hi
Just wondered whether anyone watched The code (predictability), fascinating especially when thought of in the context of trading...
The crowd behaviour and pattern modelling was interesting too..very smart guys
Regards
Peter
The code
-
- Posts: 4619
- Joined: Wed Mar 25, 2009 12:23 pm
Got it on the Sky Box to watch when I have chance, looking forward to it.
-
- Posts: 4619
- Joined: Wed Mar 25, 2009 12:23 pm
It was on the BBC so should be on their Iplayer.
-
- Posts: 309
- Joined: Thu Nov 04, 2010 7:02 am
Looks good. Hopefully shown soon here in Australia. Suppose I will have to be content with the Hollywood drama " Numbers " in the interim.
rg
rg
Here's a link relating to the program: http://www.bbc.co.uk/tv/features/code/
BTW, I question the common notion that the average of lots of independent guesses is likely to be highly accurate. Let's you were to map the guesses on a frequency histogram, with extremely accurate guesses at one end and extremely inaccurate guesses at the other. If normal distribution applies, then surely it follows that the average of the guesses will be nowhere near the extremely accurate guesses.
I think the theory put forward in this article offers a better explanation of why the Betfair markets are, on aggregate, so efficient: http://karmatics.com/docs/evolution-and ... rowds.html
Jeff
BTW, I question the common notion that the average of lots of independent guesses is likely to be highly accurate. Let's you were to map the guesses on a frequency histogram, with extremely accurate guesses at one end and extremely inaccurate guesses at the other. If normal distribution applies, then surely it follows that the average of the guesses will be nowhere near the extremely accurate guesses.
I think the theory put forward in this article offers a better explanation of why the Betfair markets are, on aggregate, so efficient: http://karmatics.com/docs/evolution-and ... rowds.html
Jeff
-
- Posts: 309
- Joined: Thu Nov 04, 2010 7:02 am
Nice article. Like Jimmy Wales' analogy of the 'sausage'. LOL.
rg
rg