Most Productive Outfielder in 2005
Title Track
I assert that the most productive outfielder in 2005 has been Brian Giles. hank, a frequent Ducksnorts commenter, asserts that Andruw Jones, Vladimir Guerrero, Carlos Beltran, Manny Ramirez, Carlos Lee, Adam Dunn and Ken Griffey Jr. have all been better. I intend to determine who is correct. Obviously, I am biased, but I will do my best to remove that bias. I imagine our three readers will point out any errors I make. I encourage that.
We will begin by determining what statistics to use. Rather than limit the discussion to one stat, we will look at several and try to draw conclusions from them. Win Shares (WS), Wins Above Replacement Player (WARP) and Net Runs Above Average (NRAA) are all designed to capture a player's entire value. Value Over Replacement Player (VORP), Equivalent Average (EQA), and the traditional batting line of Batting Average (BA), On-Base Percentage (OBP) and Slugging Percentage (SLG) all track only offense, but we will also look at those.
NRAA and BA/OBP/SLG will both use Davenport Translations in order to account for park factors.
Win Shares sees Giles and Dunn as equals and clearly ahead of the pack. WARP calls it cleanly for Giles running just ahead of Andruw Jones and well ahead of the rest of our sample. NRAA, on the other hand, prefers Manny Ramirez and Vladimir Guerrero. Win Shares and WARP are both essentially counting stats meaning that playing time matters. NRAA, however, is a rate stat and accounts for the part of the difference.
What can we conclude from this (here's where it gets subjective)? WS and WARP both have Giles in first or tied for first. Thus, both would seem to back up my initial assertion and contradict hank's. NRAA, though, leaves Giles in fourth behind Dunn, Ramirez and Guerrero, which would seem to contradict my assertion and although it backs part of hank's assertion, four of the seven players he claimed were better than Giles still fall behind him in the stat that favors him the least of the three.
VORP prefers both Dunn and Jones to Giles by a decent amount. Giles ranks #1 in EQA by a not insubstantial distance. Batting Average, the least valuable of the stats mentioned in this article gives the top slot to Guerrero followed by Giles as the only other ".300-hitter." On-Base Percentage, widely considered and statistically proven, the most valuable of the three traditional rate stats has Brian Giles with an absolutely massive lead over the pack. Slugging Percentage gives the crown to Ramirez, though Guerrero is only .001 behind with Dunn and Jones also very close.
VORP clearly shows Dunn to be the best. EQA does the same for Giles. And, although some may disagree, BA/OBP/SLG also favors Gilly. On-Base Percentage has been considered anywhere from two to three times more valuable than slugging percentage and at the low estimate, Giles blows the rest of the group away.
I think we've managed to show that by most ways of measuring productivity, Brian Giles is the most productive outfielder. However, one could reasonably make a case for Adam Dunn and to a lesser degree Vlad Guerrero or Manny Ramirez. That being said, one cannot present a stronger case than exists for Brian Giles.
I'd love to hear any differing opinions or even from those who agree. Thanks.
Lance says to hell with objective statistical analysis, use your eyes:
(Lance: I never said any such thing. I have been a SABR member and a devotee of objective statistical analysis since Richard was a zygote. Part of evaluating a player's value ought to involve seeing him play, however.)
I know, I know. It's not like I just left it at that. I included exactly what you said.
Giles' offense is likely NOT superior. The value of a HR, especially in a mediocre offense, cannot be overstated. A player who has just homered CANNOT BE LEFT ON BASE. You underestimate the impact of homers on actual runs scored. A home run is not a potential component of a run, IT IS A RUN, and is not dependant upon the contributions of others.
And the difference between Giles' and Jones' defensive contributions are so far beyond what your carefully-chosen-statistic illustrates, it's comical.
The context of their different position, ballpark, and pitching staffs played behind is simply not adequately evaluated by ANY defensive metric, no matter how much you may have fallen in love with said metric. Occcasionally, you have to trust your eyes. This is one of those times.
If you throw stats out there in hope that they will intimidate those who are less astute than you, eventually those stats turn to alphabet soup.
That being said, I still believe that, eventually, we'll be able to fully evaluate players based upon their individual numbers. It is foolhardy, however, to think that the metrics available to us now come close to telling the whole story. Until we can add the sum of our favorite stats on a team to unfailingly predict the won-loss total of that team, we are clearly missing SOMETHING.
End of rant. Don't kick me off the blog, please, Richard.
For the record, Lance has not and will not be kicked off the blog. ;-)
hank (do you want that in lower case for a reason?) responds...
Most productive outfielder. Hmmmm. That depends entirely upon how you define "productive".
