Boston's Best Tools: Best Defender

Before the 2010 season began, defense was the buzz word around town. The front office talked about the need to improve the team’s fielding and backed that talk up by signing defensive wizards Adrian Beltre and Mike Cameron. Unfortunately, the plan did not come together. Beltre was masterful at the hot corner, but Cameron and projected left fielder Jacoby Ellsbury were both hurt the majority of the season. Their replacements were woeful defensively and to make matters worse the Red Sox lost slick fielding second baseman Dustin Pedroia and first baseman Kevin Youkilis as well. All told, the 2010 Red Sox were a poor defensive team, rating 56 runs below average by the Defensive Runs Saved metric (DRS) and -15 runs by Ultimate Zone Rating (UZR).

As 2011 rolls around, the Red Sox are once again hoping for improved defense. They have added the premiere left fielder in the game in Carl Crawford and shuffled the infield to accommodate newly acquired first baseman, Adrian Gonzalez. Their top fielder last season, Adrian Beltre, is gone, replaced by Youkilis, who shifts back to his original position. With an excellent starting rotation in place and a revamped bullpen, better glove work will be a key factor in run prevention. Whose glove will have the biggest impact though?

Defensive statistics are full of problems. It is simply the nature of the beast. Among the baseball stats community two main approaches have developed for analyzing defense. The first system, represented by Sean Smith’s Total Zone system, uses play-by-play data to estimate the number of plays above or below the average from the historical norms for each hitter and pitcher. The second system, represented by Mitchel Lichtman’s Ultimate Zone Rating, uses data from Baseball Info Solutions detailing the type of hit and the area in which it was fielded to estimate the number of plays above or below average.. For our purposes, we will look at several systems and try to reach some understanding between them.

Shown above, I have two location data based systems (DRS and UZR), the modified version of Total Zone’s system (TZL) and the Fans Scouting Report data (FSR). All systems express the fielder's impact in runs, not plays made. Each one has its own advantages and disadvantages.

The main problem with all defensive metric is the high levels of variation. For this reason, I have limited this conversation to players with over 1000 innings at each position. This has left out Jed Lowrie’s excellent work at shortstop (UZR credits him with 7+ runs saved in just over 700 innings) and Jacoby Ellsbury’s time in left (where UZR has him at 10.5 runs saved in 392 innings) Both of these players deserve mention, but the uncertainty around those numbers is too great to ignore.