Bill James has been running a series all week on the vagaries of pitcher W-L records (not behind
the paywall yet, so you can still access it). Six parts:
What W-L Record Does the Pitcher Deserve?
Fortune and Fame
The Could Be Hall of Famers
Luck, and the Cy Young Award
Lady Luck and the 20-Game Winner
Cleanup
If you think W-L records are a dumb, meaningless relic from another time, don’t bother. (And
if you count “Wins suck” as insight, I can even hook you up with a couple of like-minded folk.)
But if you know they’re a relic and still find them to be a sometimes interesting, occasional-
ly even useful form of shorthand--like WAR, like FIP, like anything else--then...interesting.
James’s contention has always been (or at least has been recently) that W-L records for starting
pitchers are often unreliable season to season--sometimes wildly so--but that, over the course
of a career, they are generally reliable for most pitchers.
A couple of highlights from James’s research if you don’t want to wade through the whole thing:
He breaks the 63 HOF starters into five groups: 1) 30 inarguable picks where W-L record is irrel-
evant (e.g., Tom Seaver); 2) Another 17 in the second tier, most of them a little unlucky but
who got voted in eventually and are fully deserving (e.g., Bert Blyleven); 3) Three pitchers who
benefitted from some other kind of luck besides wins and losses (e.g., Addie Joss--he doesn’t
specify what other kind of luck; maybe Joss won the lottery); 4) Four who were a bit lucky with
their W-L record and may or may not have made it with average luck (e.g., Juan Marichal); 5) Nine
who got in primarily because of W-L luck (Catfish Hunter, Early Wynn, and Bob Lemon being the
most recent).
Of the 102 starting pitchers who won the Cy, James believes 72 deserved it or were close enough
for a valid argument; 30 got it because of W-L luck. This has gotten better over time: he contends
that 52 of the past 60 were deserving.
The bad-luck season that most fascinates me is Dave Stieb’s 1985. It’s not as famous as Felix
Hernandez's Cy year, but the disconnect is almost identical: ERA+ 174, WAR 7.1, and 13-12 for
Felix, ERA+ 171, WAR 6.9, and 14-13 for Stieb. The really bizarre thing, though, is that whereas
Felix was on a terrible team (61-101) that had the league’s worst offense by 100 runs--his W-L
record is far from shocking when viewed in the context of the team he played for--the Jays in
’85 had their big breakthrough year, winning 99 games, winning their first divisional title, and
finishing fourth in the league in runs scored. The Jays and Stieb were perfectly in sync in 1985;
you would intuitively expect that the team’s ace, having a great year and at the end of a four-
year run where he was clearly the best pitcher in the league, would have run up one of those
ungodly W-L records that almost always won Cy Youngs at the time, 22-3 or something like that.
James calls the ’85 Cy Young a toss-up between Saberhagen, the actual winner, and Stieb; WAR gives
it to Saberhagen, 7.3-6.9. With normal luck, James projects that Stieb would have ended up 20-11;
he doesn’t go through game logs, but bases that on league-average runs, Pythagorean expectations,
and various other adjustments. In trying to figure out how Stieb ended up 14-13, I instead looked
at his game logs.
Three salient points:
1) Only one of his 14 wins was cheap; on such a good team, you would think he’d pick up at least
two or three cheap wins.
2) In 26 of 36 starts, he only gave up 0, 1, or 2 runs (pitching at least 6 innings); his record in
those games was 13-5, with 8 no-decisions.
3) The Jays finished fourth-last in the league that year in save percentage; he might have lost two
or three wins that way. (Henke was great, but he didn’t arrive until August; they had Bill Caudill,
Jim Acker, and Gary Lavelle closing before that.)
So after three years of being the best starter in the AL and not having a lot to show for it, Stieb
was great again in ’85, the team finally caught up with him (they’d been good in ’83 and ’84, too,
but not this good), and he somehow ended up 14-13. Bizarre.
I don’t know--if Stieb had actually gotten lucky that year, above and beyond merely normal luck, he
might have gone 23-7. Start by adding a couple of cheap wins, games where he gave up three or four
runs but pitched deep enough into the game to still get the win. Stieb had four such games in 1985
and lost them all. If you convert two of those losses into wins, he moves to 16-11. I’m going to
assume there were a couple of games where he left with a lead that was blown by the bullpen (I know,
I could check this...); if you convert one loss, and give him another win that maybe ended up as a
no-decision, he moves to 18-10. Give him back two more of those tough losses, the games where he
held the other team to 0, 1, or 2 runs; Stieb is 20-8 now. Finally, take another two of those no-
decisions where he pitched really well and turn them into wins. Final record, 22-8--even better than
the “deserved” record of 20-11 James gives him. That’s about where I think he would have ended up
with good luck, and that's maybe even on the conservative side. The blown-save tweak may overlap a
bit with some of the other adjustments.
I’m positive that 22-8, even 20-11, plus the ERA title would have given Stieb the Cy Young that year.
Saberhagen was only in his second year, and WAR wasn’t even in the lab yet. (James had basically
laid the foundation for WAR among position players by then, but he hadn’t done as much with pitch-
ing.) That one season, reconfigured accordingly, wouldn’t have been enough single-handedly to push
Stieb into the HOF, but it would have been a start. Make some similar adjustments to other years,
as James does in “Fortune and Fame,” and you never know. He would have at least ended up with a
career line closer to what Roy Halladay will (deservedly) go into the HOF with.
Strange, strange season--even stranger than Felix’s 2010.
A complete breakdown of Stieb’s 36 starts:
Starts where he qualified for the win (5+ innings)
0 earned runs – 7 starts
1 earned run – 8 starts
2 earned runs – 11 starts
3 earned runs – 1 start
4 earned runs – 3 starts
5 earned runs – 3 starts
7 earned runs – 1 start
Starts where he didn’t qualify for the win (fewer than 5 innings)
3 earned runs – 2 starts