Thoughts On Outfield Defense
The conventional wisdom spewed forth from the mouth of Ken Williams states that by swapping Carlos Lee for Scott Podsednik and adding Jermaine Dye in right field the White Sox will follow through on promises of improved pitching and defense with vastly superior outfield defense in 2005. Considering the effort I've invested in finding some upside to the Lee-Podsednik swap beyond an additional $6 million in payroll, $4 million of which has already been invested in Orlando Hernandez, I would love to believe that the team has replaced an outfield of plodding sluggers with tremendous athletes who will routinely race to the gaps to catch balls which previously would have dropped for doubles. Unfortunately, the available data suggests that the purported improvements in the club's outfield defense for 2005 are overblown at best.
The majority of the misconceptions about the White Sox outfield defense seem to stem from flawed perceptions of Carlos Lee's play in left field. Sox lore states that Lee was an absolute butcher in left for his first several Major League seasons after making the move from third base just prior to his debut. In spring training in '03 Lee was reported to have spent considerable time on his fielding with then manager Jerry Manuel and the coaching staff. During the '03 season, it became accepted that Lee had improved his fielding and reached the skill level of an average or slightly below average left fielder. The numbers tell a somewhat different story. Clay Davenport's fielding statistics at BaseballProspectus.com do demonstrate Lee struggling as an above average fielder his first three seasons before improving to average in 2002 and 2003. However, in a development which may shock many Sox fans, Davenport's numbers peg Lee at a spectacular 13 runs (or more than a game) better than the average left fielder in 2004. Michael Lichtman's Ultimate Zone Rating (UZR) more or less confirms BPs findings through '03 rating Lee as slightly above average in range. It will be highly interesting to see if the UZR numbers incorporating '04 reflect the same leap in performance as BP's figures.
Perhaps most illustrative of Carlos' underrated defense however is the writing of Tom Tippett at Diamond Mind baseball, most likely the first to stumble upon Lee's success in the field. Examining the gold glove races in the 2003 season Tippet wrote:
There are several other AL outfielders who might be worthy of consideration if not for the presence of these three guys. Among the corner outfielders we noticed are Winn, Garrett Anderson, and (believe it or not) Carlos Lee.
I'm sure that last name will come as a surprise to many of you. It came as a big surprise to us, too, because Lee has a reputation as a defensive liability and has been removed for defensive purposes more often than any other fielder in recent years. As a result, we spent a lot of time studying his performance, and here's what we found:
- Lee stole 18 bases in 22 tries this year, and his career totals are 53 steals and a 72% success rate, so he does have some speed
- according to our analysis, Lee had no weak spots ... he was at or above the league average in all zones and depths ... and while he hasn't been this good before, he was slightly above average in 2001 and 2002, so this type of performance isn't as much of a reach as you might think
- other systems place him in the top half ... he was 24 points above average in the STATS zone rating system, and his range factor and adjusted range factor were both a little better than average
- the defensive replacements are easy to explain ... he was being replaced by two exceptional fielders, Aaron Rowand and Willie Harris, so even though Lee was getting the job done, these guys were better
- overall, Chicago's outfield converted almost as many fly balls and line drives into outs as did the Minnesota trio, so somebody was doing something right ... Rowand and Harris were major contributors, but they didn't play enough to explain this, and Lee appears to have done more than Magglio Ordonez and Carl Everett to help this outfield rank so high
Even after reviewing all of this information, I wasn't convinced. So I decided to spend some time with the MLB.com video clips service. I picked a six-week period and requested every play Lee was involved in.
(MLB.com's service isn't perfect, so I was able to get my hands on only about 80% of those plays. But think about that for a minute. I was able to call up dozens of video clips for a specific fielder in a matter of seconds, and it only cost me a few dollars. Yeah, it would have been nice if I found everything I was looking for, but how can I complain about some missing clips when such a thing wasn't even conceivable a few years ago?)
It took about three hours to view the clips that were available, and I came away very impressed. There must have been ten or eleven really good plays in that stretch. Among them were two long runs to flag down deep fly balls in the gap. On two other occasions, Lee reacted very quickly to line drives and made sliding catches to his left. Twice he went over the left field wall to save homeruns. And in what may have been his best play of that sequence, he covered a lot of ground to make a catch in foul territory while going up and over the bullpen mound at full speed.
