No announcement yet.

Defensive metrics

  • Filter
  • Time
  • Show
Clear All
new posts

  • Defensive metrics

    I've come to realize that my sabermetric undertanding of defensive metrics is quite lacking. Can you recommend some good bookS to read. Also has range factor, etc. gone the way of the dodo bird?
    Last edited by Honus Wagner Rules; 06-26-2006, 03:45 PM.
    Strikeouts are boring! Besides that, they're fascist. Throw some ground balls - it's more democratic.-Crash Davis

  • #2
    Originally posted by Honus Wagner Rules
    I've come to realize that my sabermetric undertanding of defensive metrix is quite lacking. Can you recommend some good bookS to read. Also has range factor, etc. gone the way of the dodo bird?
    There is a book out on fielding. I picked it up. Haven't gotten to read it yet.

    It's called the Fielding Bible.

    By the way, whatever happened to the dodo bird?


    • #3
      Originally posted by Balmes Pavlov
      There is a book out on fielding. I picked it up. Haven't gotten to read it yet.

      It's called the Fielding Bible.

      By the way, whatever happened to the dodo bird?
      Thanks! As for the dodo bird...
      Strikeouts are boring! Besides that, they're fascist. Throw some ground balls - it's more democratic.-Crash Davis


      • #4
        This may be able to explain some defensive metrics

        PART I – Range Factor
        PART II – Zone Rating
        PART III – Defensive Efficiency
        PART IV – Fielding Percentage

        PO = Put Outs
        A = Assists
        E = Errors
        PA = Plate Appearnces
        HBP = Hit By Pitch
        H = Hits
        K = Striekouts
        BB = Walks

        Range Factor
        ((PO + A)*9/innings played)

        The purpose of range factor is to establish a quantifiable method for the amount of outs a fielder can account for. Created by Bill James, range factor is supposed to be able to estimate what kind of range or ground a player can cover at their position by basing the equation on the total amount of plays involving the player. The most practical use of range factor is to try and determine who is best at what position given the amount of balls put in play.

        For instance, Toronto’s Orlando Hudson currently has a range factor of 6.02. Range factor is figured by the following simple equation: [((Put Outs + Assists) x 9)/ Innings)]
        Therefore, to figure out what Hudson’s range factor we just plug in his numbers:
        ((127 PO + 171 A)*9/445.1) = 6.02, tops for major league second basemen. What this tells us is that Hudson has 6.02 successful chances per every nine innings he plays, or, he is able to convert about 6 outs per game.

        The main problem that people find with range factor is that it not only depends on the defensive abilities of the player, but those around him as well. For players who play on teams with power pitchers who strikeout lots of batters, their range factor might suffer because they don’t get as many chances.

        The most important thing to remember when considering range factor is the nature of player positions. First basemen and catchers will always have a high range factor because they are involved with most putouts over the course of a game, but will have few assists. The putouts they accumulate are so high, that often the range factor for first basemen will hover near 10 with catchers between 7 and 9. Pitchers often have smaller range factors upward of 3 because they are involved with so few putouts and assists. Because of this, Bill James was led to discount those positions since they are more a part of an out than a cause of one.

        Here are the average RF numbers per position in 2004 for all other positions:
        2B, 4.98
        3B, 2.70
        SS, 4.50
        LF, 1.94
        CF, 2.58
        RF, 2.15

        Taking all these numbers into account, let’s look at infield range factors and how a team’s pitching can affect it.

        Look at the Chicago Cubs. They currently lead all of baseball with almost eight strikeouts per nine innings. The average range factor for the six fielders (excluding first baseman, pitcher and catcher) is 2.62. What you’ll generally find is that teams with higher strikeout ratios have lower range factors. If you have Mark Prior, Kerry Wood and Carlos Zambrano pitching for you and striking out one of every three batters, the fielders won’t get many chances to convert balls put into outs.

        On the flip side, the Boston Red Sox are 22nd in MLB with just under 6 Ks a game. Their average range factor for position players (minus first, pitcher and catcher) is 3.21, more than half a point higher than the Cubs. If Boston’s pitchers aren’t striking out as many batters and forcing more balls put into play on the ground or in the air, more Boston fielders will get opportunities. This will create a higher range factor.

        With that said, the biggest flaw with range factor is it depends more on the team as a whole, not necessarily an individual player. But, it can still give a decent projection at a player’s worth as far as their ability to turn balls within their range into outs.

        Zone Rating
        Formula kept through STATS, Inc.

