I was watching the game last night and Campbell or Mulliniks seemed
extremely impressed about something. I'm not sure exactly what - I
wasn't giving them my undivided attention. But I thought they said
something to the effect that Kevin Millwood has an exceptionally good
record in games where he throws more than 100 pitches.
UPDATED! Woo-hoo!
Is that really what they said? For the life of me, I'm not sure.
My ears perked up, however, like those of some rough beast, its hour come round at last. "Say what?" thought I to myself. Shouldn't everybody have an impressive record when you throw 100 pitches? Most of the time, they're not going to let you throw that many pitches unless you're actually pitching well.
So here’s today's bit of odd information, with Data Tables and everything!
As it turns out, Millwood threw more than 100 pitches in 13 of his 29 starts for Texas in 2008. He went 5-4 in those games. Which isn't very striking at all, and immediately made me think I'd misunderstood whatever it was Campbell and Mulliniks were talking about.
But by now I was rolling, baby!
Now of course - !00 pitches as a marker of anything is about as meaningless as you could possibly want. Meaningless, I tell you. First of all, it's utterly arbitrary - nothing magical happens between pitches 99 and 100. Plus, nothing is being taken into account regarding how long all this work is taking - is it an hour and forty-five minutes? Two and a half hours? That’s something I’d certainly want to know. And what the weather is like - you definitely don't want to be doing this sort of thing when you can see your breath. Ask Al Leiter. It’s probably not a good idea when they’re measuring the temperature with three digits either, as they often do in Texas. The game circumstances matter as well - every pitcher will tell you that close games and tight situations are far more taxing than blowouts.
The only thing 100 pitches has going for it is this: it's a round number. It feels bigger than 99 to the men who actually run the teams, and make the relevant decisions. Even if they know better.
There's a reason retail outlets price items at $1.99 instead of $2.00 - the reason is because it actually works.
Anyway, let's mess around with some Data Tables.
First, the Team Data, ranked by the percentage of times the starter throws 100 or more pitches.
So there. Last year in the 2268 games played by AL teams, the starting pitcher threw at least 100 pitches in 980 of those games - that's 43% of the time. In those 980 games, the starter came away with a W in 422 of them (43%), took the L 300 times (30%), and ended up with a ND in 258 games (26%.) That 422-300 record is a .584 W-L percentage.
Now as it happens, AL starting pitchers went 824-778 in 2008. So if we subtract the 100+ pitch games, we find out that when throwing fewer than 100 pitches, AL starters went 402-478 with 408 NDs - that's a winning percentage of .457, and that's sort of what you'd expect, right?
As for Kevin Millwood - coming into last night's game, Millwood was 31 games over .500 for his career (now it's 30!). He's got a 2-0 career record as a relief pitcher. As a starter, he's 23 games over when he throws 100+, and 5 games over when he throws less than that. Doesn't seem all that unusual to me at all. Here are the specifics:
So Millwood's got a .589 winning percentage when he throws more than 100+ pitches - that seems extremely normal to me. As we just learned, last year, all AL pitchers - the good, the bad, and the ugly - had a .584 winning percentage when throwing 100+ pitchers.
It is in fact Millwood's career .520 winning percentage when he throws less than 100 pitches that's really impressive. In other words, the exact opposite of the point they were making on last night's broadcast.
Sigh.
Well, having done all this work, let me share! Here are the individuals:
I like the fact that Paul Byrd threw more than 100 pitches just 3 times in his 30 starts for Cleveland and Boston. Nick Blackburn of the Twins and Andy Sonnanstine of the Devil Fishies were the only other pitchers to make at least 30 starts and have fewer than ten 100+ pitch outings.
At the other extreme... just five AL hurlers threw at least 100 pitches in 80 percent of their starts. (No, I'm not counting Nick Masset.) Two of them - A.J. Burnett and C.C. Sabathia - are now working for the Yankees. (Sabathia's NL workload, by the way, almost exactly matches his AL pattern - 17 starts, 14 100+ outings, 10-1 with 3 NDs). This is interesting if only because no AL team was less likely than the Yankees to have a starter hang around long enough to throw 100 pitches. Andy Pettitte was the only Yankee who did this in even half of his 2008 starts. The others are Ervin Santana, Justin Verlander, and Cliff Lee.
UPDATED! Woo-hoo!
Is that really what they said? For the life of me, I'm not sure.
My ears perked up, however, like those of some rough beast, its hour come round at last. "Say what?" thought I to myself. Shouldn't everybody have an impressive record when you throw 100 pitches? Most of the time, they're not going to let you throw that many pitches unless you're actually pitching well.
So here’s today's bit of odd information, with Data Tables and everything!
As it turns out, Millwood threw more than 100 pitches in 13 of his 29 starts for Texas in 2008. He went 5-4 in those games. Which isn't very striking at all, and immediately made me think I'd misunderstood whatever it was Campbell and Mulliniks were talking about.
