Archive for February, 2005
It has been brought to my attention that the permalinks on this blog are not working for some people. The problem is, I can’t get them not to work; so, I’m not sure exactly how to find the problem and fix it. If anyone has any suggestions as to what the problem is and/or a possible fix, please let me know.
Sidle up to the bar graphs; chomp on the pie charts. Thank goodness that just when you would expect the statistics to become more complicated, they remind you of simpler days.
That’s how Alan Schwartz ends his NYT Keeping Score article on salary arbitration in MLB. Salary arbitration in baseball is an interesting process that employs “final offer arbitration” as a fall-back position to determine player’s salary if the club and player cannot agree on a deal. (See pages 13-19 of the CBA for the procedure.) In short, each side submits what they think a player ought to make. Each side is given an hour to make the case for its salary and a half-hour rebuttal. From this hearing the 3-member panel selects which of the two offers will be the player’s salary for the coming season. Both the players and teams choose from a list of arbitrators to keep or strike.
This is pretty important stuff, right? So, why are arbitration panels so ignorant when it comes to assessing the value of a player? Schwartz highlights the fact that the wealth of baseball knowledge from sabermetrics is virtually ignored.
Salary arbitration hearings, from which members of the news media and public are barred, are places where traditional statistics like batting average, runs batted in and pitchers’ records still reign supreme.
Sure, adding on-base percentage to slugging percentage has gained greater currency. In recent years, relievers’ inherited runners have been discussed. But the industry’s most promising newcomers – more esoteric (and many would say more enlightening) statistics like win shares or value over replacement level – remain on the bench.
Why is this? At a recent meeting of Econodorks I enjoyed a comment by Brad Humphreys about salary arbitration. While this was several months ago, the gist of his comment was that arbitrators are typically professional arbitrators with no particular knowledge of baseball. This just blows my mind. Isn’t the whole point of this process to determine s “fair market value” of a player’s contribution to the team? How can an arbitration panel largely ignorant of all but the most basic baseball statistics do this?
Here are some possible answers and why I think they are wrong:
- The time constraints of the hearing prevent an adequate presentation of sabermetric statistics, so lawyers stick to the basics.
I don’t think so. Schwartz describes arbitration rooms to be well-equipped for speedy delivery of relevant facts, “Exhibits can be as thick as phonebooks, each side’s back-room bunker filled with enough computers and photocopiers to rival a local Kinko’s.” Also, according to the CBA, “The aforesaid time limitations may be
extended by the arbitration panel in the event of lengthy crossexamination
of witnesses, or for other good cause.”
- Teams and players are ignorant of the knowledge of sabermetrics.
There’s too much at stake here. I don’t buy the fact that teams and players will remain ignorant of knowledge common to stathead-baseball fans given the monetary incentives to understand this stuff.
- People who are trained arbitrators simply are not stat-savvy.
Schwartz alludes to this in the article,”‘You’re dealing with nonbaseball people, so the basic stats – average, homers, R.B.I., on-base percentage – are the ones they understand most,” said Mike Hill, assistant general manager of the Florida Marlins.” But, I think arbitrators would learn this stuff and pick up on it quickly, even if initially ignorant. Courts hear expert testimony all the time on all sorts of statistical evidence that would overwhelm laymen.
So what do I think is the answer? Sabermetrics lacks an authority of expert opinion, even though the sabermetric approach is correct. Bill James knows a lot about baseball, and though I believe he has consulted on many arbitration cases, there is no reason for arbitrators to value his opinion over a guy who thinks batting average and RBIs are the best statistics. When you get down to it, James is just a guy who really likes baseball. I suspect any good lawyer might bring this up. I am exaggerating a bit here; while I suspect James did and would hold sway on such panels due to his enormous reputation and intelligence, he’s now off-limits thanks to his employment with the Red Sox. If you put a prominent hobbyist sabermetrician on the stand, is he really going to have more sway than some average jackass off the street who thinks getting hits is what really matters? What about compared to an ex-scout, manager, GM who has some inside credibility? While I think Tangotiger knows a lot more about baseball than a lot of “insiders,” I’m not so sure he would hold such sway in an arbitration hearing.
Awhile back, I had an idea for how to rank the greatest
wide receivers of all time. I’m not sure if it works or
By “work,” I
mean generate a set of rankings that would agree with anyone’s intuitive notions of what such a list ought to look
like. You tell me.
Wide Receiver is the only position where even small groups of players are actually competing against each other
under nearly identical circumstances. Domanick Davis and Brian Westbrook are competing for statistics under very
different sets of circumstances and for that reason it’s extremely difficult to say with any degree of certainty
who is better. Likewise, Rod Smith and Laveranues Coles are in different environments so simply comparing their
stats isn’t necessarily a reliable way of determining who’s better.
