Sometimes I like to get straight to the point on my posts, so I’ll try that here—I don’t know.
More to the point, though, there is one fairly important point to be made about the St. Louis Cardinals and their rotation for the upcoming Wild Card Round, a best-of-three series against the San Diego Padres, which is that for the sake of this round, it doesn’t really matter in what order the three pitchers go, so long as the team goes with its three best starting pitchers. No starter is going to go twice, and if a team were to save its best starter for Game 3 and they were eliminated in two games, this may impact the length of the series, but it’s probably safe to say the team was going to lose the series anyway, barring arguments about game-to-game momentum that I do not personally believe hold much, if any, weight.
But if you work under the assumption, as I do, that the primary objective of the early rounds is to set yourself up to win later rounds, saddling your best pitcher with a Game 3 start which would prohibit him from being able to maximize the number of starts he can make going forward in the postseason doesn’t make a whole lot of sense.
The St. Louis Cardinals didn’t really have much in the way of choices for the 2020 Wild Card Round—there isn’t too much of an argument regarding which three starters should pitch. With Dakota Hudson slated for Tommy John surgery and Carlos Martinez also unavailable early in the postseason, unless you want to try to make a case for Austin Gomber (who, to be fair, has pitched decently in his four 2020 starts), the top three are, in some order, Kwang-hyun Kim, Adam Wainwright, and Jack Flaherty. The Cardinals have announced that Kim will start Game 1, Wainwright (who was unavailable for Game 1 unless he was to start on short rest) will start Game 2, and Flaherty will start Game 3, if necessary. The Cardinals, of course, hope that Flaherty will not be needed, but in the good way.
By ERA, there is actually a pretty pronounced gap among the three starters. Kwang-hyun Kim had an elite 1.62 ERA; Adam Wainwright had a respectable 3.15 ERA; Jack Flaherty had a “are we sure this guy should be a starter?” 4.91 ERA. But the problem with using ERA as the basis for making starting pitching decisions in 2020 goes beyond the typical “pitching results are so defensively dependent that ERA alone is misleading” caveat—we are basing these ERA results on shockingly low innings totals. Given the nature of how the team’s sixty games were clustered, ripe with seven-inning games and double-headers, these guys simply didn’t pitch very much in 2020. If you pro-rated Wainwright’s innings total to a 162-game pace, he’d be at 177 1/3 innings, which is the only semi-representative sample among the three. Flaherty would be at 109 innings and Kim would be at 105 1/3. But even these samples are generous because, well, they didn’t pitch this many innings. These three starters combined for 145 innings.
By defensively-independent pitching statistics, there are arguments to still be made for Kwang-hyun Kim as the Game 1 starter, but they become far less conclusive. By FIP, the most common of the defense-independent statistics, he “deserved” a ERA over two runs higher than we actually got. His strikeout rate of 5.54 K/9 is lower than that of Wainwright and far lower than that of Flaherty, while batters were extraordinarily unlucky with regards to balls hit in play—his opponent BABIP was .217.
Now, it is fair to note that Kim is a crafty type of pitcher who may truly be capable of inducing soft contact—a .217 BABIP, however, is so far removed from what any other pitcher in history has done that it isn’t even worth discussing if he is this good at it. Also, even with his much-inflated FIP, Kim still did have a lower FIP than Wainwright or Flaherty—Kim’s stood at 3.88 while Wainwright and Flaherty each posted a not-that-much-higher-but-still-by-definition-higher 4.11 (not that you should take this gap as even remotely material, but Wainwright was incrementally superior if you don’t round to the nearest hundredth). But Kim also had an unusually low home run to fly ball rate of 9.1%. This isn’t as cartoonish as, say, the rate of Chicago White Sox starter Dallas Keuchel, whose HR/FB% was a comically low 4.7%, but it is enough that, if you were to instead grant Kim a league-average HR/FB%, his FIP would stand at 4.52. If you granted the same treatment to Wainwright (whose 13.8% is still low, but much less extremely so), he’d be at 4.23. And Jack Flaherty, assumed by everybody involved to be far and away the team’s ace entering the 2020 season, would be at 3.42.
Now, a 3.42 ERA would still be a little underwhelming for Jack Flaherty based on his 2019 performance, but it is also very clearly the best of the three. By SIERA, a statistic similar to FIP but with more batted ball consideration, Flaherty is a half-run better than Wainwright and over a full run better than Kim. Unless you truly believe that Flaherty’s 23.1% home run to fly ball ratio is reflective of some new and disturbing trend in his game (he was at 13.8% last year, 15.2% in his previous full season, and only Kyle Gibson had a worse rate among qualified starters in 2020), there is a ton of evidence to suggest that Flaherty was consistently unlucky in 2020.
But even if you believe that ERA is the be-all-end-all of pitching statistics (I tend to believe that large swaths of sabermetrically-inclined analysts overly dismiss it; I’m more on your side than you might think I am), how would a 3.13 ERA for Jack Flaherty sound? This is the mark if you were to remove one particularly disastrous start, the one from September 15 against the Milwaukee Brewers. Is this statistically irresponsible cherry-picking? Well, yes and no. Yes, you should consider all of Flaherty’s starts, including the uncharacteristically poor ones. But also, if one start results in a 1.78 run increase to his ERA, doesn’t this say something about how fickle the statistics are with this small of a sample size?
Let’s expand the sample size to include last season, where Flaherty was rather lucky by batted-ball metrics. His ERA in those 236 2/3 innings was 3.12. Adam Wainwright’s over those two seasons was 3.91. Even if you considered this further cherry-picking of data, it should be more than a bit revealing that it can be done so easily.
If the Cardinals lose two games against the San Diego Padres and Jack Flaherty never takes the mound, I’m not going to blame this decision for that—maybe it’s the difference between a sweep and a three-game Padres victory, but at the end of the day, I don’t care that much about series length. But if the Cardinals advance to the NLDS following a three-game series and Jack Flaherty is only able to appear in one NLDS game, rather than potentially making two starts (the latter on three days rest) or one start plus high-leverage bullpen appearances, that could be rather noticeable. And it will be frustrating. And it will be completely avoidable.