Cardinals news from a Sabermetric point of view

NLDS Game 4: Saving your best reliever for an inning that never comes

That cagey veteran Davey Johnson did it again. His late-inning reliever strategy helped his team win its second one-run game of the series. When a game is tied late — say, from the seventh inning on — the optimal strategy to the bullpen in our opinion is a simple best to worst. Certainly other factors may weigh in the decision — rest, health, matchups, etc. But in a sudden-death or near-sudden-death (not to be confused with near-death) game, such as a Game 4 or 5 in the LDS, the idea of “saving” your best pitchers, either for a later inning or later game, is unsound.

The reasoning is simple: Without knowing how long the game will go, you may be saving your pitcher for an inning  that never comes. This is akin to not sending your best pinch hitter to the plate in the ninth inning in order to use him at some other moment. The game isn’t going to get any more leveraged than it is at that moment.

Here’s the order in which the managers deployed their relievers Thursday:

Cardinals Nationals
Boggs Zimmerman
Lynn Clippard

And here’s each team’s bullpen in order of their fielding-independent pitching (as relievers) in 2012:

Cardinals FIP Nationals FIP
Shelby Miller 1.53 Drew Storen 2.40
Edward Mujica 2.34 Sean Burnett 2.79
Joe Kelly 2.84 Michael Gonzalez 2.98
Trevor Rosenthal 3.09 Tyler Clippard 3.31
Jason Motte 3.12 Craig Stammen 3.45
Mitchell Boggs 3.42 Jordan Zimmerman* 3.51
Lance Lynn# 3.49 Christian Garcia 3.73
Fernando Salas 3.59 Tom Gorzelanny 3.88
Marc Rzepczynski 4.72 Ryan Mattheus 4.42

*FIP as starter

#FIP as both starter and reliever

Now, FIP, as helpful as it is, isn’t a stand-in for true talent. It’s close, but first, these are the FIPs for only 2012. Second, reliever innings are by definition small samples are ought to be viewed with the appropriate skepticism (See Marc Rzepczynski’s 2011 FIP with the Cardinals). But Matheny’s decision to choose Lance Lynn over Jason Motte (or Shelby Miller or Edward Mujica…) is as symbolic as anything because it showed that he was planning for an inning that never came. One of Lynn’s attributes is his ability to pitch multiple innings, so it’s reasonable to assume that Matheny was going to have him pitch at least two innings (to wit: the pitcher’s spot in the order had made the final out of the ninth inning). And with the top of the order of the Nats due up, it wasn’t as though Matheny didn’t want to “waste” a top pitcher on bad hitters.

As another cagey veteran, Bernie Miklasz, wisely observes, “Games can be “saved” before the 9th inning.” The problem for the modern manager is that there’s no stat for them. And people criticize sabermetricians for obsessing over stats.

Other notes:

  • Got to hand it to Lohse: If that was his final game before free agency, he did well for himself. He may not be worth his next contract, but he earned his current one.
  • Run differential in the series: +14, Cardinals. That’s 3.5 per game. Anyone who wants to quote records in one-run games to claim that the Nationals are the better 2-2 team this series isn’t paying attention.
  • On the plus side: The only bunting that the Cardinals did was by their pitcher.
  • In his first bunt, Lohse made a half-hearted attempt to buy time by stopping in the baseline and forcing Adam LaRoche to tag him. But with the mad dasher on first, he may have had a real opportunity to stall long enough for another two-base advance.
  • The second one turned out painfully, because Lohse executed poorly. With Kozma on first with none out, Lohse forced him at second. The run-expectancy loss wasn’t huge (around 13%) but in a one-run game, it wasn’t insignificant.
  • Mike Matheny is learning. When Chad Tracy pinch hit in the eighth inning against Mitchell Boggs, the Cardinal manager left his righty in. Fool me once, shame on you. Fool me twice, shame on me. May that reasoning prevail again tonight if it’s tied late.
  • The good news is that the Cardinals have their postseason horse — Adam Wainwright — pitching in Gam 5. His regular and postseason performance:
    79 301 29 908 1073 3.31
    1 6 1 32 23 1/3 1.83

One Response to “NLDS Game 4: Saving your best reliever for an inning that never comes”

  1. games still require Says:

    games still require

    Fungoes » Blog Archive » NLDS Game 4: Saving your best reliever for an inning that never comes

Leave a Reply

You must be logged in to post a comment.