http://www.baseballprospectus.com/article.php?articleid=2607 I understand the point, but without the threat of the stolen base some of those disruptions by a runner on first do not exist, such as the first baseman could simply play back. That suggests a team should steal just enough to keep the threat very much alive in the opponents heads, but no more. Do they factor errors into the equation when calculating the percentage though, if it be the pitcher mis-throwing the pickoff throw or the catcher throwing the ball into center field? Oh sure, if there were no threat of a stolen base, the offense would lose some advantages of having a runner at first. And there are definitely situations in which attempting a steal is well worth the risk. Dan Fox concludes from his work that the stolen base shouldn't be a "general purpose weapon," but there are situations in which it makes sense as a strategy. In other words, giving most guys a permanent green light is a bad idea, but calling for steals judiciously can help a team score runs. Fox's formula takes number of outs, occupied bases, and the base that the runner is trying to steal into account. For example, the risk/reward ratio of stealing second is far better when there are two outs and no one else on base than when there are no outs and runners on first and third. Alfonso Soriano was the best at producing runs through stolen bases in 2005, but he was among the worst in 2006. Lou should only give him the green light in low-risk, high-reward situations. If he does so, Soriano will probably be an asset. AFAIK, errors aren't currently factored in to Fox's formula, but they might be in the future. It won't make a very big difference, but it would make some difference. For example, his numbers show about 4% of runners scoring after wild throws by catchers to third.