Winning and Losing
I’ve written here before about how unusual a pitcher Nolan Ryan was. He was one of a kind in so many ways. Most strikeouts ever. Most walks ever. Most no-hitters, fewest hits per nine innings, most wild pitches (not counting Tony Mullane, who retired in 1894), most stolen bases against, most errors for any pitcher since Deadball and so on.
But in doodling around with some numbers, I came across a category I was SURE Nolan Ryan would lead. I was absolutely sure. But, as it turns out … well, you’ll see.
Let's begin here with this somewhat quirky but (in my mind) significant statistic: Since the Deadball Era ended, there have been 126 pitchers who started 200 games that their team won.
I know that's kind of a bizarre category -- starts that your team wins -- but, as many of you know, I believe this is how we should do the whole pitcher wins and losses dance. Forget the odd and illogical individual statistic and just make a team stat, the way we do with quarterbacks and goalies. The pitcher's won-loss record should be what his team's won-loss record was in his starts.
I suppose it's worth going into this again. I believe that the pitcher win, in addition to being a lousy way to judge a pitcher's ability, is an anachronism. It's from another age. It doesn't make any sense at all in a time when pitchers almost never complete games, in a time when teams use three, four, five pitchers to get through even well-pitched games. The pitcher win stat as it is now constituted is not a question of performance. It's a question of timing.
Go read the rule determining when a pitcher gets a win and when he doesn't -- Rule 9.17 if you're following along in your hymnal. It is 700 or so baffling words that offer all sorts of weird suggestions for the official scorer to choose who officially "won" the game. Determining which relief pitcher "won" the game is essentially like asking people to watch ants build an ant farm, then pointing at one and saying, "That one is the most valuable ant."
The pitcher win made some sense, as a general principle, back when starters completed a reasonably high percentage of their games. There was a baseline to work with.
It's always fun to put together a complete games chart. Here are percentages of starts that were completed::
1915: 55.1% complete games
1925: 49.1% complete games
1935: 44.5% complete games
1945: 46.3% complete games
1955: 30.3% complete games
1965: 22.8% complete games.
1975: 27.2% complete games
1985: 14.9% complete games
1995: 6.8% complete games
2005:3.9% complete games
2015: 1.4% complete games
If you only count nine-inning complete games, there have been nine this year. Nine. That is throughout major league baseball -- and Shelby Miller has thrown two of them
The complete game is not only rare, it is essentially gone as a sports concept like the drop kick and the two-hand set shot. And because we KNOW that the starter will not finish the game but we DON'T KNOW when the winning run will actually be scored, the whole pitcher win-loss thing is just a game of chance. Consider some of the stupid scenarios that are part of winning a game.
-- Starters have to pitch five innings to get a win, but relievers can win with one pitch.
-- Starters who throw an effective five innings CAN win a game, starters who throw an effective 4 2/3 innings cannot.*
*Unless it rains.
-- A relief pitcher who comes in with a three-run lead and blows it can get the win if his team scores winning runs in the bottom half of the inning.
And so on.
It's so unnecessarily baffling. And there's such a simple solution: Just give the starting pitcher the win and loss. Always. I just don’t think any of the quirky and unnecessarily complicated rules are relevant to the times now. A four-inning start is not much different from a five-inning start -- and I tend to believe that as bullpens become bigger we will get more four inning starts (and, perhaps, four-man rotations). I think the starting pitcher will ALWAYS be the pitcher of record. And we can invent much more interesting and useful statistics for relievers.
Now, you may ask (or you may not): Why not just get rid of pitcher wins and losses altogether? The whole idea is just kind of silly: Everyone knows that pitchers don't win and lose games by themselves ... and they never did. Everyone knows that some pitchers have gotten WAY too much credit simply because their teams happened to score a lot of runs for them. Why is it any more useful to know what the team record was when pitchers start?
Well, I guess my answer would go along like so: Even though starting pitchers don't win or lose games, they ARE the most important factor in a team's success that day. In that way, as mentioned, they are like quarterbacks or goaltenders. I think the reason the pitcher win has so stubbornly remained a part of baseball is because of that, because we know that starters do play a major role in victory or defeat, because we do want a bottom-line way to determine how successful they've been.
As far as why team wins are much more useful than pitcher wins, consider last year's Cy Young race between Corey Kluber and Felix Hernandez.
By pitcher wins, Corey Kluber went 18-9 and Felix Hernandez went 15-6.
Now, what does that mean? Seriously, what could you possibly learn from that? Did Kluber make more starts than Hernandez? No, they actually made the same number of starts.
Did Kluber’s Indians have more success when he pitched than Hernandez’s Mariners? No. Their teams actually had exactly the same record in their starts: 22-12.
Did Hernandez pitch fewer innings causing him to have fewer decisions? No, they pitched almost exactly the same number of innings (Hernandez actually pitched 1/3 of an inning more than Kluber).
Their won-loss records are not only meaningless, they are distracting and confusing and misleading. Think about it: If you were trying to explain the won-loss record to an alien or someone from Australia who had never seen a baseball game or my mother, how exactly could you explain Kluber’s 18-9 record? What does that even mean? You would explain that Felix Hernandez won 15 of 21 ... whats?
So now try it the other way: Let’s say that Kluber’s record was 22-12 and Hernandez’s record was also 22-12. Now, the alien or the Australian or my mother says; What does that mean?
And you say: That means their team won 22 of the 34 games they started.
That's it: Real information, right up front. Of course you can argue it's not all that useful information. You can come up with a million scenarios where it’s silly to give the starting pitcher a win or a loss: Say a starter who throws one pitch, hurts his arm and comes out or a starter who leaves the game with a five-run lead that the bullpen blows. Sometimes the pitcher will not DESERVE a win or loss.
