1. Gladwell: Marbury over there on the right drew my attention to the Gladwell-Pinker dabate as to whether the draft selects rationally with respect to quarterbacks. Gladwell says, basically, no, and Pinker says, basically, it doesn't do as badly as Gladwell says, and Gladwell, basically, skewers this rhetorically more than he skewers it actually.
I remember reading around this when the article first appeared in the New Yorker. My take, since you are so keen to hear it, was that there is a correlation which Gladwell skates over because he wants to make a point whch is rhetorically defensible enough to be getting on with, since his life's work, which he has spent over 10,000 hours on and so he is basically a genius at, is to generalise surprisingly, interestingly and with enough truth to provoke people to readjust their assumptions.
2. Hardly any Nemesis space, ironically, since Mike Tanier has his best week of the season. Seriously, just go and read it. Oh, ok, one highlight about a new head coach:
“I had all the answers when I was an assistant,” he said. “I wish I had more answers now.” It’s easy to have all the answers when none of the questions asks how to build an offense around Jamaal Charles.
3. How is your nascent top statistical study going?
You are too kind. Anyway, I have produced a set of numbers based on last week's predictions. I made some good calls - Jerricho Cotchery for instance, and keeping Miles Austin out of my top ten. I made some bad calls everyone made - Marques Colston and Hines Ward. I made some special bad calls all my own - thinking Dallas and Green Bay would have a shoot out and scoring various players accordingly.
What I have done is created a single number from my Top 10s by adding the difference between my predictions and the actual results. I have then done the same for ESPN's aggregate prediction, Matthew Berry's prediction, a robot strategy based on feeding back in last week's Top 10. Next week, I will add in a robot strategy based on cumulative season form.
The robot killed everyone on wide receiver by not picking Colston. Because I am using CBS stats, which are the only ones easily availably, I am making the not-totally-common-sense decision to rate Colston 567, like they do, because he lost a yard on the night and was therefore worse than a zillion people who didn't play.
Anyway:
QUARTERBACKS
ESPN - 68
BERRY - 77
HUDSON - 92
ROBOT 1 - 94
RUNNING BACKS
ESPN - 157
BERRY - 162
ROBOT 1 - 181
HUDSON - 199 (Mendenhall, in particular)
WIDE RECEIVERS
ROBOT 1 - 210
HUDSON - 733
BERRY - 790
ESPN - 795
Even without the Colston issue, the Robot would have won here. WRs are the most random scorers; I did best here; seems likely that luck favours you more in more unpredictable games. The sample size is very small so far. It will take several weeks before the really stop statistical journals start paying attention.
My predictions this week are, and now I've got a stinging headache so I can't bear to look at my computer for another second:
QUARTERBACKS: Brees, Warner, Rodgers, Big Ben, Peyton, All the Philip Rivers Run Into the Sea and Yet the Sea is Not Full, Brady, Schaub, Favre, Palmer
RUNNING BACKS: Jackson, Peterson, My Fair LaDainian Tomlinson, MJD, Jones, Addai, Wells, Johnson, Mendenhall, Forte
WIDE RECEIVERS: Rice (really? What am I thinking?), Fitzgerald, Ward, Andre the Giant, Wayne, White, Colston, Welker, Jackson, 85
Vision. Blurring.
2 comments:
since his life's work, which he has spent over 10,000 hours and so he is basically a genius at, is to generalise surprisingly, interestingly and with enough truth to provoke people to readjust their assumptions
Excellent insight, and it made me laugh a lot too.
It is a good line. Various other people have said it first, though.
Post a Comment