Scouting the Centerfielders - Part 1

added 12/14/2005 by Scott Barzilla

Willy Taveras has been in the news lately as a rumored bargaining chip with the Yankees and Diamondbacks. Supposedly, Taveras was the principle person involved in a rumor involving Javier Vazquez. This poses the question of whether the Astros should consider trading Taveras. There are two schools of thought on this: he will get only better from here and he will only get worse from here.

In order to determine whether the Astros should consider dealing him, we'll look at those two proposals through fielding and hitting. With the fielding we will look at three dimensions of fielding numbers. Ideally, all three dimensions will show the same thing, but when they don't we will have to look at why and make a determination.

Opportunity Dimension

This is the simplest dimension. This is where we look at putouts and assists to see where Taveras ranks. Putouts and assists are similar to RBIs and runs scored in the sense that they can give us some definition, but depend greatly on the amount of opportunities the fielder has to make plays.

              Games Putouts  Range
Randy Winn      151     410   2.74
Brady Clark     145     399   2.79	
Johnny Damon    147     394   2.71
Aaron Rowand    157     388   2.49	
Jeremy Reed     137     383   2.85	
Carlos Beltran  150     378   2.55
Grady Sizemore  155     373   2.43
Andruw Jones    159     365   2.36	
Carl Crawford   154     361   2.36
Vernon Wells    155     351   2.34	
Juan Pierre     160     332   2.12
Willy Taveras   148     332   2.31
Jim Edmonds     139     318   2.32	
Gary Mathews Jr.123     316   2.63
Brad Wilkerson  129     312   2.47
David Dejesus   119     306   2.63
Luis Matos      120     298   2.54
Mark Kotsay     137     298   2.23	
Nook Logan      123     282   2.32
Steve Finley    104     266   2.61
Corey Patterson 122     239   2.01
Bernie Williams 112     226   2.07

So, Willy is essentially near the middle of the pack in putouts, but near the bottom of the pack in range factor (total chances per game). Now, we could stop here, but it would be like saying the league leader in RBIs was the best hitter in the league. There are a couple of numbers that can help explain why this is happening. First, the Astros are a high strikeout staff, and their pitchers have a higher groundball/flyball ratio than most of the league.

Roger Clemens    1.41
Andy Pettitte    1.62
Roy Oswalt       1.45	
Brandon Backe    1.03
Wandy Rodriguez  1.34	
Ezekiel Astacio  0.77

Most of the starters have groundball/flyball ratios well above one to one. Ironically, Astacio has the worst ratio and he had the worst batting average on balls in play, but that's just an aside. The pitchers with the most innings (Clemens, Pettitte, and Oswalt) all had ratios well above the league average. That meant that he didn't get as many opportunities.

                      SO     SO/9
Houston Astros      1164     7.26 
National League     1052     6.57

So, every fielder had nearly one fewer opportunity (collectively) per game because of the high strike out rates of the pitchers. This is a good thing, but it's not a good thing when we consider only the opportunity dimension. Most of the Astros probably don't measure up in this category because of this. We have to remember that some teams obviously finish as far below average as the Astros do above average. Only two other teams finished with more strikeouts than the Astros, so we can see how skewed the numbers can look. This means we have to find another way.

Efficiency Dimension

STATS Inc. developed a new metric several years ago called zone rating. Zone rating was designed to solve many of the problems associated with range factor. Range factor is based on opportunities while zone rating calculates the percentage of balls a fielder successfully fields that go into his zone. Therefore, it doesn't matter how many opportunities a player gets in zone rating. It only matters what percentage of balls you get to. Zone rating does have one limitation: it cannot measure value, but we'll get to that later. For the time being, we need to see whether the efficiency dimension and opportunity dimensions square up. If they do, then the numbers will likely reveal the truth. If not, then we'll need to go through some more explanation.

                   Zone Rating
Jeremy Reed        .943
Aaron Rowand       .939
David Dejesus      .923  
Vernon Wells       .911
Luis Matos         .909 
Grady Sizemore     .908
Nook Logan         .904
Corey Patterson    .900
Brady Clark        .892
Jim Edmonds        .888
Carlos Beltran     .887
Juan Pierre        .882
Preston Wilson     .878
Dave Roberts       .875
Johnny Damon       .874
Andruw Jones       .873
Willy Taveras      .864 
Bernie Williams    .862			
Mark Kotsay        .850
Ken Griffey Jr.    .802

So, Willy finished 17th in zone rating and 17th in range factor among the twenty centerfielders to play in 100 or more games. That doesn't bode well for his defensive value. However, as we can see, these percentages look nice, but they don't mean a whole lot to us. This is where the value dimension comes in.

Value Dimension

The value dimension is the most important dimension because this is where we convert the opportunity dimension and efficiency dimension into something we can use. Baseball has several units of measurement (outs, runs, and wins), but for fielding, runs is the most important dimension. The good folks at Baseball Prospectus have developed two metrics we will use to measure value. Fielding runs above replacement is a measure of the number of fielding runs a fielder has over the worst possible regular at that position.

Fielding runs above replacement is a good metric we can use to measure productivity. Since Willy T played in nearly 150 games last season, he'll likely finish pretty high in FRAR. Fielding runs above average is exactly what it sounds like. It is what we use to determine value. Both are based on the number of plays that are made, but the formula also factors out discrepancies between groundball/flyball ratios, strikeouts, lefty/righty breakdowns, and ballpark peculiarities. Baseball Prospectus's exact formulas are not available, but they seem to be more accurate than fielding run formulas of the past. The good news is that we don't need to rank the centerfielders since the benchmark is based on an average.

                  FRAR   FRAA
Willy Taveras       31     15

So, obviously the value numbers show something that the others do not. Clearly though, the discrepancy means we need to do some debriefing. We discovered that fielders have fewer opportunities in Houston because of the strikeouts and the outfielders have fewer opportunities because of the groundball/flyball ratio. The zone rating is difficult to explain definitely, but when we watch Taveras we notice he sometimes doesn't get a good jump on balls. The question is whether this will improve or not.