Talk:Favourite-longshot bias

Opening header
Added Sobel and Raines as the seminal work with respect to the information model.

The What's It To Me/Risk Tolerance Threshold Hypothesis of Herb Conner

The article attributes the apparently irrational overbetting on the longshot (based on payback probability) to "risk-loving behavior, or simply inaccurate estimation", and cites references. Without being able to cite supporting reference, IMHO, there is a clearer motive more likely to be at work for a 'rational gambler' to pick a longshot despite low probability.

The usual statistician asks about win/loss odds and payback. But the gambler asks a different question: 'What is the payback worth to me?'  A gambler plans to bet $2 on each of 8 races, and feels that the $16 risk is within his Risk Tolerance Threshold.

If a $2 bet pays $4 @ 2:1 and $202 @ 100:1, even though the longshot odds may be much worse than the payback, a $2 gain on one 8 race card might only pay to offset another $2 loss. It may thus be below the gambler's threshold for a Pay-off of Useful Value.

A $200 win might pay for the gambler's whole day at the track and then some. Likewise, buying 100 Powerball tickets when the pot grows to > $100MM is a poor bet {P win = (1 - P lose)}. But, since a win would change the bettor's life enormously, and since having 1 ticket has about the same chance of losing as having 100 tickets,one ticket may be a good bet. The cost of buying one ticket may be effectively below the gambler's risk threshold, and thus is essentially a risk-less and rational bet.

Htconner (talk) 00:57, 25 October 2012 (UTC) Herb Conner