Trying to Put a Number on Outfield Speed

"I love taking hits away from guys and seeing their reaction." – Los Angeles Dodgers center fielder Andruw Jones.

"You know that kid impresses everybody. Like a Superman, right there, comes out of nowhere and jumps like that and got the ball (hit by Barry Bonds in the 2002 All-Star Game). So now I know, he's the man!" - Sammy Sosa, regarding Torii Hunter’s jumping catch.

Jones, Hunter, and Mike Cameron were just three center fielders who changed teams this past off-season, joining Juan Pierre and Gary Matthews Jr. from 2007. It's often been said that if you can play shortstop you can play anywhere in the infield, and that center field is the litmus test for outfield mobility. I can't tell you the countless times I've heard an announcer speak about an outfielder or two and throw in the obligatory "Where triples go to die," which raises a valid point, are triples a true indication of an outfield's speed and range?

By my count there are three kinds of triple catalysts: 1) the runner is fast, think Carl Crawford or Jose Reyes splitting a gap with a routine double only to stretch it to third, 2) the ballpark's environment (walls, distance, ect.) allows it, a ball high off of the Green Monster or hitting off an odd carom in Houston, 3) the defense misplays a ball or slow outfielders simply can't get there in time. With those guidelines in place I began my look into quantifying outfield speed using triples as a key ingredient, so begins SEO; Speed Estimation of Outfields, named after the usually awful Rays' pitcher Jae Weong-Seo who had some issues with the outfield, namely that the balls kept landing beyond the fences.

I must disclose that I'm awful at creating new metrics, so if you have a suggestion how to improve the formula don't hesitate to offer advice - I'm not going to be offended. Taking the amount of hits, doubles, triples, and homers into account I first formed a number that I could use to compare the propensity of defenses to allow triples between teams in accordance to the amount of non-triple hits they gave up. The formula was: (3Bs - 2Bs / (H-HR)*-100), the -100 turns the number into a positive non-decimal exclusive number that has the same affect as a negative number: the further from zero the worse the defense. Below are the ranks for each of the 30 teams.

A league average outfield sits at 20.7 for 2007; Toronto was the worst at giving up triples at 25.2 meanwhile the Dodgers were the best at avoiding it, ranking in with a 17. Of course defense and speed are two of the most difficult - if not impossible at this point - to judge, so to help set up a guideline for where the teams should be ranking I took the Tom Tango community scouting report data for outfielders "speed" rates and averaged them out for each outfield, to my dismay the Blue Jays were rated as having the best average speed while the Reds, who ranked 13th in my metric, were the worst defensive based on speed. Below are the average speeds ranked by the baseball internet community.

Earlier on I categorized each triple as being the result of one of three scenarios, I'd like to tweak that idea slightly; a triple is a result of the three variables working together with one usually acting as the dominant trait. While taking each triple from last season and examining each variable is a bit too much to ask, I do want to address the ballpark factor. Simply adding a park factor was an idea of mine, but I'm not completely sure how to implement it. Should we take the dimensions and build from the perfectly average ballpark, adding or subtracting value for distance? In cases like Colorado should we take altitude into account? What about high walls? Or even lower walls? Those are just some of the questions I'm not sure how to address when it comes to including the data.

I suppose the entire purpose of this post is simply throwing the idea out there, do you think we can actually put a number on outfield speed with the current data available?