Baseball Primer’s 2001 American League Cy Young Award
Unlike the NL award voting, the Baseball Primer staff vote for the AL CY Young
Award reflected how difficult it was to settle on one candidate this year. Eleven
voters cast first-place ballots for five different pitchers, and no pitcher
was named on all 11 ballots.
The BBWAA vote was, of course, not this close. Roger Clemens?s 20-1 start to
the season guaranteed a solid victory for him in that tally.
And of course, that W/L record, coupled with Clemens?s ninth-place finish in
ERA, his eleventh-place ranking in BP?s Support-Neutral Won-Lost record, and
the traditional stathead “W/L record doesn?t mean anything” feeling, had created
the expected backlash against the BBWAA vote.
The arguments raised against Clemens are certainly legitimate. He was the
beneficiary of run support in posting the 20-3 record, and had he been supported
at the level of his teammate Mike Mussina he would likely not have had that W/L
record. But I think that the stathead argument overlooks two other important
aspects of Clemens?s performance ? beyond the W/L record - that support his
The more important of the two factors is defense. Of the top nine pitchers in
ERA (the eight pitchers listed above plus Barry Zito), Clemens (and Mussina)
played on the team with the worst defense. I looked at this in two ways ? by
calculating Bill James?s Defensive Efficiency Rating (DER) and by calculating
Voros McCracken?s $H factor (percentage of hits per fair ball in play) from his
Defense Independent Pitching Method (DIPS).
Seattle?s DER is the highest that I remember seeing ? this in an era where
DERs have been dropping steadily. Oakland, Chicago, and Minnesota are all above
the norms for the AL (DER 0.688, $H 0.294). The Yankees were the only team below
We can estimate the impact of defense on these pitchers by applying DIPS to
the top nine pitchers in ERA. I did that, using my own calculated park factors
for 2001 (which can be found
href="http://www.geocities.com/mwemeigh">here, by clicking on the 2001 Park
Data link ? these are batter park factors, with the pitcher park factors used
in DIPS being 1/BPF).
While I believe that DIPS likely overstates the effect, I am absolutely
convinced that the difference between Garcia?s 3.05 ERA and Clemens?s 3.51 ERA
is due almost entirely to Garcia?s defensive support and the pitching-friendly
nature of his ballpark (even in comparison to Yankee Stadium).
The second aspect of Clemens?s performance that strengthens his candidacy has
to do with when he gave up his runs. The assumption in context-neutral game
methods is that a pitcher who gives up three runs in seven innings would do that
whether he?s supported by no runs, or three runs, or six runs, or nine runs over
that time frame. This assumption is not necessarily warranted. Teams tend to
change the way that they defend when they have a lead ?they will often trade
runs for outs. Pitchers tend to make changes to the way that they approach
hitters when they have a lead ? the focus is more on throwing strikes and making
the batter put the ball into play. If a pitcher is allowing a fair percentage of
his runs when his team has a good-sized lead, he is ? in the context of the
games that he is pitching ? costing his team less than if he is allowing those
runs with the game close, or with his team trailing.
I looked at five of the top nine pitchers in ERA: Clemens, Garcia, Mussina,
Mays, and Buehrle. Due to the publication deadline, I didn?t have time to look
at the other four pitchers (the Oakland trio and Moyer), but I will post the
results for those four pitchers as a comment to this article. I looked at how
they pitched in three situations:
The game situation was figured at the start of the inning, and all of the
results in that inning were credited to that situation; e.g. if Team A was ahead
2-0 at the start of the inning and the pitcher allowed 3 runs, that became 1 IP
and 3 runs allowed in the “leading by 2 or more” category.
Clemens was the best of these five pitchers both when his team was behind and
when his team was in a close game ? in the game situations where it was most
important to prevent runs from scoring. Mussina and Garcia weren?t all that far
behind him; Buehrle and Mays are further down the list.
These two factors ? defense and pitching within game context ? to me lift
Clemens and Mussina out of the pack. I voted for Clemens in the Primer poll
largely for these two reasons, but after doing the more detailed comparisons I?d
have voted for Mussina? he had more innings in key situations than did Clemens
and didn?t pitch significantly worse than Clemens in those situations, so he was
likely more valuable overall.
I know I haven?t addressed quality of competition. The two worst teams in the
AL were in the East, and with the unbalanced schedule Clemens and Mussina got
more opportunities against Baltimore and Tampa than did their competition. The
AL West teams also faced stronger interleague competition from the NL West.
Theoretically, then, the Yankees could be knocked down a peg or two because they
played weaker teams more frequently. The counter-argument to be made is that
early in the season Seattle was the only team playing well in the West, while
both Boston and Toronto looked like they?d be giving New York a good run, thus
it?s only in hindsight that the Western teams appear to be tougher competition.
That?s another article; this one is already long enough. I tend to downplay the
argument, but I?m willing to be convinced otherwise.
Posted: November 23, 2001 at 06:00 AM | 17 comment(s)
Login to Bookmark