Prediction Markets + Market Predictions = Collective Forecasting That Pays Off

User Interface & Target Audience: BetFair, TradeSports-InTrade, MatchBook, etc.

No Gravatar

Yahoo! research scientist David Pennock comes to the rescue of British prediction market consultant Jed Christiansen.

Yahoo! research scientist David Pennock comments:

I think Jed Christiansen is correct to a large degree. Betfair speaks the punter’s (gambler’s) language. TradeSports speaks Wall Street’s language. I have a bookie friend who upon first look at TradeSports couldn’t make heads or tails of it. Chris is right in that both BetFair and TradeSports perform the same service. However their target audience, at least initially, is different.

#1. BetFair and TradeSports-InTrade do perform the same service.

#2. Their user interface (how they display the prices) could be easily tweak so that any of these prediction exchanges (betting exchanges) express prices (which we interpret as probabilities, thanks to Pennock et al.’s works) in their competitor’s way. There are actually four ways to express prices: 0–100, American, fractional or decimal. It’s all equivalent and interchangeable. No need to make a big fuss out of it. (For how MatchBook managed to have an interface user displaying “American odds”, please see: Americans Need to Embrace Betting Exchanges as the Preferred Way to Bet.)

#3. BetFair did first target the sophisticated British horse race punters. TradeSports did first target the sophisticated American sports bettors (and was unsuccessful at targeting the American horse race punters for the reason that it would require proximity, I suppose). I don’t see that the target audience is “different”. In both cases, it’s the sophisticated sports bettors.

External Link: Nisan Gabbay on BetFair

Previous: BetFair Case Study – Betting Exchange – Prediction Markets

[…] Betfair on the other hand was built like a stock market exchange, where odds functioned as the share prices. […]

Previous: BetFair vs. TradeSports-InTrade

UPDATE: Jed Christiansen…

Chris, regarding your point #2, it’s in fact a lot harder than you think.

The reason is because in choosing odds or 0-100 scale, you’re choosing a step size, and the two types don’t match up. On a 0-100 scale, the traditional step size is 0.1, or 0.1%. On an odds exchange, the step size varies from 0.01 when trading near 1 to much higher when you reach odds of 1000-1. For fiduciary reasons, you need to choose one or the other as your base system, and then “translate” for the other audience, and that translation process won’t always be clean.

Again, it’s always possible to make it work, but it is NOT an “easy tweak.” (Like so many good ideas, the tough part is in the implementation.)

Leave a Reply

You must be logged in to post a comment.

Search

Post Categories