If we are to believe that The Most Productive Outfielder is the player with the highest "win shares", or should that be "wins above replacement player", or maybe "Net Runs above average"? Also worthy of consideration, apparently, is "Value over replacement player" and Equivalent average".
The traditional methods of measuring performance are not valid unless we use the "Davenport Translations" to allow for park factors?
In suggesting that Giles could hardly be the most productive outfielder in baseball, I carelessly tossed out a few names without research and was immediately admonished for this transgression.
I do not consider these above named stats are valid to determine the most productive outfielder.
Again I ask, what is PRODUCTIVE?
I believe that runs and runs batted in are the final say in productivity. Production is: Scoring runs for the team. Both scoring the runs himself and driving in runs, in my opinion, should determine the most productive player.
If you want to talk about the Most Valuable Player, then a lot more factors must be considered but for now we will just concern ourselves with production.
Brian Giles scored 63 runs in 444 plate appearances and drove in 56. Subtracting his homers(so they wont be counted twice) it leaves him with 109 runs produced in 444 plate appearances or a run produced every 4.07 plate appearances.
some others for consideration:
My bottom line is to be the most productive outfielder you must be the best at producing runs for your team. Although Giles is a very good player and could be in the running for MVP of not only the Padres but the NL west div, I would say Manny Ramariz [sic] is the most productive outfielder.
(sorry the blog wont accept my tags for columns)
Added them for you... and I'll grant you that an argument can be made for Manny, though I don't believe you have made it. You've made a strong argument for the fact that he hits in the best lineup, though.
Kevin responds to hank...
You've got to be kidding. Runs and RBIs? Come on.
And if you use that, why remove home runs? Runs produced measured both parts of producing the run -- getting on base to score and knocking the runner home. Why penalize the player who "produced" both parts of the equation?
I tend to agree.
Lance: As do I. The huge value of home runs warrants being counted twice. But hank, runs & RBI?!? You're kidding, right?
Kevin quotes James...
Yes, we know a home run counts as a run and an RBI. But to subtract the home runs out is a flawed way of doing things, as Bill James said in his 1984 Baseball Abstract:
"My problem with "Runs Produced" has to do with this bit about subtracting the player's home run count? What you're figuring here isn't whole runs produced, it's half runs produced. One player provides the first half of the run--the on-base part of it; the A Factor; the run scored--and the other provides the second half of the run--the advancement part of it; the B Factor; the run batted in. Each is credited with one run produced. But when a player provides both the on-base act and the advancement act, he receives no more credit than if he had done only one. Does this make sense?"
Edit: One oversight, Miguel Cabrera is far and away the most productive outfielder this year.
I assert that the most productive outfielder in 2005 has been Brian Giles. hank, a frequent Ducksnorts commenter, asserts that Andruw Jones, Vladimir Guerrero, Carlos Beltran, Manny Ramirez, Carlos Lee, Adam Dunn and Ken Griffey Jr. have all been better. I intend to determine who is correct. Obviously, I am biased, but I will do my best to remove that bias. I imagine our three readers will point out any errors I make. I encourage that.
We will begin by determining what statistics to use. Rather than limit the discussion to one stat, we will look at several and try to draw conclusions from them. Win Shares (WS), Wins Above Replacement Player (WARP) and Net Runs Above Average (NRAA) are all designed to capture a player's entire value. Value Over Replacement Player (VORP), Equivalent Average (EQA), and the traditional batting line of Batting Average (BA), On-Base Percentage (OBP) and Slugging Percentage (SLG) all track only offense, but we will also look at those.
NRAA and BA/OBP/SLG will both use Davenport Translations in order to account for park factors.
| Player | WS | WARP | NRAA |
|---|---|---|---|
| Giles | 22 | 6.5 | 30.1 |
| Jones | 16 | 6.3 | 38.1 |
| Guerrero | 13 | 4.4 | 21.8 |
| Beltran | 16 | 3.4 | 39.8 |
| Ramirez | 19 | 4.4 | -0.5 |
| Lee | 18 | 3.9 | 39.1 |
| Dunn | 22 | 5.2 | 21.8 |
| Griffey | 17 | 4.5 | 31.8 |
Win Shares sees Giles and Dunn as equals and clearly ahead of the pack. WARP calls it cleanly for Giles running just ahead of Andruw Jones and well ahead of the rest of our sample. NRAA, on the other hand, prefers Manny Ramirez and Vladimir Guerrero. Win Shares and WARP are both essentially counting stats meaning that playing time matters. NRAA, however, is a rate stat and accounts for the part of the difference.