But by now I was rolling, baby!
Now of course - !00 pitches as a marker of anything is about as meaningless as you could possibly want. Meaningless, I tell you. First of all, it's utterly arbitrary - nothing magical happens between pitches 99 and 100. Plus, nothing is being taken into account regarding how long all this work is taking - is it an hour and forty-five minutes? Two and a half hours? That’s something I’d certainly want to know. And what the weather is like - you definitely don't want to be doing this sort of thing when you can see your breath. Ask Al Leiter. It’s probably not a good idea when they’re measuring the temperature with three digits either, as they often do in Texas. The game circumstances matter as well - every pitcher will tell you that close games and tight situations are far more taxing than blowouts.
The only thing 100 pitches has going for it is this: it's a round number. It feels bigger than 99 to the men who actually run the teams, and make the relevant decisions. Even if they know better.
There's a reason retail outlets price items at $1.99 instead of $2.00 - the reason is because it actually works.
Anyway, let's mess around with some Data Tables.
First, the Team Data, ranked by the percentage of times the starter throws 100 or more pitches.
100+ Starter's
Team Starts pitches Pct. W Pct. L Pct. ND Pct. W/L Pct.
TORONTO 162 91 .562 48 .527 24 .264 19 .209 .667
LOS ANGELES 162 91 .562 39 .429 20 .220 32 .352 .661
KANSAS CITY 162 83 .512 31 .373 31 .373 21 .253 .500
CHICAGO 163 80 .491 36 .450 27 .338 17 .213 .571
SEATTLE 162 76 .469 21 .276 32 .421 23 .303 .396
TAMPA BAY 162 74 .457 39 .527 18 .243 17 .230 .684
BOSTON 162 74 .457 41 .554 13 .176 20 .270 .759
DETROIT 162 71 .438 24 .338 28 .394 19 .268 .462
AL TOTAL 2268 980 .432 422 .431 300 .306 258 .263 .584
BALTIMORE 161 66 .410 24 .364 24 .364 18 .273 .500
CLEVELAND 162 63 .389 33 .524 15 .238 15 .238 .688
TEXAS 162 57 .352 26 .456 12 .211 19 .333 .684
MINNESOTA 163 54 .331 27 .500 13 .241 14 .259 .675
OAKLAND 161 51 .317 15 .294 24 .471 12 .235 .385
NEW YORK 162 49 .302 18 .367 19 .388 12 .245 .486
So there. Last year in the 2268 games played by AL teams, the starting pitcher threw at least 100 pitches in 980 of those games - that's 43% of the time. In those 980 games, the starter came away with a W in 422 of them (43%), took the L 300 times (30%), and ended up with a ND in 258 games (26%.) That 422-300 record is a .584 W-L percentage.
Now as it happens, AL starting pitchers went 824-778 in 2008. So if we subtract the 100+ pitch games, we find out that when throwing fewer than 100 pitches, AL starters went 402-478 with 408 NDs - that's a winning percentage of .457, and that's sort of what you'd expect, right?
As for Kevin Millwood - coming into last night's game, Millwood was 31 games over .500 for his career (now it's 30!). He's got a 2-0 career record as a relief pitcher. As a starter, he's 23 games over when he throws 100+, and 5 games over when he throws less than that. Doesn't seem all that unusual to me at all. Here are the specifics:
100+ pitches 99 or less
Starts W L ND W L ND
1997 8 1 1 0 3 2 1
1998 29 11 2 0 6 6 4
1999 33 9 4 5 9 3 3
2000 35 7 6 5 3 7 7
2001 21 1 5 0 6 2 7
2002 34 5 2 3 12 6 6
2003 35 12 6 4 2 6 5
2004 25 4 1 5 5 5 5
2005 30 4 9 5 5 2 5
2006 34 9 5 3 7 7 3
2007 31 7 6 3 3 8 4
2008 29 5 4 4 4 6 6
2009 4 1 2 1 0 0 0
TOTAL 348 76 53 38 65 60 56
So Millwood's got a .589 winning percentage when he throws more than 100+ pitches - that seems extremely normal to me. As we just learned, last year, all AL pitchers - the good, the bad, and the ugly - had a .584 winning percentage when throwing 100+ pitchers.
It is in fact Millwood's career .520 winning percentage when he throws less than 100 pitches that's really impressive. In other words, the exact opposite of the point they were making on last night's broadcast.
Sigh.
Well, having done all this work, let me share! Here are the individuals:
100+ Starter's
Pitcher Team Starts pitches Pct. W Pct. L Pct. ND Pct. W/L Pct.
