Explaining Stats: Ultimate Zone Rating
By Jeff Ellis
February 24, 2012
Ultimate zone rating (UZR) was devised by Mitchel Lichtman using the zone rating stat used by STATS, INC. It is one of the many defensive stats that have gained prominence in an attempt to find something that can measure defensive success. It measures a fielder's success in getting to balls which are in the zones of his specific position.
For many years the main stats for defense were errors and fielding percentage, but as time has passed these numbers have shown to not really display a player's true defensive skill. This has lead to a growth in defensive stats, but there is not yet a distinctive best stat for defense; however, the one I do see the most is UZR.
The way UZR is calculated is a bit complex so bear with me. The field is divided into 78 zones and 64 of them are used in calculations (the ones left out are infield line drives, infield flies, and outfield foul balls) . So the next step is to calculate the out rate and the percentage of balls that result in outs in each zone. Once you have this league average you then subtract the players' out rate. Now it gets less complicated because negative is bad and positive is good, so it is pretty easy to read and understand the stat. This number is multiplied by the numbers of balls hit into that player's zone, and then you have a zone rating for one specific zone. Then all of the zones of the position are added together and that yields a basic unadjusted UZR.
This is then adjusted based on park factors, batted ball speed, pitcher groundball/flyball ratio, outs at the time, base runners, and even the side of the plate the hitter was hitting from. The idea is that these factors can all have a significant impact on defense. Then you take this data and using run expectancy charts everything is converted to runs (If you are lost that's okay because so am I). One more important note is that it is a stat like home runs, runs, etc. in that the more games you play the higher the values tend to be.
What makes UZR great is it is an all purpose stat that takes into account a ton of variables and in spite of its complex nature it is very logical.
Of course there are a few problems. One problem is that there is no UZR for pitchers or catchers, so it eliminates two positions to be able to judge defensively. Next, UZR can have a huge variance, and frankly, on a year to year basis it is useless. It is best used as a way to look back over a long period of time and not s single year because one year is considered too small a sample size. In other words, this stat should never be used in talks for the Gold Glove Awards. Fangraphs has a UZR/150, which is a way to compare players that have played in a different number of games, but this stat is really not effective because three years is considered the minimum amount of data needed to get a useable stat.
UZR is an interesting stat and more importantly is an easy to understand defensive stat. It is better than using errors or fielding percentage to determine who the best defender is, and it is easier to understand than those stats. Really, the important thing to take away from it is that it is not really a year to year stat. Next time you get into a defensive argument and that person brings up UZR, remember it is best used as a career summing stat and that UZR for the year is too small a sample size to make it a valid tool for an argument. It is really meant as a multi-year tool to show how well a player has preformed defensively, meaning it is great for a Hall of Fame debate than a Gold Glove debate.
Follow me on Twitter @JEFFIPI
Follow Jeff on Twitter @jeffmlbdraft, or email him at firstname.lastname@example.org
What really kills it for me is from 2008-2012, Russell Branyan had a positive UZR. I watched him play many games in 2010, and he was largely immobile, and balls that even a Laporta would've gotten to skipped past him. He was not a better than average 1st baseman during that time frame, he was a solid contributor to Masterson and others' inflated BABIPs.
It also makes no sense to use a stat that's been discredited as being useful on a year-to-year basis in calculating WAR, so i kind of wish UZR would go away