Last week, I set out to determine what player attributes or tools fans look for when they cast their ballots for the MLB All-Star Game. Raw measures of power, contact skill, pitch selectiveness, speed, ability to find holes, fielding, durability, presence on a winning team, and clutchness combined to explain about 50 percent of variation in the voting in 2011. But my original analysis suffered from a crucial flaw: It made the false assumption that fans' decision-making processes are consistent throughout their ballots.
Right or wrong, stereotypes about certain roles and roster spots have become fully ingrained in the game. The best bullpen arm should pitch the ninth inning. Leadoff hitters have to be scrappy and speedy, while the cleanup spot demands light-tower power. Sabermetric researchers and unorthodox-thinking managers have questioned the efficiency of some of these traditions, but they nonetheless continue to shape the way we—and professional coaches—think about baseball.
Similar stereotypes can be identified among players from different positions. First basemen and left fielders are your power guys. Second basemen and center fielders need wheels and good gloves. Catchers and shortstops are team leaders. And your DH's...well, obviously they'd better hit.
It therefore stands to reason that the way fans evaluate All-Star candidates varies by position. What, then, do fans look for as they consider different parts of the ballot?
I performed another series of correlation analyses between last year's All-Star votes (using my estimates for the missing values) and each candidate's numbers in every aforementioned category except clutch (I discounted it here because the original relationship was insignificant and actually slightly negative), but rather than looking at the player population at large I broke down the results by position. I then calculated the R-squared for each of the relationships (essentially, the amount of variation in the voting that can be explained by the skill in question).
Here are the results. Note that the negative signs indicate that the attribute in question had an inverse relationship with votes at that position (it is impossible for a real number to be negative when squared). Also consider that these numbers are all based on small sample sizes (all but outfield have at most 30 candidates, and there were only 14 at DH) from one year of voting, so take them with a grain of salt. (click to embiggen)
Playing on a good team remains the biggest factor in voting at every position but outfield; I would posit that outfield is the lone outlier because a disproportionate number of star players make their homes in the big green, but that seems to be countered by the fact that outfielders' votes are harder to explain across the board. No surprise to see that catchers' support is disproportionately influenced by the other 24 men on the candidates' rosters, though I would have expected it to be higher for shortstops—perhaps they are not seen as team leaders as much as they used to be.
Moving down the list, it makes sense that corner infielders get the biggest boosts in support from increased playing time. More plate appearances leads to gaudier counting stat totals in the kinds of categories that (unlike these ones) fans actually care about. Interesting that the effect of durability is diminished for catchers, who have the most grueling jobs in baseball. And despite the popular rhetoric that designated hitters have it easy, fans don't really punish those who don't play as much.
Outfielders, DH's, and first basemen are the expected top three in terms of power (if not in that order), but second basemen are right on their heels. Surprisingly, third basemen lag far behind, while catchers' support wasn't related to slugging ability at all.
Towards the bottom of the list, the relative proportions get interesting just because the baseline is so low. BABIP is meaningless for third basemen but reflects about a quarter of vote variation for backstops. Defense is a minor factor for infielders but are irrelevant everywhere else. Catchers, first basemen and DHs aren't usually known for their wheels, but they're the ones for whom votes were most based on speed. And fans pay so little mind to plate discipline that some of the relationships I found were actually negative.
For quick reference, I also relativized how much players at each position's support was affected by each factor on a scale like OPS+ where 100 equals the overall average. I hesitate to post this because the high relative variation in the lower-causation categories makes the numbers extreme, but here goes: (click to embiggen)
Overall, I'd say the results here don't really match up with what we would expect to see from positional stereotypes; the greater effect seems to be that fans pay more attention to specific skills in general for some positions (first and second base) than others (catcher and outfield). But a little extra pop really does seem to make more difference for a DH than for a catcher.