Tuesday, September 21, 2004
Look Out, Buster Olney
What is the point of this stat? I guess it should correlate pretty well with runs scored and runs against. There's no evidence, however, that scoring runners on second and third is more important than scoring runners from first or hitting home run. The closest this comes to being useful is to create a luck factor (or, generously, a RISP factor) by adjusting it for a team's overall offensiive productivity. That seems pretty useless, though. Isn't that what second and third order wins are for (and don't those measure it much better)?
The O Zone, a creation of MLB.com, measures a team's ability to score runners once they have reached second or third base, as well as its ability to prevent opponents from scoring runners from second or third.
It is a statistic not unlike football's red zone and hockey's power play statistics -- stats that have become the accepted standards when measuring a team's ability to succeed when inside the 20-yard line, in the case of football, or when it has a man advantage, as in hockey. And conversely, in preventing opponents from cracking the end zone or finding the back of the net.
Until now, baseball has not offered a similar statistic. Sure, a fan will hear how Gary Sheffield is hitting with runners in scoring position, or how Eric Gagne performs when he enters with runners in base, but never before has the fan been able to gauge how successful, or unsuccessful, an entire team is in these situations.
The O Zone Factor was created to accomplish this task.
How does it work?
The Cardinals, owners of the best record in baseball, have scored 43 percent of runners who have reached second or third, while only allowing 37 percent of opposing teams' runners in scoring position to reach home. Subtract .370 from .430 and the result is .060 -- that is St. Louis' O Zone Factor, and it is by far the top rating in baseball this season.
In most cases, teams with higher O Zone Factors are teams that are currently in the playoff hunt, while teams with lower rates are no longer in postseason contention.
This O-Zone business shouldn't bother me as much as it apparently does, but wouldn't it be nice if mlb.com actually paid homage to stats that have already been developed and are shown to have actual correlation with winning rather than forcing more inane stats into the world? I don't know about you, but my appreciation for baseball grows by having statistics available that explain past performance and predict future performance. There's simply no evidence to support O-Zone as doing either of those things, and all of my knowledge of baseball indicates that evidence is never going to come. Oh, how I wish that mlb.com could focus on heightening the average fan's understanding of advanced metrics. Is EqA really a more difficult concept than O-Zone? It's not like O-Zone has some obvious accessibility; it's not a stat the average fan can calculate at home any better than they can calculate EqA.
The worst part might be that this article acts as if the correlation between O-Zone and winning implies causation. Their numbers show that most of the contenders are doing swell in O-Zone and that all the teams that are doing well in O-Zone are doing well overall. Umm, yeah, but maybe that has to do with the fact that the better teams avoid more outs and hit more extra-base hits and get their opponents out and so forth. I would give 10,000 to one odds that over any three-year period of a baseball league a team's Offensive GPA minus defensive GPA correlates better to winning than their O-Zone factor. Since I'd rather not waste my time researching that one now, any takers are welcome. It's maddening that the mlb.com article acts as if there's some supreme correlation that cuts through all the mystery surrounding how teams win baseball games:
Outside of winning percentage, no statistic is ever going to be 100-percent accurate when predicting team success. The O Zone Factor, while simple is theory (sic), illustrates the importance of a team's performance with RISP in respect to winning ballgames.Did I miss something? Where's the illustration of this importance? I know it's important, but not because of O-Zone. Furthermore, the article presents no evidence to suggest that O-Zone predicts future performance yet implies that it does. The article demonstrates some correlation between winning and O-Zone, but no evidence that suggests that a team's O-Zone will remain consistent over any period of time. O-Zone might be a good tool to show that a team's good fortune came from unusually high success with RISP in the past (or that bad fortune came from lack of said success), but that alone has little to do with predicting future success, and it would still only make sense in the context of the differential between RISP OBP/SLG and overall OBP/SLG.
Additionally, the article makes the curious claim that a team's winning percentage will be "100-percent accurate when predicting team success." A string of question marks is in order here, as I know that teams' winning percentages change frequently and I'm pretty sure Bill James has done the work to illustrate that run differential better predicts future success. Couldn't someone over there point out that there are decent ways of predicting future success? At the all-star break, third-order wins showed that the Dodgers were the class of the west, the Giants a solid second place, the Padres not that good, and the Cubs and Astros the best bets for the wild card. The results as they stand now seem to validate that; would O-Zone have predicted each team's success as well? The fact that the Giants are, shall we say, struggling in their collective O-Zone according to MLB's stats would seem to indicate it would not have.
So let's say you have a near monopoly over your sport and need only to convince potential consumers that your sport is superior to the other sports. Your sport works on logic, consistent individual performances, large sample sizes, and playing six or seven times a week. The other sports revolve around people slamming into each other for three hours one to three times a week. Should you a) market your sport as mysterious and containing all sorts of cool stats that can't tell you much except what happened in the past or should you b) emphasize stats that are backed by empirical research and have great explanatory power going forward? Why use stats that make you look like the competition when you can use stats that make you look superior to the competition? As a fan, I'd much rather have a good idea of what should happen. Baseball is like music or film: patterns and rhythms emerge that create expectations and sensations are produced by surrounding individual events with suspense so that it's a payoff both when the expected and the unexpected happen. Football, basketball, and hockey are lacking in this department. "Any Given Sunday" explains football to a tee: these people only do this sixteen times a year. Baseball has all the intrigue of "What Will Happen Next?" but with all the fun of "I can tell you what's going to happen in the next few weeks." As a baseball fan you can be both surprised and stimulated by searching for explanations in a systematic way. Football fans have a maddening array of non-stats: the Sacks total for each player will change substantially from year to year for any good pass rusher, Pancake Blocks tell you almost nothing about a player's overall performance, and so forth. Yet MLB wants to market all the Sacks and Pancakes it can come up with while shunning Equivalent Average, Gross Production Average, Wins Above Replacement, and the like. This is not what I would do.
Song of the Day: "How Am I Different?", Aimee Mann
UPDATE: Mike Carminati posted within moments of me, and shows quickly that O-Zone is indeed worthless.