Santana LAA 32 28 .875 14 .500 6 .214 8 .286 .700
Burnett Tor 34 28 .824 14 .500 8 .286 6 .214 .636
Verlander Det 33 27 .818 10 .370 14 .519 3 .111 .417
Halladay Tor 33 26 .788 18 .692 7 .269 1 .038 .720
Lee Cle 31 25 .806 18 .720 2 .080 5 .200 .900
Meche KC 34 24 .706 10 .417 7 .292 7 .292 .588
Hernandez Sea 31 22 .710 6 .273 8 .364 8 .364 .429
Vazquez Chi 33 22 .667 7 .318 11 .500 4 .182 .389
Buehrle Chi 34 21 .618 8 .381 8 .381 5 .238 .500
Matsuzaka Bos 29 20 .690 14 .700 0 .000 6 .300 1.000
Guthrie Bal 30 20 .667 7 .350 6 .300 7 .350 .538
Bannister KC 32 20 .625 6 .300 9 .450 5 .250 .400
Garza TB 30 18 .600 10 .556 6 .333 2 .111 .625
Cabrera Bal 30 18 .600 6 .333 4 .222 8 .444 .600
Lester Bos 33 18 .545 12 .667 1 .056 5 .278 .923
Pettitte NY 33 18 .545 7 .389 7 .389 4 .222 .500
Washburn Sea 26 17 .654 3 .176 9 .529 5 .294 .250
Beckett Bos 27 17 .630 7 .412 7 .412 3 .176 .500
Weaver LAA 30 17 .567 6 .353 4 .235 7 .412 .600
Greinke KC 32 17 .531 7 .412 4 .235 6 .353 .636
Kazmir TB 27 16 .593 7 .438 4 .250 5 .313 .636
Padilla Tex 29 16 .552 9 .563 3 .188 4 .250 .750
Jackson TB 31 16 .516 8 .500 3 .188 5 .313 .727
Garland LAA 32 16 .500 5 .313 4 .250 7 .438 .556
Floyd Chi 33 16 .485 10 .625 3 .188 3 .188 .769
Sabathia Cle 18 15 .833 6 .400 6 .400 3 .200 .500
Lackey LAA 24 15 .625 7 .467 3 .200 5 .333 .700
Galarraga Det 28 15 .536 7 .467 3 .200 5 .333 .700
Baker Min 28 15 .536 5 .333 3 .200 7 .467 .625
Saunders LAA 31 15 .484 7 .467 3 .200 5 .333 .700
Shields TB 33 15 .455 8 .533 4 .267 3 .200 .667
Blanton Oak 20 13 .650 4 .308 7 .538 2 .154 .364
Litsch Tor 28 13 .464 8 .615 2 .154 3 .231 .800
Millwood Tex 29 13 .448 5 .385 4 .308 4 .308 .556
Marcum Tor 25 12 .480 4 .333 4 .333 4 .333 .500
Mussina NY 34 12 .353 5 .417 4 .333 3 .250 .556
Davies KC 21 11 .524 6 .545 3 .273 2 .182 .667
Hernandez Min 23 11 .478 4 .364 3 .273 4 .364 .571
Smith Oak 32 11 .344 3 .273 6 .545 2 .182 .333
Contreras Chi 20 10 .500 4 .400 3 .300 3 .300 .571
Duchsherer Oak 22 10 .455 3 .300 5 .500 2 .200 .375
Wakefield Bos 30 10 .333 3 .300 2 .200 5 .500 .600
Danks Chi 33 10 .303 6 .600 2 .200 2 .200 .750
McGowan Tor 19 9 .474 3 .333 2 .222 4 .444 .600
Slowey Min 27 9 .333 5 .556 4 .444 0 .000 .556
Robertson Det 28 9 .321 3 .333 4 .444 2 .222 .429
Rogers Det 30 9 .300 2 .222 5 .556 2 .222 .286
Dickey Sea 14 8 .571 1 .125 4 .500 3 .375 .200
Carmona Cle 22 8 .364 5 .625 2 .250 1 .125 .714
Liz Bal 17 7 .412 2 .286 4 .571 1 .143 .333
Burres Bal 22 7 .318 2 .286 5 .714 0 .000 .286
Feldman Tex 25 7 .280 1 .143 3 .429 3 .429 .250
Perkins Min 26 7 .269 6 .857 0 .000 1 .143 1.000
Silva Sea 28 7 .250 1 .143 4 .571 2 .286 .200
Sonnanstine TB 32 7 .219 5 .714 1 .143 1 .143 .833
Waters Bal 11 6 .545 3 .500 3 .500 0 .000 .500
Rowland-Smith Sea 12 6 .500 3 .500 2 .333 1 .167 .600
Bonderman Det 12 6 .500 1 .167 2 .333 3 .500 .333
Bedard Sea 15 6 .400 3 .500 2 .333 1 .167 .600
Batista Sea 20 6 .300 2 .333 3 .500 1 .167 .400
Rasner NY 20 6 .300 1 .167 5 .833 0 .000 .167
Olson Bal 26 6 .231 4 .667 0 .000 2 .333 1.000
