SABR 101 - Relative and Absolute Scales (June 6, 2003)
The following was an e-mail I wrote a few months ago. I didn't intend to write this as an article, but I suppose that the information here might act as a good reference point for some readers. Consider it to be a primer on relative and absolute scales.
--posted by TangoTiger at 04:56 PM EDT
Posted 5:32 p.m.,
June 6, 2003
(#1) -
Patriot
It's amazing that someone as smart and prominent as Bill James either does not recognize the difference here or purposefully ignores it to advance his agenda.
Posted 11:51 p.m.,
June 6, 2003
(#2) -
Greedo
Oota goota Solo?
Posted 6:11 p.m.,
June 7, 2003
(#3) -
MGL
Good stuff Tango!
Patriot, would like to see you expand on that thought. I always thought that BJ was brilliant (and a very good writer), but like many brilliant people, he can be considerably one-track...
Posted 7:03 p.m.,
June 7, 2003
(#4) -
Patriot
I was referring(although I failed to mention so) to the section about negative value and playing time. In Win Shares, he says things like
"Total Baseball tells us that Kaline was better than Clemente, that Billy Herman was three times the player Buddy Myer was"
No, Bill, it's saying that Herman contributed three times as many wins above what an average player would have as Myer did.
"So, in the Linear Weights analysis, if you trade a nobody, a player who never plays in the major leagues, and you only get some guy who 1500 major league games and hits .260 with 180 homers--not an average outfielder--well, Linear Weights would say that you've made a bad trade."
Well, Bill, only a fool would use RAA to compare a guy who never played in the majors to one who did. Speaking of which, I have a higher TPR than Rey Ordonez. Only an idiot would take that to mean that I am in fact a better player.
Then he goes on to say, "I'm not saying that Linear Weights is a bad system. It simply is not designed to analyze trades, and it won't." True. But then why did you rip on it, Bill, as if you expected it to analyze trades for you?
"The fatal error in the method of measuring players as better or worse than .500 is that it forces one into the assumption that value consists in being better than average."
True, if the temperature being -2 degrees means that there is no heat.
Posted 1:45 a.m.,
June 9, 2003
(#5) -
Greg Tamer(e-mail)
"The fatal error in the method of measuring players as better or worse than .500 is that it forces one into the assumption that value consists in being better than average."
I'm confused...need help...how is this assumption forced?
Posted 7:43 a.m.,
June 9, 2003
(#6) -
tangotiger
Because James views "negative" value as being "bad", so bad that you'd be better off as 0 runs above average in 5 PAs, than 5 runs below average in 1000 PAs. But, the key to understanding this issue is the point I made in the article regarding the "key" point.
Posted 9:35 a.m.,
June 9, 2003
(#7) -
Greg Tamer(e-mail)
Was being a bit sarcastic in my last post. My thought is that how can Bill James believe he's forced into his assumption, and then on top of that, change it to "forces one", so that the reader will follow him down the error of his ways? But then looking at the statement again, "forces one into the assumption," I almost read it as that the writer of this statement feels the assumption is an incorrect one, and that is the fatal error, the forcing of the assumption. Maybe.
BTW, I understand the "key" point. Just nitpicking here and trying to determine if Bill James really has the opinion on this topic that Patriot and you maintain he has.
Posted 12:41 p.m.,
June 9, 2003
(#8) -
David Smyth
Bill James may be guilty of 'piling on' against LW too much. But the point he makes is correct. Sure, if you interpret LW in the accurate way, like Patriot and Tango do, then LW is just fine. But those lists in Total Baseball are implicitly intended to be a ranking of players according to their value. If not, can someone show where Palmer makes that clear?
Posted 12:59 p.m.,
June 9, 2003
(#9) -
tangotiger
But those lists in Total Baseball are implicitly intended to be a ranking of players according to their value
Palmer is wrong in how he sells Linear Weights by doing this.
James is wrong in saying that because he doesn't buy the TB ranking, then he can't possibly buy Linear Weights, and then go on and start telling you why he can't buy Linear Weights. James can't get past the concept of "zero" not meaning zero in an absolute sense, where zero is "absence of something". Zero is defined in LWTS as average.
The issue I have with James is not that he doesn't buy it, because I really don't care. I also don't care too much that he gives weak arguments against Linear Weights. My issue with James is that he has an enormous amount of influence (more than all of us put together), and the reader has a certain amount of trust in James, that they won't feel they have to do the dirty work to validate what James is saying, and that James then puts out these weak arguments that it takes us 20 years (and counting) to undo the damage.
That run-on sentence means: James has to be responsible with what he says, becauses people treat him as judge, jury and executioner. James derives (or at least derived) his income by getting people to buy into what he says, and he should be more responsible with his analysis.
Posted 2:18 p.m.,
June 9, 2003
(#10) -
Patriot
Measuring players' value from above average may not be the best option, but at least I can defend it as saying that it is his value above his opponent. The baseline(s) used in Win Shares are, as far as I know, completely unjustifiable in every way because 1)they are not really absolute values and 2)even if they were, they would need to be considered with losses. So BJ attacking the baseline of the Palmer system is really quite ironic. The truth is somewhere in the middle, but at least Palmer could explain what his is doing. James could not truthfully explain what the baseline of Win Shares is. We know what he passes it off as(absolute), but it is quite obvious that this is hocum.
And getting off the issue of which baseline is right or defensible, and back to my original point about James, is that that first quote that I posted proves that he either does not understand or purposefully does not state the truth about the average baseline. To say that "TB says X is 3 times the player Y is" is just false. Even a true absolute method would not be saying "X is 3 times the player"--it would be saying that "X is 3 times more valuable to a major league team". Whether you consider the semantics to be important is up to you. I do.
Posted 4:13 a.m.,
June 26, 2003
(#11) -
Sylvain(e-mail)
Ok, I'm not very familiar with Replacement Level, so I won't add any comments, but I found this article on Prospectus very interesting and thought it might fit pretty well here.
Link: http://premium.baseballprospectus.com/article.php?articleid=2032
S
Posted 10:56 a.m.,
November 12, 2003
(#12) -
tangotiger
Just bringing this one forward as a companion to Patriot's article.