More key analysis NOT from talking heads.

UK_WhoDat · 02-02-2010, 05:32 PM

Originally Posted by Danno

wow, thats a tough read. Gotta link?

Tru Dat! Good write up though and the original source diagrams are helpful.

Here's one for you:
Be Skeptical of My Super Bowl Prediction: Colts in a Close One
The Fifth Down Blog - NYTimes.com
By BRIAN BURKE January 28, 2010

How accurate should a good N.F.L. prediction system be? Picking simple winners and losers can’t be that hard. It’s obvious who the better teams are, right? What would you say, 85 percent? 80 percent? The answer may surprise you. The best systems out there, even the consensus favorites from betting lines, couldn’t break 70 percent this year.

When the Fifth Down approached me about posting my game probabilities last summer, I was a little apprehensive. Although my system had a strong record of success, I feared it could be due for a down year. Mark Twain wasn’t completely correct when he said, “It’s better to stay silent and look a fool, rather than speak and remove all doubt.” If you really want to remove all doubt about your foolishness, publish predictions in The New York Times. Fortunately, my model had its best year ever, exceeding 70 percent accuracy.

There will be some people in office pools that might beat 70 percent, but unless that person knows something the rest of us don’t, and they’re able to exceed 70 percent for more than one year, it’s almost certainly luck. If 100,000 monkeys randomly picked winners, at least one of them could be expected to be 70 percent accurate.

I think many people overestimate how easy it is to pick straight-up winners. Much of our overconfidence is probably due to what psychologists call hindsight bias, the inclination to think that past events were more predictable than they really were. People also tend to whitewash their own predictions, recalling them as more accurate than they truly were.

Publishing predictions in public each week has helped me appreciate concepts like hindsight bias. I’ve learned two big lessons over the past four seasons. First, our ability to predict the future, even in bounded and controlled events with plentiful information, is terrible. Second, we quickly forget how wrong our predictions were.
.......
On to the numbers for Super Bowl XLIV........ read rest in link

02-02-2010, 05:32 PM	#4
UK_WhoDat 1000 Posts + Join Date: Aug 2003 Location: Wherever I lay my hat Posts: 4,001	Re: More key analysis NOT from talking heads. Originally Posted by Danno wow, thats a tough read. Gotta link? Tru Dat! Good write up though and the original source diagrams are helpful. Here's one for you: Be Skeptical of My Super Bowl Prediction: Colts in a Close One The Fifth Down Blog - NYTimes.com By BRIAN BURKE January 28, 2010 How accurate should a good N.F.L. prediction system be? Picking simple winners and losers can’t be that hard. It’s obvious who the better teams are, right? What would you say, 85 percent? 80 percent? The answer may surprise you. The best systems out there, even the consensus favorites from betting lines, couldn’t break 70 percent this year. When the Fifth Down approached me about posting my game probabilities last summer, I was a little apprehensive. Although my system had a strong record of success, I feared it could be due for a down year. Mark Twain wasn’t completely correct when he said, “It’s better to stay silent and look a fool, rather than speak and remove all doubt.” If you really want to remove all doubt about your foolishness, publish predictions in The New York Times. Fortunately, my model had its best year ever, exceeding 70 percent accuracy. There will be some people in office pools that might beat 70 percent, but unless that person knows something the rest of us don’t, and they’re able to exceed 70 percent for more than one year, it’s almost certainly luck. If 100,000 monkeys randomly picked winners, at least one of them could be expected to be 70 percent accurate. I think many people overestimate how easy it is to pick straight-up winners. Much of our overconfidence is probably due to what psychologists call hindsight bias, the inclination to think that past events were more predictable than they really were. People also tend to whitewash their own predictions, recalling them as more accurate than they truly were. Publishing predictions in public each week has helped me appreciate concepts like hindsight bias. I’ve learned two big lessons over the past four seasons. First, our ability to predict the future, even in bounded and controlled events with plentiful information, is terrible. Second, we quickly forget how wrong our predictions were. ....... On to the numbers for Super Bowl XLIV........ read rest in link
	Our Saints, Our City, Our Soul Saints are individuals of exceptional holiness

Thread Tools
Show Printable Version Email this Page
Display Modes
Switch to Linear Mode Switch to Hybrid Mode Threaded Mode