Scott Podsednik, the Problems with OPS, and the Value of a Stolen Base

By Keith Glab 7/30/2005

There are currently nine players on the Chicago White Sox who have over 300 Plate Appearances. Scott Podsednik ranks 8^{th} among them in OPS with a paltry .722 mark. Although we sometimes rely heavily on OPS for a quick quantification of a player’s worth, this is one case where a player’s OPS does not correlate to his value to his team.

Firstly, here’s where Scott ranks among those nine White Sox in some other important categories:

BA: 1^{st}-.299

OBP: 1^{st}-.365

Hits: 1^{st}-100

2B: 1^{st}(T)-19

Runs: 2^{nd}-57

BB: 2^{nd}-36

BB/K: 1^{st}-.78

SB: 1^{st} (by 36)-51

Obviously, the problem with Podsednik’s OPS lies in his anemic .356 SLG. The problem with OPS itself is that it weighs SLG too heavily. Generally, players have considerably higher SLGs than OBPs, even though OBP is much more valuable (three times more valuable, according to a run production model of Paul DePodesta’s, though I don’t believe it’s that extreme). OBP is even more valuable for a leadoff hitter, and more valuable still for a leadoff hitter who projects to have eighty stolen bases on the year.

A high OBP is also very difficult for a speedy leadoff hitter to achieve. 14 players have over 20 SB this year, and among them only Ryan Freel and Bobby Abreu have OBPs higher than Podsednik’s. This is partly because the average player hits four points higher when he isn’t leading of an inning. In addition, where sluggers like Frank Thomas and Mark McGwire get intentionally walked and pitched around all the time, pitchers are trying their very best not to walk the likes of Scott Podsednik and Maury Wills. 604 of Barry Bonds’ 2302 walks have been intentional, while only 61 of Rickey Henderson’s 2190 were ordained by the opposition.

Another problem with OPS is that it does not factor in Stolen Bases, which is naturally the main part of Podsednik’s game. The value of a stolen base in the 20^{th} century averages out to be .2 runs if stolen bases are distributed randomly during the course of a game. (This means that a stolen base is worth less than half of what a single is and less than two thirds of what a walk is). Caught Stealings cost a team an average of .35 runs per event if they are indeed distributed randomly. So throughout history, if your SB% is less than 63.5%, you are costing your team runs.

However in 1980, statistician Dave Smith convinced Pete Palmer that the frequency of SBs and CSs is not random, but situation-dependant (i.e., more likely to occur in close games and important situations. Mitchel Lichtman refuted this claim while constructing his Super-LWTS formula, but I don’t believe him; if it’s 7-0 in the 8^{th} inning no one will try to steal a base, or if they do, Davey Lopes will issue death threats and official scorers will issue Defensive Indifference. No one tries to steal 3^{rd} with 2 outs, managers avoid stolen base attempts in front of their cleanup hitters, etc.). Palmer therefore bumped up the value of the Stolen Base to .3 and the Caught Stealing to -.6 in his Linear Weights Formula.

But I postulate that someone who steals lots of bases is worth much more than the sum of his SB’s linear weight values. With a Podsednik-like runner on base, hitters will get many good hitter’s counts (because of pitchouts) and lots of fastballs to hit (because pitchers are afraid that an offspeed pitch will allow the runner to steal uncontested). Marcus Giles’ OPS is over 100 points higher hitting in the #2 hole than out of the #3 hole, largely due to Rafael Furcal’s prowess on the basepaths.

A player who excels at stealing bases also has more value in the postseason, when the best pitching in baseball tends to curtail scoring runs by conventional means. In fact, many of the game’s best pitchers who frequent postseason play (Maddux, Johnson, Clemens, etc.) are awful at holding runners, and stolen bases are often the best way to get to them.

But getting back to the regular season and the quantifiable value of Podsednik’s stolen bases, his 51 SB and 14 CS have created eight runs above average using linear weights values of .3 and -.52, which is more than twice his overall linear weights value. Listing his OPS value isn’t statistically accounting for Scott’s main value on offense, let alone the intangible contributions of those stolen bases.

Among those nine regulars of the White Sox, Scott Podsednik ranks third in RC/27. I’m not saying that RC is the ultimate statistic… it still doesn’t account for the different roles played by a leadoff hitter and a run producer. But Scott Podsednik is a lot closer to being the 3^{rd} best offensive player in the Sox’ lineup than he is to being the 8^{th}.