This is a question I have asked before, and Bill Simmons touched on some relevant answers in a recent mailbag:
We knew something shifted in baseball a few years ago; it’s definitely happening in basketball right now. Whether it transforms the other sports remains to be seen. I do think we could reach a ceiling with performance-related formulas some day soon — if we’re not getting there already — and complicated analysis will shift to less definable quantities like injury recovery and behavior. But that’s a few years away. As I mentioned at the conference, the big challenge for sabermetricians this decade will be learning how to educate a mainstream audience in a relatable and entertaining way. Easier said than done.
There are some quantifiable areas that have yet to be fully explored, with defensive metrics still waiting for technologies such as Hit f/x to help take them to the next, more accurate level. However, there are some elements of the game, particularly offensive production and pitching, where the innovation seems to be about building upon existing ideas and adding a higher degree of accuracy rather than reinventing the wheel.
Outside of defense, where might we see some revolutionary ideas? Simmons mentions behavior and health, and I think health in particular will become a new frontier for statistical analysts, as we try and predict injuries based on workloads, pitch and swing types, and other observable factors. Teams that can find some measure of predictability in terms of player durability will find themselves at a strong advantage when it comes to building an effective, consistent team. Injury projections represent a logical evolution of the “Moneyball” philosophy that encourages teams to exploit market inefficiencies.
Where do you think the sabermetric revolution will take us next?
If you enjoyed this article, please consider sharing it!
2 Responses to Discussion: Where Are Sabermetrics Headed?
Leave a Reply
TIQ IQ
As I mentioned in an earlier post some time ago,that while it is not necessarily important for the public to understand how the advanced metrics are derived formulaically, it is imperative to educate as to what’s good, average and not so good for each new statistic that is developed.As one example, I am not certain that many above average fans can say with some precision these things about even a relatively archaic stat such as OPS and how that may vary along the defensive spectrum.By contrast, virtually everyone who follows baseball to some degree, grew up knowing that a BA of 300+ was good and 250- was not
As for particular areas to be explored, I think baserunning is one area that is often overlooked and underappreciated though it appears that some newer metrics are beginning to capture some of the more subtle elements. A liitle off point, but I think it would be really kind of neat if broadcasts could incorporate the times of various baseruuners as they run down to first base, go from first to third on a single, leg out a triple , score from second on a base hit, etc. It would be interesting to see the precise difference between a Carl Crawford or Jose Reyes and slower runners in this aspect of the game.Also, it would be interesting to see the difference in velocity of the throwing arms of various position players.We all know Johnny Damon had a terrible arm but how much worse is it in MPH than a Jeff Francoeur for example and how does this translate in feet on the basepaths if they are throwing from the same distance? Just a couple of areas which I think can be easily incorporated to give people a clearer picture.
Baserunning, arm ratings are both good examples, if only because they have not done a perfect job yet removing all external variables.