- FanGraphs Baseball - http://www.fangraphs.com/blogs -

How Many Runs Would Strasburg Have Saved?

The Washington Nationals face elimination from the playoffs today at home, down 2-1 in their NLD Series against the St. Louis Cardinals. Pitching for the Cardinals is Kyle Lohse, who is a fine, if not excellent, starter. The Nationals counter with left-hander Ross Detwiler — or, as he will likely be referred to more than once by TBS broadcasters Dick Stockton and Bob Brenly, “Not Stephen Strasburg.”

The reader probably hasn’t heard about it even once, so allow me to say: Stephen Strasburg is an excellent pitcher for the Nationals. He had Tommy John surgery towards the end of the 2010 season. He rehabbed for the majority of the 2011 season. Entering 2012, the Nationals suggested that they’d enforce some manner of innings limit with Strasburg — just as they had the previous year with other, young Tommy John-survivor Jordan Zimmermann. Then both Strasburg and the Nationals were really good — like, good enough, at one point, that the playoffs became a foregone conclusion (which is weird for the Nationals). Then people were like, “Are you really going to shut down Stephen Strasburg?!?” And then the Nationals were like, “Yes.” And then they did. In September.

It’s not necessarily the case that Stephen Strasburg would be pitching this game today, but it’s also the case that the only reason Ross Detwiler is pitching this afternoon is because Stephen Strasburg was shut down. Otherwise, the Nationals would have likely deployed a playoff rotation of Strasburg, Gio Gonzalez, Zimmermann, and Edwin Jackson — with Strasburg pitching a hypothetical fifth game.

What the people are certainly wondering — and which question Stockton and Brenly will certainly ask today — is “How much is Strasburg’s absence actually worth in terms of run prevention?” Or, alternatively: “How many runs would Strasburg have saved over his replacement(s)?”

The are a number of ways to answer that question, a number of variables to consider, but the simplest method is to begin by looking directly at Strasburg versus Detwiler.

Here are Stephen Strasburg’s and Ross Detwiler’s lines from this season, including only innings compiled as a starting pitcher:


Name GS IP IP/GS K% GB% BB% ERA- FIP- xFIP-
Stephen Strasburg 28 159.1 5.7 30.2% 44.2% 7.4% 80 73 72
Ross Detwiler 27 151.0 5.6 15.0% 52.3% 7.2% 91 101 109

While we don’t know how Strasburg would have pitched in Game One — and don’t know how Detwiler will pitch in Game Four, this afternoon — we do know that, on average, both pitchers threw about 5 and 2/3 innings per start. We also know that, when he pitched, Strasburg (on account of his various index figures) was the sort of pitcher to allow about 75% of the league-average number of runs. Detwiler, meanwhile, would allow runs (more or less) at a league-average rate.

Another thing we know is that the Cardinals scored 4.72 runs per game — presumably, against league-average pitching. That means, on average, we would expect them to score 2.97 of those runs over 5.2 innings. Against the league-average Detwiler, in other words, the Cardinals would likely score 2.97 runs. Against Strasburg, however, who allowed runs at 75% of the league-average rate, the Cardinals would score only 2.23 runs.

In this case — considering none of the other variable that a smart reader is definitely already considering — Strasburg would save 0.74 runs over Detwiler.

But wait, there’s more!

Because Strasburg would start Game One, he’d also start Game Five — i.e. a game that will now, if it’s played, be started by Gio Gonzalez. In order to calculate the impact of Strasburg’s absence, we also have to figure out how many runs (if any) Strasburg would have saved over Gonzalez.

Like before, we consider the pitchers’ lines side-by-side:


Name GS IP IP/GS K% GB% BB% ERA- FIP- xFIP-
Stephen Strasburg 28 159.1 5.7 30.2% 44.2% 7.4% 80 73 72
Gio Gonzalez 32 199.1 6.2 25.2% 48.2% 9.3% 73 73 87

In this case, the pitchers are much more evenly matched: Strasburg’s peripherals are better, but Gonzalez actually prevented real-life earned runs at a greater rate than him. Also, their FIPs are identical. Also, Gonzalez pitched more innings per start. The Cardinals would, very hypothetically, score about 3.32 runs in 6 and 1/3 innings — that is, Gonzalez’s average during the regular season. If we assume that Gonzalez also allows runs at 75% the league-average rate, then we would expect him only to concede 2.49 runs over that many innings. That’s slightly more runs than Strasburg, but also in slightly more innings. I feel comfortable calling it even. I hope you do, too.

Beyond this very basic look at run prevention, it is, of course, necessary to consider context: during a regular-season game, manager Davey Johnson will have let his starters get into trouble. In the playoffs, he is forced to have a significantly quicker hook. According to Baseball Reference, opposing batters had just a .610 OPS against Strasburg for their first plate appearance of a game. In the second plate appearance of a game, batters posted a .641 OPS. For the third, it was a .741 — that is above the National League average of .718. In a playoff situation, Davey Johnson is much more likely to call upon his bullpen — any of whom could likely hold the opposition to a below-average OPS for a short stretch — than let even the Stephen Strasburg face an opposing lineup a third time. Accordingly, Strasburg is unlikely to have saved even as much as the three-quarters of a run noted above.

In conclusion: does Strasburg’s absence make a difference? Yes, but one that is more likely to be overstated than not.