        In the late 1980s, STATS, Inc. created the zone rating as a means to further what the range factor wasn’t able to explain. With zone rating, STATS was able to assign specific “zones” for each position. While range factor looks at how players turn their chances into outs, zone factor gives a percentage that translates all the balls hit into the area where a player can be expected to record an out, then arrives at the percentage of outs that are actually made.

        The beauty of the zone rating is that the system doesn’t provide ridiculous zones for players to cover. That way, zones don’t overlap in the infield and outfielders. Since all ballparks aren’t created the same, you can obviously expect differences in how zones are created depending on the specific stadium. Where range factor uses numbers that a fielder actually produces through putouts and assists, zone rating is more helpful because it provides a more accurate portrayal of what each player should be able to do.

        For an example, let’s look at Edger Renteria. Generally considered as one of the top defensive players in baseball, Renteria won consecutive Gold Gloves with the St. Louis Cardinals in 2002 and 2003. In 2002, Renteria recorded a zone rating of .856, good for only 11th in the majors for players who logged at least 880 innings at shortstop. In the National League Renteria was 5th overall, behind Jimmy Rollins, Cesar Izturis, Rey Ordonez, and (surprise) Jose Hernandez, who was first overall with a zone rating of .883. Renteria was able to sneak away with the top defensive award even though he wasn’t tops at his position in zone rating, range factor or fielding percentage.

        In 2003, Renteria’s case was a little stronger. He had the best zone rating in the NL at .871 and third best out of all shortstops who played on a regular basis (minimum 880 innings). Perhaps his award was a bit more deserving in 2003 (he had a ZR of .855 in 2004, good for 5th best in the NL). However, Renteria is back to his old ways again this season in Boston, ranking 20th in zone rating with .804 (minimum 340 innings). Maybe he’ll get lucky if he turns in a few highlight reel plays a night like he did in 2002.

        The fact that a player who isn’t tops among their league or MLB statistically, but can still win major awards leads to one of the bigger criticisms of zone rating, which is that it doesn’t count all the balls hit to a player. For example, infielders aren’t charged if a ball is hit up the middle, down the foul lines or through the gaps. In short, a player is simply responsible for the balls hit within their small range. Also, players who turn more double plays will get awarded two outs in zone ratings because their play resulted in two outs (although that is a lesser problem).

        Even still, zone rating helps give a more accurate look at the defensive abilities of players because it will actually matter if a fielder can create plays or not. While extra points aren’t awarded for pretty plays, zone rating does show how hard working players who put forth strong efforts can still stack up with the highlight reel boys.

        Defensive Efficiency
        (PA-H-K-BB-HBP- (0.6*E)))/(PA-HR-K-BB-HBP)

        This statistic really requires less explanation than the others. Defensive efficiency measure the percentage of balls put into play and how many are turned into outs (without including home runs). At the end you’re left with a percentage that represents just how productive your team is at creating outs with the opportunities given.

        DE relies greatly on the performances of pitchers since they can keep plate appearances and balls put into play at a minimum. This in itself makes the defensive opportunities easier. If you have a pitcher like Greg Maddux, Jake Westbrook or Zack Greinke who all rely on their defenses to back them up, their teammates might produce a lower defensive efficiency rating when they’re on the hill.

        On the Cubs, Maddux currently has a defensive efficiency rating of .717 when he is pitching. That is lower than Carlos Zambian’s .770. One stat to look at in relation to DE is strikeouts per nine innings. Maddux is historically a groundball pitcher who relies on defense. Zambrano is the opposite - an overpowering pitcher with great strikeout potential. On the season, Zambrano is striking out nine batters every nine innings pitched. Maddux, meanwhile, is fanning only 5.5. One reason Zambrano might have a better defensive efficiency rating is because less balls are put into play when he pitches, which means fewer chances for players to botch any potential outs.

        Continuing our look at the same pitchers, Zambrano has a groundball to flyball ratio of 1.64:1, much lower than Maddux’s 2.16:1. This provides a more solid way of seeing how pitchers who need stronger defense tend to have lower defensive efficiency ratings.

        Fielding Percentage

        Fielding percentage measures the number of fielding opportunities a player has without an error. This is perhaps the most basic of all the statistics mentioned, simply because it’s easy to calculate and easy to understand. Good players who make fewer errors will have higher percentages and players who make more mistakes will see their fielding percentage drop.

        go sox.



        • #5
          Also has range factor, etc. gone the way of the dodo bird?
          Bill James actually has a chapter or two about range factor in Dewan's Fielding Bible. He talks about where it went wrong and then discusses a revised version.


          Ad Widget