0 comments have been left for this article. BP Comment Quick Links
| |
|
|
August 1, 2001 Doctoring The NumbersPitch Counts in 2001Many of you are aware of a column that I write in conjunction with ESPN.com columnist Rob Neyer. While "Rob & Rany on the Royals" is, at its heart, a discussion of the affairs of that fallen franchise, our comments frequently touch upon all major-league teams. I recently made the following point about the Royals' pitch counts:
[O]verall, the Royals' starters haven't been abused as much as you might
think. Here's the breakdown of their pitch counts:
However, as impressive as this decline in pitch counts was, I did not
discount the possibility that this might only be part of a larger,
league-wide phenomenon.
Those of you unfamiliar with the concept of Pitcher Abuse Points
can click
here for an explanation.
I have chosen not to write about PAP this year, as I have turned the reins
over to Keith Woolner, who has a far greater understanding of the complexity
of the issue, and whose groundbreaking work in BP 2001 resulted in a radical
redesign of the formula used to calculate PAP. But Keith has provided me
with the up-to-the-moment PAP statistics for this season, and after checking
the numbers, I felt that I should get back to all of you, because what I
found was a real sea change in pitch counts this season.
In the chart below, "Stress" is PAP^3/NP, the same formula that
was used in BP 2001. The Roman numerals I-V refer to the five different
categories of pitch counts:
Category I: 1-104 pitches
Keep in mind that, as a result of Woolner's study in this year's book, we
can state with some confidence that the majority of abuse meted out to a
pitcher's arm occurs in Category IV or V starts.
Here is the data for this season as well as for 2000:
The Royals are, in fact, taking significantly better care of their pitchers
than a year ago; the team's Stress level has dropped from 31 to 21.
And yet the team is actually higher in the PAP standings than they were a
year ago. In 2000, six AL teams and 11 NL teams--more than half the teams in
baseball--had an average Stress level of 20 or more. This season? Only the
Royals and the Yankees in the AL, and the Diamondbacks (with Randy
Johnson and Curt Schilling) and Giants (with Livan
Hernandez and Russ Ortiz) in the NL, have Stress levels that
high. If we look at how many teams have seen their Stress level rise or fall
this season, we get the following chart:
Increased by 4 or more: 2
Only two teams have had a significant increase in their Stress levels: the
Marlins, as their young staff matures into a playoff-caliber rotation, and
the Tigers, where Jeff Weaver has been handed the mantle of staff
ace, with all the responsibilities that go with it, and where knuckleballer
Steve Sparks has been in the rotation all season.
Meanwhile, more than a third of major-league teams have seen their Stress
level drop by 13 or more, including teams like the Orioles, who have dropped
from 36 to 10; the Rangers, from 37 to 13; the Cubs, who despite Oscar
Acosta's insinuations to the contrary have dropped from 24 to 9; and even
the Giants, who have descended from the truly foolhardy height of 39 to a
more calculated risk of 26.
Every team in the NL had a Stress level of at least 14 in 2000. This year,
seven of the 16 teams have a Stress level in the single digits. There have
been just eight Category V starts in all of baseball this season (Chad
Durbin just pitched the ninth); last season, there were 40 of them.
This isn't progress measured in inches or feet, but in miles.
To what do we owe this most pleasant development? Certainly, the
re-definition of the strike zone must take a bow.
As Keith Woolner has
investigated elsewhere on this website, Sandy Alderson's mandate to increase
the number of strikes called by umpires has been motivated by a desire to
see overall pitch totals decline. This has worked by two different
mechanisms: decreasing the number of batters per game (in the AL, for
instance, OBP is down 15 points this season, which works out to an average
of 0.93 fewer batters per 27 outs), and decreasing the number of pitches per
batter.
This is certainly not the first time in history that we have seen a marked
downturn in offense, with a concomitant decline in pitches thrown per game.
What's interesting, though, is that in low-offense eras, the continuous
decline in complete games (an adequate substitute for high-pitch outings for
eras where pitch counts weren't kept) that has occurred throughout baseball
history has tended to slow down some. For instance, during the enormous
strike zone years from 1963-68, the percentage of complete games briefly
reversed course and began to increase:
1943 44.2%
1948 36.3%
1953 34.8%
1958 30.1%
1963 26.7%
1968 27.6%
1973 27.3%
1978 24.6%
1983 17.7%
1988 14.8%
1993 8.2%
1998 6.2%
This is intuitively correct. In an era when pitchers hold the advantage, and
are able to go nine innings and still have gas left in the tank, managers
are going to be slower to go to the bullpen. But what is truly interesting
is that, despite the downtick in offense this season, we are not seeing an
increase in complete games this season. In fact, complete games have
continued to drop dramatically, with a 19% decrease just from last season.
