Tuesday night offered baseball fans a rare opportunity: not only the chance to see a rematch of last year's World Series participants, the Philadelphia Phillies and New York Yankees, but also the chance to see each team send its respective ace to the Yankee Stadium mound, Philly's Roy Halladay and New York's CC Sabathia.
Unfortunately, no pitchers' duel materialized. Halladay conceded three home runs, Sabathia wasn't exactly at his sharpest (walking three in seven innings) and the Yankees won by a distinctly unduelish score of 8-3.
Meanwhile, in a less publicized (and considerably less attended) affair, C.J. Wilson of the Texas Rangers and Josh Johnson of the Florida Marlins gave us the game we might have expected from Halladay and Sabathia, allowing only six hits and three runs between them over 13 collective innings.
Yet, despite the cosmetic difference in run total (11 on the one hand, five on the other), these two games help demonstrate that simple runs-allowed numbers are hardly the best way to determine whether a pitcher has truly "shut down" the opposition.
More on that in a second. But first, let's consider the Wilson-Johnson matchup.
Again, in terms of superficial returns, we see Wilson allowed two earned runs and Johnson allowed only a single earned run. But even a casual glance at the box score reveals that while Johnson struck out seven and allowed only one walk, Wilson struck out six but also walked six. Intuitively, we understand that Johnson controlled the opposition's batters better than Wilson. The question is: How much better?
Luckily, we can find out. Graham MacAree of StatCorner has done work that gives us the expected run values for every event within a pitcher's control. Those events and their respective run values are as follows. (Note: In the version below, the expected run value for home runs has been integrated into the outfield fly ball run value according to the principle that home runs occur on approximately 11 percent of outfield flies.)
Run Value For Each Event
Graham MacAree of StatCorner calculated the expected run values for every event within a pitcher's control.
Of course, it's not as if every time a pitcher records a strikeout, it takes 0.105 runs from the other team's score. Anyone who's watched a game knows that striking out the opposing pitcher with two outs in the bottom of the third is a lot different than striking out the other team's cleanup hitter with the bases loaded, no outs, etc. Still, these events are generally the things over which a pitcher has control, and all of them stabilize pretty quickly.
So what happens if we look at the Wilson-Johnson game in the context of expected runs? This:
Expected Runs for Wilson and Johnson
This allows us to calculate exactly how effective these pitchers were.
Here, we see the degree to which Wilson's walks penalized him -- to the tune of roughly two runs. All told, we should have expected Wilson to allow three runs over his six innings pitched. That's not a huge difference from the two he actually allowed, but it's still noteworthy.
Now here's what happens if we do the same thing for the Halladay-Sabathia game:
Expected Runs for Halladay and Sabathia
In this game, luck wasn't with these pitchers.
Two notes here. First, look at Halladay's expected runs: a hair under four. Why so much lower than the six he actually gave up? Because Halladay allowed three homers, but he did so on only eight balls to the outfield. Again, these expected run totals don't take into account Halladay's opposition (in this case, the heavily armed Yankees), but still: Three home runs on eight fly balls is bad luck any way you slice it.
The careful reader will note a second something as well: Although he allowed more actual runs than Wilson (three to two), Sabathia conceded fewer expected runs. And it makes sense, too. Just look at Sabathia's line compared to Wilson's. More strikeouts? Check. Fewer walks? Check. More grounders and fewer flies? Double-check. Sabathia controlled the game better than Wilson, even if the results don't reveal such a thing.
In a season that has seen two perfect games and a should-have-been perfecto, it's important to recognize that sometimes luck isn't on a pitcher's side. On Tuesday, Wilson benefited from luck. Halladay? Not so much.
Carson Cistulli is a writer for FanGraphs.