Admittedly, sabermetric analysis of defense is far from perfect. However, performance based observations have clearly demonstrated that subjective evaluations from scouts are highly susceptible to biases and arbitrary observations. Therefore, defensive evaluations based purely on scouting do not represent a desirable alternative to sabermetric methods. Ideally, both approaches should be combined as with Tippett's use of video clips to coraborate statistical analysis. That three different methods of analysis (plus Tippett's scouting observations) rate Lee above average, with two of the methods suggesting his defensive performance is outstanding, provides a rather reliable indication that Lee is a highly underrated and highly valuable defender in left field. This revelation immediately weakens notions of a highly improved outfield defense in 2005. One of the strongest assumptions responsible for predictions of improved defense is that Lee is a plodding slugger with average to above average defensive ability. The logic follows that replacing Carlos in left with an athletic, defensive-minded outfielder will therefore generate huge gains in the field. Once one considers the fact that Lee is most likely a well above average defender, it becomes far more difficult to reap benefits through the elimination of his defensive production.
Aaron Rowand is an outstanding defensive outfielder. Davenport's metrics rate him as slightly above average in CF in 2005, not an easy feat to accomplish. I have little doubt that Rowand will play tremendous defense in left field and I suspect he will be a better left fielder than Lee in 2005. The question though, is how much better than Lee will Rowand be in '05? To represent a one game improvent over the course of the season Rowand would need to perform at more than 20 runs above average next season, a rather difficult feat. Unless Rowand manages to redefine the position of left field in 2005, it will be difficult for the White Sox outfield defense to provide the significant improvement envision by Williams and the front office. UZR rates Podsednik as a centerfielder with below average range while Davenport's methods peg Pods at 6 runs worse than Rowand in '04, suggesting the outfield may actually need to compensate for poorer defensive production in center. BP's metrics also suggest that Dye generally been a below average defender the past six seasons and claims persist that Dye has failed to regain the mobility he possessed prior to breaking his leg in the playoffs with the A's several years ago. Magglio Ordonez, the man he replaces, was above average in 43 games last season and an impressive 11 runs above average in his last full season in 2003.
It should be fun for Sox fans to watch Rowand dominate left field after battling to play average defense out of position in center field the last few seasons. Unfortunately, the bottom line is that the man he is replacing did an outstanding job in left last season. Furthermore the team is potentially looking at net losses defensively in centerfield and right. While I'm not ruling out the possibility of slightly improved outfield defense in 2005, trading your best defensive outfielder (relative to position) and downgrading in the two other outfield spots is not a surefire recipe for reconfiguring your outfield in the name of game altering defense.
The majority of the misconceptions about the White Sox outfield defense seem to stem from flawed perceptions of Carlos Lee's play in left field. Sox lore states that Lee was an absolute butcher in left for his first several Major League seasons after making the move from third base just prior to his debut. In spring training in '03 Lee was reported to have spent considerable time on his fielding with then manager Jerry Manuel and the coaching staff. During the '03 season, it became accepted that Lee had improved his fielding and reached the skill level of an average or slightly below average left fielder. The numbers tell a somewhat different story. Clay Davenport's fielding statistics at BaseballProspectus.com do demonstrate Lee struggling as an above average fielder his first three seasons before improving to average in 2002 and 2003. However, in a development which may shock many Sox fans, Davenport's numbers peg Lee at a spectacular 13 runs (or more than a game) better than the average left fielder in 2004. Michael Lichtman's Ultimate Zone Rating (UZR) more or less confirms BPs findings through '03 rating Lee as slightly above average in range. It will be highly interesting to see if the UZR numbers incorporating '04 reflect the same leap in performance as BP's figures.
Perhaps most illustrative of Carlos' underrated defense however is the writing of Tom Tippett at Diamond Mind baseball, most likely the first to stumble upon Lee's success in the field. Examining the gold glove races in the 2003 season Tippet wrote:
There are several other AL outfielders who might be worthy of consideration if not for the presence of these three guys. Among the corner outfielders we noticed are Winn, Garrett Anderson, and (believe it or not) Carlos Lee.