What can we conclude from this (here's where it gets subjective)? WS and WARP both have Giles in first or tied for first. Thus, both would seem to back up my initial assertion and contradict hank's. NRAA, though, leaves Giles in fourth behind Dunn, Ramirez and Guerrero, which would seem to contradict my assertion and although it backs part of hank's assertion, four of the seven players he claimed were better than Giles still fall behind him in the stat that favors him the least of the three.
| Player | VORP | EQA | BA/OBP/SLG |
|---|---|---|---|
| Ramirez | 36.7 | 0.319 | .275/.383/.603 |
| Guerrero | 37.4 | 0.321 | .323/.385/.602 |
| Dunn | 44.8 | 0.318 | .250/.386/.597 |
| Giles | 40.8 | 0.328 | .303/.442/.523 |
| Griffey | 40.4 | 0.300 | .280/.363/.542 |
| Jones | 43.0 | 0.301 | .268/.355/.590 |
| Lee | 34.3 | 0.288 | .267/.336/.532 |
| Beltran | 14.6 | 0.259 | .263/.316/.437 |
VORP prefers both Dunn and Jones to Giles by a decent amount. Giles ranks #1 in EQA by a not insubstantial distance. Batting Average, the least valuable of the stats mentioned in this article gives the top slot to Guerrero followed by Giles as the only other ".300-hitter." On-Base Percentage, widely considered and statistically proven, the most valuable of the three traditional rate stats has Brian Giles with an absolutely massive lead over the pack. Slugging Percentage gives the crown to Ramirez, though Guerrero is only .001 behind with Dunn and Jones also very close.
VORP clearly shows Dunn to be the best. EQA does the same for Giles. And, although some may disagree, BA/OBP/SLG also favors Gilly. On-Base Percentage has been considered anywhere from two to three times more valuable than slugging percentage and at the low estimate, Giles blows the rest of the group away.
I think we've managed to show that by most ways of measuring productivity, Brian Giles is the most productive outfielder. However, one could reasonably make a case for Adam Dunn and to a lesser degree Vlad Guerrero or Manny Ramirez. That being said, one cannot present a stronger case than exists for Brian Giles.
I'd love to hear any differing opinions or even from those who agree. Thanks.
Lance says to hell with objective statistical analysis, use your eyes:
(Lance: I never said any such thing. I have been a SABR member and a devotee of objective statistical analysis since Richard was a zygote. Part of evaluating a player's value ought to involve seeing him play, however.)
I know, I know. It's not like I just left it at that. I included exactly what you said.
Giles' offense is likely NOT superior. The value of a HR, especially in a mediocre offense, cannot be overstated. A player who has just homered CANNOT BE LEFT ON BASE. You underestimate the impact of homers on actual runs scored. A home run is not a potential component of a run, IT IS A RUN, and is not dependant upon the contributions of others.
And the difference between Giles' and Jones' defensive contributions are so far beyond what your carefully-chosen-statistic illustrates, it's comical.
The context of their different position, ballpark, and pitching staffs played behind is simply not adequately evaluated by ANY defensive metric, no matter how much you may have fallen in love with said metric. Occcasionally, you have to trust your eyes. This is one of those times.
If you throw stats out there in hope that they will intimidate those who are less astute than you, eventually those stats turn to alphabet soup.
That being said, I still believe that, eventually, we'll be able to fully evaluate players based upon their individual numbers. It is foolhardy, however, to think that the metrics available to us now come close to telling the whole story. Until we can add the sum of our favorite stats on a team to unfailingly predict the won-loss total of that team, we are clearly missing SOMETHING.
End of rant. Don't kick me off the blog, please, Richard.
For the record, Lance has not and will not be kicked off the blog. ;-)
hank (do you want that in lower case for a reason?) responds...
Most productive outfielder. Hmmmm. That depends entirely upon how you define "productive".
If we are to believe that The Most Productive Outfielder is the player with the highest "win shares", or should that be "wins above replacement player", or maybe "Net Runs above average"? Also worthy of consideration, apparently, is "Value over replacement player" and Equivalent average".
The traditional methods of measuring performance are not valid unless we use the "Davenport Translations" to allow for park factors?
In suggesting that Giles could hardly be the most productive outfielder in baseball, I carelessly tossed out a few names without research and was immediately admonished for this transgression.
I do not consider these above named stats are valid to determine the most productive outfielder.
Again I ask, what is PRODUCTIVE?
I believe that runs and runs batted in are the final say in productivity. Production is: Scoring runs for the team. Both scoring the runs himself and driving in runs, in my opinion, should determine the most productive player.
If you want to talk about the Most Valuable Player, then a lot more factors must be considered but for now we will just concern ourselves with production.