2000 4.81%
2001 3.89%
The philosophy of increased reliance on the bullpen, with larger and larger
pitching staffs (remember when 10-man staffs were the norm?) and LaRussian
specialization of the middle-relief roles, has become even more pervasive
this season, even as the overall workload on a pitching staff has declined
significantly. As teams continue to rely on their bullpens to pick up a
share of the workload even as the workload itself has gone down, it is not
surprising that the expectations placed on the starting pitchers has
lessened.
Here is the breakdown of the overall percentage of starts in each Category
for both 2000 and 2001:
59.0% 21.1% 16.5% 3.2% 0.3% 53.4% 20.2% 19.6% 6.0% 0.8%
While there has been an overall dramatic leftward shift in the numbers, it's
worth pointing out that the overall percentage of Category II and III starts
(i.e., pitch counts between 105 and 121) has only declined slightly, from
39.8% to 37.6%. Category IV and V starts have been cut nearly in half, from
6.8% to 3.5% of overall starts.
(In the interest of full disclosure, I should point out that this year's
numbers are artificially--if only minimally--depressed because the season is
just four months old. Pitchers are more likely to be stretched out towards
the end of the season, so we should expect the percentage of high-pitch
outings to increase slightly over the next two months. But certainly not
enough to make up the difference between this season and last.)
What are the possible ramifications of this across-the-board decline in
pitch counts? The potential long-term impact on young starters today should
not be underestimated. Consider the number of starting pitchers who began
their careers in the mid-to-late 1960s who would go on to enjoy extremely
long and productive careers. Steve Carlton, Nolan Ryan, Tom
Seaver, Phil Niekro, and Don Sutton all debuted in the
major leagues between 1964 and 1967, and all went on to win 300 games. By
comparison, only nine pitchers total that debuted since 1920 have gone on to
win 300 games. More than half of the 300-game winners in the live-ball
era made their debut in the mid-1960s. And one of the four others,
Gaylord Perry, debuted in 1962 and spent most of his formative years
in the 1963-68 heyday as well.
Three other pitchers who debuted between 1963 and 1968 won 250 games (Jim
Palmer, Tommy John, and Ferguson Jenkins.) Including
Roger Clemens and Greg Maddux, a total of 23 pitchers in the
live-ball era have won 250 games. Eight of them debuted in the mid-'60s, and
if we include Perry, Jim Kaat (debuted in 1959, was still just 24
years old in 1963) and Bob Gibson (debuted in 1959, threw the first
of his ten 240-inning seasons in 1963), we can claim that nearly half of the
250-game winners since 1920 were positively impacted by the low-scoring
levels of the 1960s.
True, the "pitching era" of 2001 doesn't hold a candle to 1968.
But let's also remember that 30 years ago pitchers were still throwing
complete games in a quarter of their starts, and even accounting for the
fewer pitches it took to get through nine innings, it seems likely that some
of today's young starters are throwing as few, if not fewer pitches, than
guys like Don Sutton and Tom Seaver were throwing in their early twenties.
So here's hoping that today's young starters, guys like C.C. Sabathia
(Stress level of 12, just one Category IV start all year), and Ben
Sheets (Stress level of 4, no outings of more than 114 pitches), and
Brad Penny (Stress level of 6), and Roy Oswalt and Tim
Redding (no 120-pitch outings between them), and the Tim
Hudson/Mark Mulder/Barry Zito trio in Oakland (Stress
levels under 20 and just one Category IV start each), continue to be handled
with the same care that they've received all season, in the hopes that some
of them may still be winning 15 games a year 15 years from now.
At the very least, each of them has a legitimate chance at such a long and
prosperous career, something we probably can't say for Livan
Hernandez, who by the way has a Stress level of 78 this season. Some
things, unfortunately, have not changed.
Rany Jazayerli is an author of Baseball Prospectus. You can contact him by
clicking here.
Rany Jazayerli is an author of Baseball Prospectus.
|
Playoff Prospectus: PECOTA Odds and ALCS Gam...