Stealing Pays?

Carl Bialik at the Wall Street Journal who authors a column titled "The Numbers Guy" ran an interesting piece on his blog today related to the trend in stolen bases success rates.

Until he called my attention to it I hadn't realized that the success rate in 2007 was over 74% marking the highest it has ever been. In looking into the historical trend I found that three of the four previous highs occurred in 2004-2006 and I produced the graph he shows in the post. Because of its steady increase since the second World War I then ventured that it recorded a systemic change in the game perhaps related to increasing athleticism by baserunners coupled to a smaller degree with refinements in technique and strategy.

For those familiar with the strategic analysis of baserunning you'll notice that the current rate is dangerously close to the oft-repeated claim that 75% is the break even success rate for stolen bases in the big picture (it varies by base, out, and score of course). As a result, using the tenets of game theory, we would expect that as the actual success rate catches up with the break even rate, defenses would be a little more vigilant in protecting against the running game in order to keep the rate close to the break even. Whether they'll be able to do so, however, is dependant to a large degree on the constraints that are inherent in the game. For example, pitching out more often may indeed keep the rate down but at the cost of increasing the productivity of the current batter perhaps making it a wash. Balk rules also constrain attempts by the pitcher to keep the runner close. As a result, if the increase is primarily a result of increasing athleticism on the part of runners outstripping the combined throwing velocity of pitchers and catchers (and their technique), then we might indeed see the rate continue to climb.