Eveland Oak 29 6 .207 2 .333 2 .333 2 .333 .500
Ponson Tex 9 5 .556 4 .800 0 .000 1 .200 1.000
Chamberlain NY 12 5 .417 2 .400 1 .200 2 .400 .667
Miner Det 13 5 .385 1 .200 0 .000 4 .800 1.000
Wang NY 15 5 .333 2 .400 1 .200 2 .400 .667
Hochevar KC 22 5 .227 1 .200 3 .600 1 .200 .250
Blackburn Min 33 5 .152 3 .600 1 .200 1 .200 .750
Nippert Tex 6 4 .667 1 .250 1 .250 2 .500 .500
Tomko KC 10 4 .400 0 .000 4 1.000 0 .000 .000
Bonser Min 12 4 .333 1 .250 2 .500 1 .250 .333
Harden Oak 13 4 .308 2 .500 0 .000 2 .500 1.000
Harrison Tex 15 4 .267 4 1.000 0 .000 0 .000 1.000
Sowers Cle 22 4 .182 1 .250 1 .250 2 .500 .500
Morrow Sea 5 3 .600 2 .667 0 .000 1 .333 1.000
Westbrook Cle 5 3 .600 0 .000 2 .667 1 .333 .000
Masterson Bos 9 3 .333 1 .333 1 .333 1 .333 .500
Gallagher Oak 11 3 .273 0 .000 1 .333 2 .667 .000
Liriano Min 14 3 .214 3 1.000 0 .000 0 .000 1.000
Buchholz Bos 15 3 .200 1 .333 2 .667 0 .000 .333
Laffey Cle 16 3 .188 1 .333 1 .333 1 .333 .500
Mathis Tex 4 2 .500 0 .000 0 .000 2 1.000 .000
Hammel TB 5 2 .400 1 .500 0 .000 1 .500 1.000
Hurley Tex 5 2 .400 1 .500 0 .000 1 .500 1.000
Reyes Cle 6 2 .333 0 .000 0 .000 2 1.000 .000
Jennings Tex 6 2 .333 0 .000 1 .500 1 .500 .000
Duckworth KC 7 2 .286 1 .500 1 .500 0 .000 .500
Byrd Bos 8 2 .250 2 1.000 0 .000 0 .000 1.000
Braden Oak 10 2 .200 0 .000 2 1.000 0 .000 .000
Purcey Tor 12 2 .167 1 .500 1 .500 0 .000 .500
Ponson NY 15 2 .133 1 .500 1 .500 0 .000 .500
Masset Chi 1 1 1.000 1 1.000 0 .000 0 .000 1.000
Lewis Cle 4 1 .250 1 1.000 0 .000 0 .000 1.000
Loewen Bal 4 1 .250 0 .000 1 1.000 0 .000 .000
McCarthy Tex 5 1 .200 1 1.000 0 .000 0 .000 1.000
Parrish Tor 6 1 .167 0 .000 0 .000 1 1.000 .000
Gaudin Oak 6 1 .167 0 .000 1 1.000 0 .000 .000
Colon Bos 7 1 .143 1 1.000 0 .000 0 .000 1.000
Gonzalez Oak 7 1 .143 1 1.000 0 .000 0 .000 1.000
Hughes NY 8 1 .125 0 .000 0 .000 1 1.000 .000
Feierabend Sea 8 1 .125 0 .000 0 .000 1 1.000 .000
Trachsel Bal 8 1 .125 0 .000 1 1.000 0 .000 .000
Jackson Cle 9 1 .111 0 .000 1 1.000 0 .000 .000
Gabbard Tex 12 1 .083 0 .000 0 .000 1 1.000 .000
Byrd Cle 22 1 .045 1 1.000 0 .000 0 .000 1.000
I like the fact that Paul Byrd threw more than 100 pitches just 3 times in his 30 starts for Cleveland and Boston. Nick Blackburn of the Twins and Andy Sonnanstine of the Devil Fishies were the only other pitchers to make at least 30 starts and have fewer than ten 100+ pitch outings.
At the other extreme... just five AL hurlers threw at least 100 pitches in 80 percent of their starts. (No, I'm not counting Nick Masset.) Two of them - A.J. Burnett and C.C. Sabathia - are now working for the Yankees. (Sabathia's NL workload, by the way, almost exactly matches his AL pattern - 17 starts, 14 100+ outings, 10-1 with 3 NDs). This is interesting if only because no AL team was less likely than the Yankees to have a starter hang around long enough to throw 100 pitches. Andy Pettitte was the only Yankee who did this in even half of his 2008 starts. The others are Ervin Santana, Justin Verlander, and Cliff Lee.
