## What Can We Learn From Video-Based Defensive Statistics?

Garbage in, garbage out. This is an everlasting truth of just about any system. If you’re writing code or brewing beer, the quality of your ingredients and inputs greatly effects the quality of your product.

When we’re evaluating baseball players, the quality of the inputs has the same impact on the product that comes out on the other end. For fans arguing in the bleachers or front offices carefully weighing their trade options, there is no way around this. Many folks balk at the comprehensive measures like Wins Above Replacement because of the cryptic and unreliable nature of defensive measurement. It is much easier to take offensive stats at face value. The math is just as gory but the events are discrete and easily countable.

This is slowly changing, of course. As we saw earlier this week (read Ben Lindbergh for a deeper look at what the new tracking system can do), the manner in which we measure and monitor defensive contributions are changing but quick. Beyond the “zone” based approach of most defensive measures, we can track the reaction time, running speed, distance traveled and overall “efficiency” of the route each fielder takes to a batted ball.

It’s a huge step in an incredibly promising direction, though one with a future very much in doubt. How much information will trickle down to the fans? What will this information mean for the way we watch the games?

For now, this remains a bit of a mystery. Fangraphs unveiled a new tool to better tackle the present as we patiently wait for the future to arrive. Inside Edge and their video-based fielding carefully pours over each and every ball in play, grouping plays based on their likelihood of being fielded.

Fangraphs and Inside Edge make use of six different buckets, grouped by the chances of fielding a given ball in play: impossible (0%), remote (1-10%), unlikely (10-40%), about even (40-60%), likely (60-90%), and almost certain/certain (90-100%). While it is a less precise measure than the tracking systems provide, it does offer a window into how we think about players.

Looking over the numbers (Inside Edge stats are available for 2012 and 2013 only), a few key observations jump out. Please enjoy them in assorted thoughts form:

• Mike Trout goes and gets it. Mike Trout had a terrific defensive season in 2012, posting huge def. stats in his rookie year. He saved home runs and ran down balls in the gap. Watching the games, we know this. The Inside Edge numbers support this. Over the last two years, 20 balls were hit in Trout’s direction with a “remote” chance of being successfully taken. Trout converted six of these into outs, a rate of 30%. Among all fielders at all non-catcher positions, this is the highest rate of any player with at least 15 chances.
• Shortstop is such an interesting position J.J. Hardy has a reputation as a steady shortstop who “makes all the plays.” The scout-based fielding numbers support this claim, as he ranks very highly in the “certain” and “likely” categories but below-average on the tougher balls to field.
• The most important defensive position on the diamond runs the gamut in terms of the way they play the position. On the other side of the Hardy coin sits Adeiny Hechavarria. Rated as one of the worst players in baseball last season, Hechavarria remains a dynamic defensive player. Watching him take infield could cure the sightless and place a smile on the face of a hardened war criminal. The Inside Edge numbers show that no other shortstop converted “unlikely” or “remote” plays better than the slick Cuban infielder, though his play on more pedestrian balls shows he has room to improve.
• In terms of the general trends, it becomes clear first base really is a defensive dead end. There just aren’t that many chances, especially difficult to field balls. As you’ve read in this space many times before, the duties of first do not allow even a superlative fielder to make a big difference.

Interesting as this information might be, it still must come with the required grains of salt. The percentage differences are so slight and represent just a few made plays between worst and best. And it still doesn’t account for differences in positioning, reaction times and the other stuff the next level of numbers provides.

Hopefully, some enterprising baseball fanalyst with database expertise can pour over these numbers in greater detail. Is there a connection between age and range? More importantly, is range more important than sure-handedness or are we in a “take care of the cents and the dollars take care of themselves” situation?

As our own Jack Moore wrote earlier this week, the new MLBAM player tracking system can only increase the way fans recognize and appreciate the various contributions players make to the greater cause of winning ball games. The desire to generate a grand unified theory of baseball is understandable but misguided. Most of us can only hope to sharpen our focus and learn to better appreciate what and how the best players make their contributions.