Peeps Place, or anyone else interested, do you think that we could use a prediction market to provide a measure of trust for bitcoin sportsbooks?  This would provide a means for skeptical bettors to hedge against the risk of fraud. 

One problem with this plan is whether anyone would be willing to bet that the book was not going to disappear (or whatever the contract specified).  At a minimum, the sportsbook itself should be willing to provide some bids to help build confidence.  After all, if they are legit, they would be betting on an outcome that they control.  The only excuse they would have to not do this is a desire not to tie up the capital.  However, they could roll this over as contracts expired.  Any sportsbook lacking the capital required to do this should be extremely worrisome for that reason alone.

An additional issue with this plan is, of course, the reliability of the prediction market itself.  There seem to be a few of these that have now emerged as fairly stable.  Some, such as Predictious, have also demonstrated a level of transparency that is, so far, unheard of among the sportsbooks themselves.  (At least they have an address, a verifiable company name and a legal jurisdiction!)  Ironically, some of the legal limitations that are a principal reason for both sportsbook owners and players to prefer anonymity might not apply to a bet on whether the sportsbook would "go Madoff".

The final challenge I see would be structuring the contract in an objective way, allowing verification without requiring the prediction market to launch a huge investigation.   This is an issue, but I think that it could be addressed.

A couple of near-term examples would be:

  "Nitrogensports / Anonibet to pay out 20 BTC World Cup contest bonus." 

Since these books have taken pains to provide transparency to these contests, it should be possible to verify these outcomes.  In effect, the bonus amount could be held in escrow by the market. Players participating in these contests could do so with confidence, and could hedge against the fraud risk, if desired.

Has this been suggested, or even tried, before?