But the same does not apply to Rod Smith and Ashley Lelie. Smith and Lelie are working in the same system with the
same quarterback, the same offensive line, even the same game conditions. Raw numbers probably are a good
determine to what extent Smith is better than Lelie. Likewise, Coles and Rod Gardner can be fairly compared.
Every season, every team has a group of 3 to 5 guys that can, for the most part, be rank-ordered by their numbers.
situation is unique to wide receivers.
But how does this help us compare Rod Smith to Laveranues Coles? Think college football. USC didn’t play Auburn
So who was better? Well, we know USC is good because, among other reasons, they crushed
Oklahoma, who we suspect was
pretty good; they beat Texas, for example. We know
Auburn was good, in part, because they beat Tennessee, Georgia, and LSU, all solid teams. While there is
unfortunately no direct
evidence to help us settle the Auburn/USC debate, there are piles and piles of indirect evidence. Every game
played by either team, or the opponents of either team, or the opponents of those teams, serves as a tiny
sliver of indirect evidence about how good USC and Auburn were. And
very intelligent people have devoted lots of their time and talent to convincing computers to assimilate all
So why not put this technology to work ranking wide receivers? Rod Smith “played” Ed McCaffrey several times, and
McCaffrey was good. He also played Anthony Miller and Willie Green and Eddie Kennison (remember that?). And
McCaffrey has played Jerry Rice and Stephen Baker, Willie Green has played Mark Carrier and Don Beebe, Eddie
Kennison has played — well, who hasn’t Eddie Kennison played? Likewise, there is loads of indirect evidence —
mind you, much of it is extremely indirect — about how good Laveranues Coles is compared to Michael
Jackson and Marvin Harrison and Troy Brown and even Randy Moss.
A big tip of the cap to Wes Colley, who developed one of the computer schemes utilized by the BCS. He has
published all the details of his ranking system, and I think it’s
pretty nifty. So I’m going to modify it for my purposes here and use it to rank wide receivers. Here is a very
rough sketch of a few of the details.
1. I looked at all wide receivers whose career started in 1960 or later. Essentially, I treat every pair of WRs on
the same team in the same year as a “game.” I only looked at
receiving yards, although TDs and receptions could be easily integrated into the scheme. The winner of the game is
the guy who had a higher receiving yards per game average during that season. I do take into account margin of
2. Before scoring a game, I make an age adjustment. So, for example,
when Terrell Owens had 1451 yards and Jerry Rice had only 805 yards for the 2000 49ers, I tweak those numbers to
account for the fact that Owens was 27 and Rice was 38 at the time.
3. Just as college football ranking schemes have to decide how to deal with games against I-AA teams, I have to
decide how to treat the multitudes of receivers like Ron
Lewis and Jeff Campbell who are
basically irrelevant to the exercise of ranking the all-time greats. So I deemed a receiver to be relevant if he
is in the all-time top 300 in receiving yards and he averaged 50 yards per game in at least one season. All
the irrelevant receivers in a given year on a given team were combined to form one pseudo-receiver. So Randy Moss
in 1998 played games against not only Cris Carter and Jake Reed, but also Joe PseudoReceiver, who was a composite
of Chris Walsh, Robert Tate, and Matthew Hatchette.
4. Feed all these games into (my modified version of) Colley’s algorithm and let it work.
Before I end today’s entry with the list, keep in mind that a methodology stands or falls on its merits,
independent of the results it generates. If you like the methodology, you are not allowed to complain about the
list. If you don’t like the methodology, you shouldn’t even be looking at the list. But no fair changing your
mind after peeking.
That said, I promise that my next entry will be entitled Why in the World is Joey F. Galloway #9 on This
4. Tim Brown
5. James Lofton
6. Andre Reed
10. Harold Jackson
11. Art Monk
12. Cris Carter
13. Anthony Carter
14. Stanley Morgan
15. Irving Fryar
16. Rob Moore
17. Fred Biletnikoff
18. John Stallworth
19. Michael Irvin
20. Charley Taylor
21. Haven Moses
22. Wesley Walker
23. Eric Martin
25. Lynn Swann
26. Mel Gray
27. Gary Clark
28. Mark Duper
29. Jimmy Smith
30. Al Toon
31. Brian Blades
32. Alfred Jenkins
33. Andre Rison
34. Sterling Sharpe
35. Reggie Rucker
36. Charlie Joiner
37. Ken Burrough
38. Terry Glenn
39. Randy Moss
40. Otis Taylor
42. Gene Washington
43. Carroll Dale
44. Isaac Bruce
45. John Gilliam
46. Marvin Harrison
47. Drew Hill
48. Lance Alworth
49. Gary Garrison