The day has arrived. Rather than stress out about the game too early, let’s have some fun with another preview. One of the smaller, but more interesting battles during the NLDS will be that between Cardinals catcher Yadier Molina behind the plate and Dee Gordon on the bases. We all saw the kind of difference a running game can make in a playoff game earlier this week. However, Yadier Molina is no Derek Norris, who was catching for the A’s while the Royals were running wild during the AL play-in game.
Molina has thrown out 45% of the baserunners who have tried to run on him through his career, well above the league average caught stealing rate of 28%. He has also faced off against Dee Gordon a few times in the past, so let’s take a look back:
Attempt 1: 9/13/12
Gordon entered the game as a pinch runner after Luis Cruz reached on an error. The pitcher, Edward Mujica, threw a splitter which is not necessarily an easy pitch for a caught stealing. However, Molina threw an absolute dart to second. It took a perfect throw to get Gordon, and the throw was perfect. Relievers don’t generally do a great job of holding runners on, but Mujica has been relatively good at limiting the running game, even while away from the Cardinals. Given Gordon’s jump, Molina deserves most of the credit.
Attempt 2: 9/15/12
This was another pinch running appearance for Gordon, this time with Jason Motte on the mound. As the GIF shows, the pitch was 95, and in a decent place for Molina to make a throw. Gordon was ruled safe, but…
Good thing for the Dodgers that this was before replay review was implemented, because was clearly out. The missed call led to a Dodger victory on a Juan Rivera walk-off single two batters later. Hah, Juan Rivera. Remember him?
Motte has allowed a league average stolen base success rate while mostly pitching in front of Molina, so he isn’t exactly great at holding runners on. It didn’t change the result here.
Attempt 3: 9/16/12
Trevor Rosenthal has pitched nearly 170 innings at the big league level. This Dee Gordon caught stealing (again, on a late innings pinch run appearance) was one of six times anybody has even attempted to steal against Rosenthal. This was Molina’s worst throw of the four attempts, but Gordon’s jump wasn’t great and Daniel Descalso deserves some credit for a fast tag.
Attempt 4: 6/29/14
Due to a combination of Molina injuries and a lack of playing time, the next rematch wasn’t until earlier this year. This time, the pitcher was Shelby Miller, currently slated to start game four of this series. Like Motte, Miller allows about a league average stolen base success rate, even with Molina behind the dish for most of that span. Gordon got a great jump, but the Cardinals pitched out. Gordon was so fast that Molina’s great throw barely got him.
Overall, Dee Gordon has attempted four stolen bases against Yadier Molina in 15 “stolen base chances” (on base with second or third base open in front of him). That’s a 27% attempt rate, which is a bit lower than Gordon’s career attempt rate of 35%. Already, Molina has limited Gordon’s running game before using his arm.
Of the four attempts, Dee was caught three times. He was “safe” (actually out) once. That’s a 25% (0%) success rate, compared to Gordon’s career success rate of 77%. If we go by what should have been called, Dee has been out all four times he has attempted to steal against Molina, which is pretty incredible.
In order to maintain a positive contribution to overall offensive output, the break-even point for stealing bases (without including leverage) is a 2/3 success rate (as in, if a runner steals 66.6% successfully, his net stolen base runs output will be zero). Gordon’s past four attempts against Molina aren’t really a predictor of his future success rate, which will probably be closer to Molina’s 48% caught stealing rate (despite Gordon’s speed, he still steals around the league average rate, even this year). With a large enough sample, it’s tough to imagine Gordon hitting the 2/3 threshold. Basically, if Gordon doesn’t have a perfect read on a pitcher, he should not go. Depending on Molina to make a mistake would be, well, a mistake.