Brian Giles scored 63 runs in 444 plate appearances and drove in 56. Subtracting his homers(so they wont be counted twice) it leaves him with 109 runs produced in 444 plate appearances or a run produced every 4.07 plate appearances.
some others for consideration:
| Player | Runs | HR | RBI | PA | PER RP |
|---|---|---|---|---|---|
| Damon | 83 | 8 | 51 | 456 | 3.61 |
| Dunn | 75 | 31 | 72 | 426 | 3.67 |
| Cabrera | 75 | 23 | 76 | 440 | 3.43 |
| Ortiz | 75 | 25 | 88 | 454 | 3.29 |
| Suzuki | 74 | 9 | 42 | 466 | 4.27 |
| Bay | 72 | 20 | 55 | 454 | 4.24 |
| Sheffield | 71 | 21 | 81 | 440 | 3.35 |
| Ramirez | 69 | 29 | 97 | 408 | 2.97 |
| Jones | 62 | 32 | 80 | 436 | 3.89 |
| Lee | 58 | 26 | 89 | 455 | 3.82 |
| Guerrero | 57 | 21 | 70 | 362 | 3.41 |
My bottom line is to be the most productive outfielder you must be the best at producing runs for your team. Although Giles is a very good player and could be in the running for MVP of not only the Padres but the NL west div, I would say Manny Ramariz [sic] is the most productive outfielder.
(sorry the blog wont accept my tags for columns)
Added them for you... and I'll grant you that an argument can be made for Manny, though I don't believe you have made it. You've made a strong argument for the fact that he hits in the best lineup, though.
Kevin responds to hank...
You've got to be kidding. Runs and RBIs? Come on.
And if you use that, why remove home runs? Runs produced measured both parts of producing the run -- getting on base to score and knocking the runner home. Why penalize the player who "produced" both parts of the equation?
I tend to agree.
Lance: As do I. The huge value of home runs warrants being counted twice. But hank, runs & RBI?!? You're kidding, right?
Kevin quotes James...
Yes, we know a home run counts as a run and an RBI. But to subtract the home runs out is a flawed way of doing things, as Bill James said in his 1984 Baseball Abstract:
"My problem with "Runs Produced" has to do with this bit about subtracting the player's home run count? What you're figuring here isn't whole runs produced, it's half runs produced. One player provides the first half of the run--the on-base part of it; the A Factor; the run scored--and the other provides the second half of the run--the advancement part of it; the B Factor; the run batted in. Each is credited with one run produced. But when a player provides both the on-base act and the advancement act, he receives no more credit than if he had done only one. Does this make sense?"
Edit: One oversight, Miguel Cabrera is far and away the most productive outfielder this year.
posted by Richard B. Wade at 9:03 PM
4 Comments:
As a follower of sorts of Win Shares, I looked up the Win Shares leaders among outfielders and found that many of the top guys are not in your study.
Sheffield is No. 1 with 23 win shares. Edmonds is No. 4 with 21. Bay is No. 5, Cabrera is No. 7, Abreu is No. 8 and Johnny Damon is No. 9.
Beltran and Griffey should not have even been in the study. Rich, I know you just picked the guys that Hank pointed out, but I wanted to mention Sheff and Edmonds as elite guys also.
Jones doesn't do well at all in Win Shares. I'm not sure why, other than the on-base is not great.
I think a great case can be made and has been made for Giles as the best OF this season. It would be nice to have a guy like Andruw, Sheff, Dunn or Manny that made Petco look small. Not instead of Giles, but along with him.
All offseason, I was kept throwing out a Nevin for Andruw trade. At that time, before Nevin lost his skills and Andruw had a career year, it didn't seem laughable.
You've got to be kidding. Runs and RBIs? Come on.
And if you use that, why remove home runs? Runs produced measured both parts of producing the run -- getting on base to score and knocking the runner home. Why penalize the player who "produced" both parts of the equation?
Yes, but hank, that's on a team level. A runner on first base has no control over whether the guy after him GIDP or hits a home run. Can't you see that?
Yes, we know a home run counts as a run and an RBI. But to subtract the home runs out is a flawed way of doing things, as Bill James said in his 1984 Baseball Abstract:
"My problem with "Runs Produced" has to do with this bit about subtracting the player's home run count? What you're figuring here isn't whole runs produced, it's half runs produced. One player provides the first half of the run--the on-base part of it; the A Factor; the run scored--and the other provides the second half of the run--the advancement part of it; the B Factor; the run batted in. Each is credited with one run produced. But when a player provides both the on-base act and the advancement act, he receives no more credit than if he had done only one. Does this make sense?"
Post a Comment
<< Home