchers who lost with
lots of strikeouts, pitchers who lost with few Ks."
Then, on looking at the list (and the box scores) some more, a few anomalies
stuck out Martinez, with the second highest score, had only 5 Ks. Mulholland,
with the highest score among losers, had only 3 Ks. Martinez threw a 7-inning,
one-hit shutout, and got 4 points less than McDowell, who threw a complete-game
4-hitter, surrendering 1 earned run. The two 9-strikeout losers are second and
third among losers. Tanana and Knudson, with the fewest Ks among winners, are
near the bottom.
"Uh, oh," you say, "Lose the stat."
Weeellllll, maybe not. Check out the innings and earned runs:
McDowell 9 1
Martinez 7 0
Clemens 8 1
Gooden 8 1
Browning 8.3 2
Mulholland 7 2
Saberhagen 7 1
Leary 6 4
Knudson 5.3 3
Ryan 7 5
Swindell 6 3
Drabek 5 3
Stieb 5 5
Tanana 5 4
Ballard 5.7 6
Scott 4 5
Now, obviously, this isn't a perfect correlation. Leary with 6 innings/4 ER is
ahead of Swindell with 6/3 ER. But although Swindell had 7 fewer Ks, he also
gave up an unearned run and 3 more hits that Leary didn't. You certainly can
argue that the earned run shouldn't hurt him, but the game score stat says
it should. James' rule.
I also don't mind that McDowell's 2 extra innings and 5 extra K's outweigh
1 earned run advantage by Martinez. That seems more than fair.
All in all, I think there's a stronger correlation between game score and
ER/IP (or really IP/ER) than there is between GS/K or even GS/K/IP.
I like the stat.
Steve, if you were going to do this, sorry -- I just thought it was a good
chance to examine the question about strikeout weight in the stat. Thanks for
bringing it up -- I may just have to start a notebook, too.
The Reds are only 1449 outs away from a 162-0 season!