I'm sure that last name will come as a surprise to many of you. It came as a big surprise to us, too, because Lee has a reputation as a defensive liability and has been removed for defensive purposes more often than any other fielder in recent years. As a result, we spent a lot of time studying his performance, and here's what we found:
- Lee stole 18 bases in 22 tries this year, and his career totals are 53 steals and a 72% success rate, so he does have some speed
- according to our analysis, Lee had no weak spots ... he was at or above the league average in all zones and depths ... and while he hasn't been this good before, he was slightly above average in 2001 and 2002, so this type of performance isn't as much of a reach as you might think
- other systems place him in the top half ... he was 24 points above average in the STATS zone rating system, and his range factor and adjusted range factor were both a little better than average
- the defensive replacements are easy to explain ... he was being replaced by two exceptional fielders, Aaron Rowand and Willie Harris, so even though Lee was getting the job done, these guys were better
- overall, Chicago's outfield converted almost as many fly balls and line drives into outs as did the Minnesota trio, so somebody was doing something right ... Rowand and Harris were major contributors, but they didn't play enough to explain this, and Lee appears to have done more than Magglio Ordonez and Carl Everett to help this outfield rank so high
Even after reviewing all of this information, I wasn't convinced. So I decided to spend some time with the MLB.com video clips service. I picked a six-week period and requested every play Lee was involved in.
(MLB.com's service isn't perfect, so I was able to get my hands on only about 80% of those plays. But think about that for a minute. I was able to call up dozens of video clips for a specific fielder in a matter of seconds, and it only cost me a few dollars. Yeah, it would have been nice if I found everything I was looking for, but how can I complain about some missing clips when such a thing wasn't even conceivable a few years ago?)
It took about three hours to view the clips that were available, and I came away very impressed. There must have been ten or eleven really good plays in that stretch. Among them were two long runs to flag down deep fly balls in the gap. On two other occasions, Lee reacted very quickly to line drives and made sliding catches to his left. Twice he went over the left field wall to save homeruns. And in what may have been his best play of that sequence, he covered a lot of ground to make a catch in foul territory while going up and over the bullpen mound at full speed.
Admittedly, sabermetric analysis of defense is far from perfect. However, performance based observations have clearly demonstrated that subjective evaluations from scouts are highly susceptible to biases and arbitrary observations. Therefore, defensive evaluations based purely on scouting do not represent a desirable alternative to sabermetric methods. Ideally, both approaches should be combined as with Tippett's use of video clips to coraborate statistical analysis. That three different methods of analysis (plus Tippett's scouting observations) rate Lee above average, with two of the methods suggesting his defensive performance is outstanding, provides a rather reliable indication that Lee is a highly underrated and highly valuable defender in left field. This revelation immediately weakens notions of a highly improved outfield defense in 2005. One of the strongest assumptions responsible for predictions of improved defense is that Lee is a plodding slugger with average to above average defensive ability. The logic follows that replacing Carlos in left with an athletic, defensive-minded outfielder will therefore generate huge gains in the field. Once one considers the fact that Lee is most likely a well above average defender, it becomes far more difficult to reap benefits through the elimination of his defensive production.
Aaron Rowand is an outstanding defensive outfielder. Davenport's metrics rate him as slightly above average in CF in 2005, not an easy feat to accomplish. I have little doubt that Rowand will play tremendous defense in left field and I suspect he will be a better left fielder than Lee in 2005. The question though, is how much better than Lee will Rowand be in '05? To represent a one game improvent over the course of the season Rowand would need to perform at more than 20 runs above average next season, a rather difficult feat. Unless Rowand manages to redefine the position of left field in 2005, it will be difficult for the White Sox outfield defense to provide the significant improvement envision by Williams and the front office. UZR rates Podsednik as a centerfielder with below average range while Davenport's methods peg Pods at 6 runs worse than Rowand in '04, suggesting the outfield may actually need to compensate for poorer defensive production in center. BP's metrics also suggest that Dye generally been a below average defender the past six seasons and claims persist that Dye has failed to regain the mobility he possessed prior to breaking his leg in the playoffs with the A's several years ago. Magglio Ordonez, the man he replaces, was above average in 43 games last season and an impressive 11 runs above average in his last full season in 2003.
It should be fun for Sox fans to watch Rowand dominate left field after battling to play average defense out of position in center field the last few seasons. Unfortunately, the bottom line is that the man he is replacing did an outstanding job in left last season. Furthermore the team is potentially looking at net losses defensively in centerfield and right. While I'm not ruling out the possibility of slightly improved outfield defense in 2005, trading your best defensive outfielder (relative to position) and downgrading in the two other outfield spots is not a surefire recipe for reconfiguring your outfield in the name of game altering defense.
posted by Jeremy @ 7:45 PM
0 Comments:
Post a Comment
<< Home