clock menu more-arrow no yes mobile

Filed under:

Worth the Risk?

Reading Clack’s article on base running got me to thinking about Run Expectancy. Specifically, Run Expectancy in relation the stolen base. After last season, it was clear that Ed Wade wanted to increase the overall team speed of the Astros. Trading for Michael Bourn and signing Kaz Matsui were the two most notable means of accomplishing this goal. Did it end up working, is the question that rattled in my brain.



Tom Tango is a sabermetrician who has taken data form thousands upon thousands of major league games, and come up with specific run expectancy values based on specific in game situations. The weight given to a stolen base is valued at .19 of a run, because it moves the runner over one base, often into scoring position. In other words, a team’s run expectancy goes up nearly 1/5 of a run when a stolen base attempt is successful. However, a caught stealing reduces run expectations by .44 of a run. Why does a caught stealing so extremely limit the amount of runs a team can score in a given inning? Well, it has an inning-killing effect of -.16, as well as a "moving the runner over effect" of -.02. The remaining -.26 exists because by a player recording an out by being caught stealing, the next few batters chances of getting on base are inherently smaller, due to the decreased number of outs available.

Now that we've gone over the specific run expectancies that go into a stolen base/caught stealing, we can examine the ratio for the 2008 Astros, and see just how many runs have been created/been reduced by our boys being given the green light.

Successful Stolen Base Attempts 90 Run(s) Gained *(.19) = 15.75 Runs
Caught Stealing 37 Run(s) Lost *(-.44) = (-17.279) Runs
Totals 71% SB% Total Runs Lost = 1.529

What does this mean?

Basically, the Astros have cost themselves a little over one run this season in their attempts to steal bases.

Their success rate of 71% is a tad over the Major League Average in 2003 (the most recent year's data I could find) of 69% (Source: The Book on The Book, Bill Felber).

Going back to our friend Tom Tango, the Linear Weight associated with a runner on first and two outs (the most common steal situation), is .239. By completing a successful stolen base attempt, your run expectancy value has increased to .347, a whopping .108 of a run. If the runner is caught, then with two outs the Run Expectancy falls to zero, a net loss of .239.

The Astros may have gained some pure team speed, and their stolen base totals may have gone up some, but is their stolen base percentage (SB%) any higher than the team that was ranked last in team base-running by Mr. James? The 2007 Houston Astros had a SB% of 67%. So, yes, Ed Wade succeeded in upping the team's proficiency at stealing bases.

Just for fun, let's break down the 2007 Astros:

Successful Stolen Base Attempts 64 Run(s) Gained *(.19) = 12.16 Runs
Caught Stealing 31 Run(s) Lost *(-.44) = -13.64
Totals 67% SB% Total Runs Lost = 1.48

It may just be in tenths of a run, but the lead-footed 2007 squad actually cost themselves less runs, by virtue of simply not running all that much. What we can see from this is that, those players that are proficient base-stealers should steal with greater frequency, but those that are not should stick closer the bag. The situations described by Clack in his article (going first to third on a single, scoring from first on a double) in addition to defensive plays could tip the balance in favor of the moves made by the Astros front office in the offseason. Judging purely on stolen bases though, they did not succeed, at least not 70 percent of the way through the season.