Steve Carlton's '72, King Felix's '10 and the Myth of Knowing How to Win

By Bill

I suspect that what we're seeing right now in the American League is something we'll see in one or the other of the leagues almost every September for the next mess o' years: one pitcher has the most wins, one has been the best at preventing runs, and there's a heated debate (among, increasingly, fans of the team the wins guy pitches for on the one hand and almost everybody else in the world on the other) over which one deserves the Cy Young Award.

You can guess which side I come down on in the Felix-vs.-C.C. debate, and it's been covered more than adequately elsewhere (including last week's podcast), so I won't discuss it here; I'll just heartily endorse this, and this, and this.

What I do want to address is one line of argument the pro-Sabathia crowd keeps bringing up in this debate. You see it mostly in angry comments to articles like the ones I just linked to, but it's probably phrased most ably by courant.com's Dom Amore (who doesn't ultimately pick C.C., but uses this argument to take Felix out of the equation right away; the bolded emphasis is mine):

Pitching isn't like a golf tournament where the lowest aggregate score wins, it's Match Play - the job of a starting pitcher is to match the other guy on each given time out. This is why a great pitcher can win a ton of games for a bad team, like Steve Carlton in 1972, while others, no matter how talented, seem to find their way to .500 no matter what kind of team they're on, such as A.J. Burnett.

Sounds good, right? Well, not to me, and probably not to you either since you've found your way to this blog, but you can see why it would sound good to a lot of pretty smart people. Most serious baseball fans get the allusion right away: in 1972, the Philadelphia Phillies won just 59 games, and Carlton was credited with the "win" in nearly half of those victories, going 27-10 in his 41 starts. If he can win with a terrible team like that, why is it unfair to expect Felix Hernandez to win with the similarly terrible Mariners?

Well, you've got your standard responses. Chief among them is probably that that's just one example out of 140 years of baseball, kind of the exception that proves the rule. But what I'm interested in knowing is: is this even a good single example? Did Carlton really rise above that much adversity to put up that pretty record? Because, you know, it's not just being on a bad team that sinks you. If you've got an okay-to-good offense and defense, but your staff is made up of one in-his-prime Steve Carlton and four Daniel Cabreras, the team might lose almost every game he doesn't start, but Lefty's still going to get his wins. So while Carlton obviously had a supremely great season in '72 -- putting up a 1.97 ERA in the 11th-most innings since 1920 -- I thought I'd try to see just how incredible and otherworldly that won-lost record was. 

And you can tell straight away that the 1972 Phillies were bad in every aspect of the game. Their pitching was below average even with the league's best pitcher handling 25% of their innings, but their offense was even worse; the team's 3.22 runs per game was second to last, just three hundredths ahead of the recent expansion Padres for last and seven tenths of a run per game behind the league average (3.91). 

But that doesn't tell the whole story, either. Just as it doesn't matter how the other starting pitchers on the team do in judging Carlton's ability to win, it also doesn't matter how the offense does in games in which Carlton doesn't pitch. And this where the Carlton-just-knew-how-to-win myth starts to run into trouble. According to baseball-reference.com, the Phillies scored 3.8 runs per game in which Carlton pitched, which is very close to league average. If you're being supported by a league-average offense, the fact that the team is terrible when you're not pitching doesn't matter much (unless you're going to argue that he's somehow responsible for getting more runs out of his offense, but if that's the case, I don't want to hear it -- though Carlton himself did put up a .501 OPS, much better than average for pitchers, just not nearly worth more than half a run per game).

Compare that to Felix's Mariners. The M's have scored 3.23 runs per game, almost exactly the same as the 1972 Phillies, but the differences are that that's easily the worst figure in this league (and in all of baseball), and the American League average (even in what's perceived as a down year for scoring) is 4.46 runs a game, more than half a run higher than the 1972 NL. And they've performed exactly that poorly for Hernandez, scoring 3.2 runs a game in his starts. 