But, as Eastwood said, deserve's got nothing to do with it. We're now getting into what I call judgmental baseball stats. We all know pitchers don’t win or lose baseball games by themselves … and we all know that sometimes they will get hosed or they will get unfairly rewarded. But that's the point anyway. If we're going to give them wins and losses anyway (and we are) let's stop pretending that it makes sense and just give the starting pitcher his real record.
Whew, that was a longer preamble than I expected. Anyway, from here forward in this post, when I talk about a pitcher winning I’m actually talking about the team winning a game the pitcher started. I understand this might be confusing and irritate some but, hey, you will notice I at least did not call the Cleveland Indians the “Spiders.”
OK, since 1920 here are the all-time leaders in wins:
1. Roger Clemens, 433
2. Greg Maddux, 431
3. Don Sutton, 418
4. Nolan Ryan, 405
5. Steve Carlton, 404
6. Warren Spahn, 401
Those are those six pitchers whose teams won 400 of their starts. Now, here are the Top five winning percentages (remember, this is among pitchers whose teams have won 200 or more of their starts):
1. Whitey Ford, 297-139, .681.
2. Lefty Grove, 304-150, .670
3. Lefty Gomez, 212-108, .663
4. Sandy Koufax, 206-108, .656
5. Bob Lemon, .220-128, .624
You will probably notice that the top four were all left-handed pitchers. Just kind of interesting.
The lowest winning percentages? Well, as it turns out, eight pitchers who won at least 200 wins also had losing records.
8. Charlie Hough: 219-221
7. A.J. Burnett: 205-207
6. Jeff Suppan: 207-210
5. Bob Friend: 240-254
4. Rick Wise: 221-234
3. Bobo Newsom: 233-247
2. Livan Hernandez: 225-249
1. Mike Moore: 207-233
An aside: Frank Tanana — this is fitting — won exactly as many games as he lost: 308-308.
Now, we’re getting to the fun Nolan Ryan stuff (or, anyway, I think it’s fun). As you might expect, pitchers' ERA's are a much better — typically about three runs better — when they win than when they lose. That just goes without saying, right? Some days you have it, some days you don’t. When you pitch well, you usually win, and when you don't pitch as well you usually lose.
My expectation was that nobody, absolutely nobody, better fit this description than Nolan Ryan. When Ryan was on, forget it. No chance. And when he was off ... yuck.
Ryan had a lot of off days. His teams lost THREE HUNDRED SIXTY EIGHT times when he took the mound. That is thirty more losses than any other pitcher since Deadball. So my guess was that Ryan HAD to have the biggest difference between when he was on and when he was off in the history of baseball. And that is sort of true.
When Nolan Ryan’s team won, his ERA was 1.78. He gave up — and this seems ludicrous — 1,821 hits in 3,099 innings: That’s 5.2 hits per nine innings. He also allowed just 109 home runs. To give you an idea, that would be like giving up seven home runs in a 200-inning season.
When Ryan’s team lost, however, his ERA was 5.14, his WHIP was an unseemly 1.54 and he gave up 205 homers in 2,227 innings. Teams stole 442 bases against him (nobody seemed to care less about holding on base runners). He still struck out a lot of guys, and he still gave up fewer than a hit per inning (though a much more human 8.3 per nine innings). But the numbers show the Nolan Ryan I have grown to love: On his good days, you couldn’t beat him. On his off-days, he'd make so many mistakes that you could.
So, what is the category that I expected Ryan to lead but he doesn’t? Well, here are the Top 5 ERAs for pitchers on their winning days:
5. Bob Gibson, 1.86
4. Sandy Koufax, 1.86
3. Don Drysdale, 1.81
2. Nolan Ryan, 1.78
1. ??????????, 1.77
What? There was a pitcher with a lower ERA in victories than Nolan Ryan? I just wasn't expecting that at all. Who could that possibly be?
First guess, no doubt, Pedro Martinez. But it isn't him. You have to remember: Martinez played in such a huge offensive time, his ERA could only be so low. As it turns out, his ERA when his team won was 1.91, eighth on the list.
All my next guesses would have been wrong too: I would have guessed Tom Seaver (1.88), Greg Maddux (2.01) or Roger Clemens (2.04). I would have guessed Randy Johnson (2.26) or Bert Blyleven (1.98) or Whitey Ford (2.00). I would have kept guessing.
Jim Palmer had an interesting career — his teams went 325-196, and he had a 1.90 ERA when his teams won and a 4.62 ERA when his teams lost. Did you know, as a total aside, that Jim Palmer never struck out 200 batters in a season?
The best career ERA in losses, by the way, belongs to Seaver. His teams lost 278 games, but he had a 4.44 ERA in those losses. In other words, his record probably should have been better -- Seaver pitched on a quite a few low-scoring teams.
The worst ERA in losses? No surprise: It’s Tim Wakefield. When that knuckler wasn’t knuckling, look out — he had a 6.55 ERA in his losses.
OK, enough suspense? Who had the lowest ERA when his team won the game?
Answer: Gaylord Perry. Yeah, in Perry’s 360 wins, he had a 1.77 ERA.
In his 329 losses, he had a 4.96 ERA.
Perry really had quite a remarkable career; underrated, I think, when telling the grand story of baseball. He's become known in history simply as the spitball guy ("A hard slider" as his daughter called it) but I tend to think of him as the pitcher who came up more ways to get hitters out than anyone since Satchel Paige. He would throw spitballs, puffballs, any kind of off-speed junk to win. And he spent most of his career on lousy teams. That -- the lousy teams part -- sums up WHY I think Perry had the lowest ERA among winning pitchers. He didn't have any choice.