Baseball Musings
Baseball Musings
August 10, 2006
The Value of a Steal

In the bottom of the first inning in today's Padres/Mets game, New York stole three bases and did not score a run. This underscores one of the offensive shortcomings of the stolen base. It's seldom valuable in isolation. Without a offensive help from following batters, stolen bases seldom score runs. It's another reason sabermetricians tend to value steals less than fans and the media, and another reason why some great pitchers don't care about base runners and concentrate on getting the batter out. If you get the batters, the thiefs tend to be stranded.

Update: The Mets score 2 in the bottom of the second. They didn't get a stolen base, but a bases clearing triple by Jose Reyes. That's 14 three-baggers for the shortstop.

Update: Reyes scores on a sac fly to make the score 3-2 Mets. Power once again is a much better way to move base runners and set up scoring.


Posted by David Pinto at 12:49 PM | Base Running | TrackBack (0)
Comments

"This underscores one of the offensive shortcomings of the stolen base. It's seldom valuable in isolation."

Nothing except a homerun is valuable by that criteria.

If the next batter doesn't get a hit, of course it doesn't matter if the baserunner steals or not. If the batter gets a base hit, that stolen base could equal a run.

Posted by: Robert Daeley at August 10, 2006 02:47 PM

A stolen base moves one runner one base. A hit or a walk can move multiple runners multiple bases.

Posted by: David Pinto at August 10, 2006 02:52 PM

The stolen base may have worked against the Cubs in the top of the 3rd. Pierre got on, Theriot (who?) got on behind him, and then they executed a double-steal to put runners on 2nd & 3rd with 1 out. With 1st now open, the Brewers walked Aramis Ramirez, and pitched to the less-imposing Phil Nevin & Jacque Jones, who both obligingly struck out swinging. If the Cubs don't steal, then Milwaukee probably pitches to Chicago's most dangerous hitter, instead of skipping him.

Posted by: Barron at August 10, 2006 03:02 PM


To me, it seems like not all stolen bases are created equal. Stolen bases from the 1st or 2nd batter I bet are worth more than thos from the 7th batter (due to the higher batting average of following hitters). Or maybe not, since these hitters slug more...

The stolen base from 1st base to 2nd base is probably better than from 2nd base to 3rd (the difference being scoring position for a single). And stolen bases with 0 or 1 out is probably more valuable than those that occur with two outs.

Why doesn't the stathead community differentiate (sp?) ?

Mike

Posted by: Mikee at August 10, 2006 03:08 PM

Your reasoning is specious.


What's done in the batter's box doesn't preclude the value of what's done on the basepaths, and conversely, a stolen base doesn't come at the expense of a hit... which is your inference by saying the stolen base has a shortcoming because it's not a hit. That doesn't make sense because the two plays are mutually exclusive... for the most part.

Granted, you won't score a lot without hits, but the stolen base enhances your chances of scoring should there be a hit. In a game where 65% failure is the norm, you have to expect the out rather than another hit.


Advancing the runner is good.
Advancing the runner with a hit is better.
Advancing an already advanced runner with a hit is best.

Posted by: bmc at August 10, 2006 03:56 PM

All of the above comments make sense, but the only way to value the SB is by looking at the whole picture, including success rates. If you are 100% successful, steal every time. Otherwise, you have to look at how much a steal increases your potential to score, discounted by the chance you will be thrown out. All things being equal, stealing only makes sense if you are better than 75% successful, otherwise you will decrease your scoring on average.

Posted by: Eric at August 12, 2006 06:33 PM

Mike: I believe the stats community has looked into it, with some research on this in baseball prospectus 2005, I think. I'm not sure that's where I saw it, but it was somewhere. They looked at the % success rate necessary to make the steal a good choice in every situation. For example, in Y year, player X on average chose good spots to steal and only needed a 60% success rate to justify the risk. Since he stole bases at an 80% rate, stealing when he did was a good play. Or, player Z didn't chose good spots to steal, needing an 85% success rate but only being safe on 70% of attempts. He cost his team runs.

Posted by: Ben B. at August 13, 2006 04:46 AM

The stealing percentage is irrelvant in this example because A) nobody go thrown out, and B)The Mets' success rate is about 88%.

Posted by: bmc at August 13, 2006 09:02 AM
Post a comment









Remember personal info?