In Carlton's 10 losses and 4 no-decisions in '72, the Phillies scored an average of exactly three runs a game (14 games, 42 runs). In Felix's 11 losses and 9 no-decisions, the Mariners have scored an average of 2.15 runs a game (20 games, 43 runs). And that's probably even less favorable to Felix than it looks, actually, because of that higher run-scoring environment he's toiling in to begin with. Felix also has a lower ERA in non-wins than Carlton, 3.47 to 3.93 (he has a better ERA in wins, too -- 0.79 to 1.12 -- but Carlton of course just has many, many more of those wins and pitched more innings in them).

Carlton gave up 0-2 runs in 25 starts (61%), 3-4 13 times (32%), and 5+ thrice (7%). Felix has given up 0-2 runs 17 times (55%), 3-4 eight times (26%), and 5+ five times (16%). There are some differences there, but not the kind that begin to explain the difference between 27-10 and 11-11.

Carlton's other big advantage, of course, is pitching in an era (the only such period in post-war baseball history) in which ace pitchers really were expected to "finish what they started." He started a bit more than one in every four games, and completed an unbelievable thirty of the 41 games he started--tied for the second most since World War II. Hernandez, by contrast, has left 14 games in which he had permitted 3 or fewer earned runs in 6 or more innings, and did not ultimately get the win. So as if it weren't bad enough that Felix suffered from much worse run support, especially relative to the league and era, than Carlton did, the strategies of the time dictate that he also has to rely much more on the Mariners' well-below-average bullpen. Carlton's ability to withstand that kind of workload is commendable and a huge part of the reason he's in the Hall of Fame today, but there's no chance that he would be allowed to stay out there nearly as long today (or that Felix would be pulled nearly so early were he pitching in 1972).

If you were able to pluck the 27 year old Steve Carlton out of 1972 and put him on the Mariners in 2010 (leaving aside any questions about how the average player's conditioning and abilities may have improved or deteriorated since then), he'd be pitching for a team that gave him fewer runs; he'd be pitching less often; and he'd be pulled from games early and more often, pitching in a world in which three complete games would be a lot more likely than thirty. That may make him even more effective on an inning-by-inning basis, since he'd be angling for 7 or 8 innings rather than 9-plus, but if the Mariner hitters were their usual futile selves in his starts, I don't see any reason to believe that Carlton would be sitting significantly better than 11-11 right now, and there's absolutely zero chance he'd be anywhere near 27 wins. And remember, we're talking about a guy who went 13-20, 16-13 and 15-14 over his three post-1972 seasons, and for much better Phillies teams than the 1972 edition; if it's true that Carlton just "knew how to win," he sure forgot again quickly.

So Carlton starts out looking like a good argument against giving a low-win, bad-team guy the Cy Young Award, but if you dig a little deeper, it's a good lesson about the uselessness of pitcher wins as a statistic. Carlton had an absolutely phenomenal season, but he got to 27 wins because (a) he pitched often and generally completed his games and (b) his team scored a pretty decent number of runs for him. It wasn't his ability to win, it was the situation he was placed in. Switch up his offense, environment and some luck, and he could have had exactly the same season, in effectiveness terms, and gone 32-5 or 17-20. The point, again, isn't to impugn Carlton's great season or to attempt to argue that Felix has been as good as Carlton was; the point is merely that putting any stock in the pitcher's win-loss "record" doesn't really do justice to either of them.

Match play in golf would be a very odd thing, and ultimately pretty uninteresting, if one competitor were playing his local public course and his opponent had to tackle St. Andrews, or if one played the course in July and the other in October. But that's pretty much what Dom Amore and others who think like him expect a pitcher to do when they say the pitcher's job is to "match the other guy." The thing is that when, say, Felix faces off against the Rangers, he's playing a different course than is the pitcher from the other dugout. A pro golfer playing St. Andrews could be much more on his game than the other guy playing the public course, and still finish with a 75 to his 65 (which is why golf has all those complicated handicap adjustments); in the same way, Felix could match the other guy in every way possible -- he could destroy the other guy, even -- and the team might well still lose, because the guys trying to score runs for Felix just don't pose nearly the same challenge to a pitcher as do the ones trying to score runs off of him. Those are the facts, and this is just common sense. And if you're going to try argue against common sense, you're going to have to come up with a better example than Steve Carlton (and a whole lot more of them).