Earlier this month I introduced the idea of a Probabilistic Model of Steals. In that post, I looked at how the stage of the game, represented by the inning, changed the probability of attempting a steal, and the success rate on those attempts. Subsequent posts examined outs, the score difference and the combination of all three. This posts looks at running against catchers, comparing actual stolen base attempts against the expected number of attempts based on the model.

To review, I’m only concerned with the pure steal situation, only a runner on first. I define score difference from the point of view of the offensive player (Offensive score – defensive score). In looking at the data, a lead of seven runs in either direction seemed to be the point where teams really stopped running, so any lead of seven runs or greater is either placed in the 7 or -7 bin. I also put all extra innings in the 10 bin for that category, since each extra inning starts the same, with the score tied. The data is from 1996 through 2008. The following table displays all catchers with at least 1500 steal situations against them defined by the three parameters. The catchers are ranked by the ratio of actual steals to expected steals, times 100. A ratio of 100 means that runners stole exactly as often as expected against the catcher. A ratio over 100 means runners attempt steals more often than expected, under 100, less often. There are 105 catchers in the study:

With Yadier Molina injured this Wednesday afternoon, it seemed like a good time to look at this data. Molina is pretty amazing. Not only does he completely shut down the running game, the few who try to run against him make it less than 50% of the time. Note that his potential replacement, Jason LaRue, does a great job at stopping the running game as well.

Note also that Ivan Rodriguez and Benito Santiago earned their reputations as great arms behind the plate. Joe Mauer joins Yadier Molina as the next generation of outstanding throwers.

At the other end, Mike Piazza certainly earned his reputation as someone who was challenged at stopping the running game, and Scott Hatteberg needed to become a first baseman for more than just the injury.

The real interesting catchers to me are Yadier’s brother Jose, Wiki Gonzalez, and Joe Girardi. They all posted very low stolen base percentages against, yet that didn’t stop runners from trying to steal against them. There appears to be more than the catcher’s arm at work here.

The next post concentrates combining pitchers and catchers. As always, I’m interested in your feedback. You can follow the series here.

Please consider supporting this work with a donation to the Baseball Musings Pledge Drive.

The information used here was obtained free of charge from and is copyrighted by Retrosheet. Interested parties may contact Retrosheet at 20 Sunset Rd., Newark, DE 19711.

Last week I introduced the idea of a Probabilistic Model of Steals. In that post, I looked at how the stage of the game, represented by the inning, changed the probability of attempting a steal, and the success rate on those attempts. Subsequent posts examined outs, the score difference and the combination of all three. This posts looks at running against pitchers, comparing actual stolen base attempts against the expected number of attempts based on the model.

To review, I’m only concerned with the pure steal situation, only a runner on first. I define score difference from the point of view of the offensive player (Offensive score – defensive score). In looking at the data, a lead of seven runs in either direction seemed to be the point where teams really stopped running, so any lead of seven runs or greater is either placed in the 7 or -7 bin. I also put all extra innings in the 10 bin for that category, since each extra inning starts the same, with the score tied. The data is from 1996 through 2008. The following table displays all pitchers with at least 1000 steal situations against them defined by the three parameters. The pitchers are ranked by the ratio of actual steals to expected steals, times 100. A ratio of 100 means that runners stole exactly as often as expected against the pitcher. A ratio over 100 means runners attempt steals more often than expected, under 100, less often. There are 107 pitchers in the study:

At the top of the list are a number of pitchers one might expect, left-handers with good moves to first base like Mulholland and Rogers. I’m a bit surprised Andy Pettitte doesn’t rank higher, but I think it’s a combination of two things. Pettitte pitched in a lot more stolen base situations than many of the others at the top of the list, so the larger sample size might bring him back closer to average. The bigger factor, however, may be his high number of pickoffs. Most pickoffs are actually scored as caught stealings, since the runner often heads toward second base hoping for an error. Note the low success of runners against Pettitte.

Chris Carpenter is simply amazing at not only stopping the running game, but also preventing a stolen bag once a runner commits. Part of that is having Yadier Molina as a catcher, but Carpenter’s only had Molina behind the plate for part of his career. We’ll look into that relationship in more detail in a later post. I’m hoping this research eventually leads to a way to distinguish between catcher defense of the steal and pitcher defense of the steal.

At the other end of the scale, where runners attempt steals more often than expected against pitchers, lie three of the greatest hurlers of the period, Roger Clemens, Greg Maddux and Randy Johnson. These three appeared to pitch to a philosophy that the stolen base didn’t matter. They paid little attention to base runners, concentrating on getting the out at the plate. If the hitter makes an out, the chance of the runner scoring drops close to zero.

I’m not surprised the two easiest pitchers to steal on were knuckleballer Tim Wakefield and windup artist Hideo Nomo. The slow speed of Wakefield’s pitches combined with the difficulty of catching them makes Tim an easy target. Nomo’s back arch delivery allowed runners to get an extra step as well.

The next post concentrates on the catchers. As always, I’m interested in your feedback. You can follow the series here.

Please consider supporting this work with a donation to the Baseball Musings Pledge Drive.

The information used here was obtained free of charge from and is copyrighted by Retrosheet. Interested parties may contact Retrosheet at 20 Sunset Rd., Newark, DE 19711.

Earlier today I introduced the idea of a Probabilistic Model of Steals. In that post, I looked at how the stage of the game, represented by the inning, changed the probability of attempting a steal, and the success rate on those attempts. This posts examines the second parameter of interest, outs.

To review, I’m only concerned with the pure steal situation, only a runner on first. The following table shows the probability of attempting a steal based on the number of outs, and the success rate for each number of outs:

The farther along in an inning in terms of outs, the more likely a runner will steal. This makes sense, as getting a runner in scoring position with two out means he’ll likely score on a single. Note two that runners succeed more in this situation as well, as getting the batter becomes the primary task of the defense. Even if the runner reaches second base, the offensive team only has one out to spend to get him home. It appears teams are much more willing to defend against a steal in a one-out situation, as shown by the difference in success rates.

Teams seem perfectly willing to wait to see if a situation develops with a man on first and no one out. The idea that a runner caught stealing harms the chances for a big inning are reflected in this number. Note how the SBPct lies between the other two outs situations. Teams don’t want to run themselves out of a big inning, so they don’t run as often and they try to run when they have a better chance of success.

As always, I’m interested in your feedback. The next installment will look at the score difference.

Please consider supporting this work with a donation to the Baseball Musings Pledge Drive.

The information used here was obtained free of charge from and is copyrighted by Retrosheet. Interested parties may contact Retrosheet at 20 Sunset Rd., Newark, DE 19711.