"Baseball is 75% pitching." "Pitching and defense are what
win baseball games. " We’ve heard these cliches spouted countless
times, but thanks to the work of analysts like Bill James, among many
others, we know they don’t stand up to scrutiny. Scoring runs plays just as
important a role in winning as preventing runs.
However, while pitching is not the primary determinant of, it is still
unusual for a team to win consistently without it. There aren’t many teams
who get to the playoffs with pitching that’s out-and-out bad. The
1998 Texas Rangers were one such team. The Rangers managed to win 88 games
despite ranking 27th out of 30 major league teams in runs allowed. That’s
lousy even after taking the hit-happy Ballpark in Arlington into
(At this point I could provide pitcher-by-pitcher assessments of last
year’s Ranger staff, but with the language I’d have to use, our web site
would end up "banned" by family-friendly search engines. Suffice
to say that the 1998 Rangers were an example of a team that succeeded
despite their pitching rather than because of it. The offense was
good to an even greater extent than the pitching was bad.)
That was last year. The question now is: Does the Rangers’ "all hit,
no pitch" 1998 tell us anything about what to expect from the 1999 team?
When I started thinking about this issue in the offseason, my first thought
was that the Rangers were in good shape. I speculated that a team with bad
pitching whose offense was good enough to carry it to a +.500 season is
more likely to maintain its success than other types of successful teams.
The reasoning is this: Pitchers are flaky. Their performance levels vary
more from year to year than those of hitters’. Therefore, team pitching
performance should also vary more from year to year than team hitting,
i.e., a team’s pitching performance in a given year reflects luck and
randomness to a greater extent than a team’s hitting. If that’s true, then
you’d expect a team that scored a lot of runs to be more likely to maintain
that high level of offense, while a team that gave up a lot of runs would
be more likely to regress toward the mean the next year, just because of
the greater randomness of pitching performance. If a team has both good
hitting and bad pitching, on average I’d expect pitching to improve about
as much or more than hitting declines. As a result, I’d expect those kinds
of teams to maintain or improve on their overall W/L success.
Designing the Study
That’s a theory, and an easily testable one. I looked at all major league
teams that met two criteria over the past 20 years:
- they allowed 40 more park-adjusted runs than a league-average team
- they scored more runs than they allowed.
I use 40 here to make the tables a manageable size, but I looked at the
numbers for 20, 30 and 50 runs above average, and the general results are
still the same. Note that the 1998 Rangers don’t even meet the 40-run
I call these the All-Hit-No-Pitch teams. The question we want to answer: To
what extent did these teams maintain their success the following year? The
results for all qualifying teams are below.
|TEAM||R||OR||Pyth||Next Year Pyth||Change|
R: park-adjusted runs scored compared to an average team
OR: park-adjusted runs allowed compared to an average team
Pyth: the Pythogorean winning percentage, a projection based on runs
scored and runs allowed, originally published by Bill James
Next Year Pyth: the Pythagorean-projected winning percentage for the
Change: how much the team’s projected winning percentage improved or
declined from one year to the next.
The All-Hit-No-Pitch teams did drop off the following year, but only a
little. They went from a collective .528 Pythagorean winning percentage in
their All-Hit-No-Pitch year to .526 the following year. Clearly, these
All-Hit-No-Pitch teams maintained their success pretty well; that drop of
.002 in winning percentage represents a loss of less than half a game in
But maybe all successful teams maintain their success equally well. How do
the All-Hit-No-Pitch teams compare to other types of successful teams, in
particular their pitching-dominated counterparts? That’s the interesting
part. I did the same evaluation for All-Pitch-No-Hit teams of the past 20
years: those that scored more than they allowed despite scoring 40 fewer
park-adjusted runs than an average team.
|TEAM||R||OR||Pyth||Next Year Pyth||Change|
|87 White Sox||-60||69||.501||.410||-.091|
|91 Blue Jays||-69||123||.547||.567||+.019|
The All-Pitch-No-Hit teams’ Pythagorean winning percentage dropped like a
rock the following year, from .529 to .507. That .022 drop in winning
percentage represents a loss of more than three-and-a-half games in the
The bottom line: a winning team with lousy pitching will, on average, have
a better season next year than a winning team with lousy hitting. Of
course, this result does not mean that having lousy pitching is a good
thing, or that teams should be ignoring pitching when making personnel
decisions, or anything of the kind. Naturally a team with good hitting
and good pitching is going to be better than one with only good
hitting. What the results do give us is a useful predictive indicator for
teams when you already know their makeup.
What does it say about the 1999 Rangers? Something encouraging for Ranger
fans. But it certainly doesn’t mean the Rangers can stand still and expect
to become a top-tier team. In fact, one of the reasons I expect that the
crummy pitching teams maintained their success so well is that pitching
problems are easier to fix than hitting problems. The right front-line
starter can save a team 50 or more runs.
The Rangers certainly have not fixed their pitching problems; they
have failed so far in their attempts to land an ace for the staff, and that
failure has shown up in the R and RA columns, if not the standings. Through
June 7, the Rangers have allowed about as many runs as they’ve scored–317
allowed vs. 319 scored–for a .503 Pythagorean winning percentage. That
their actual winning percentage is .600 can be attributed in large part to
good luck, with a nod to the excellent performance of their bullpen.
Bringing in a Curt Schilling, Kevin Appier or Brad Radke before the trading
deadline will improve the Rangers’ chances of being another
All-Hit-No-Pitch success story.