-
-
After skimmiing this article, I think the idea was: If someone runs a play money on-line betting shop (or stock market), the results of can be as good as real money situations, and be used as a predictve indicators for such situations. Sounds reasonable (if I got the hypothesis correct)
While I would agree with the idea, I don't see how that concept is generally any different than taking a public opinion poll.
Also, have you ever played an at the play money or freeroll tournament games at the on-line poker sites? In situations, where players are interested in actually learning the game, the table can mimic a real money game. However, when you get players who just want to irritate the table (for their own personal kicks) or understand this is just play money and make idiotic and silly plays at an acclerated rate (that they would never make with large amounts if real money at stake), the situation breaks down rapidly.
For example, in the early rounds of a freeroll, there are people who just love to go all-in and never stop (mimicking a tilt). The table just sits back and watches the guy accumulate tons of money, waiting to get something to challenge him with and take him down. If the guy "tilting" was smart - he then backs down and plays "correctly"; however, most people "tilting" this way eventually fade out, but I have seen situations where the guy wins the tournament or sit and go. Perhaps using the play money predictor model applied to freeroll on-line poker is not appropriate as a comparison, but it may bring up some ideas regarding error or bias.
I believe we have good play money prediction models regarding sporting events all over the place, and I look at them at times. The Prick contest is there (with totals next to picks) for all to see. I go to my free pick-em spread contest and watch how people are picking........ They are quite accurate at times, but I still try to do my handicapping homework as well......
Does money matter in a predictive market? I my humble (not being schooled in economics) opinion: No.
The players of a market (free or real) want to try to achieve some type of reward when they enter. Otherwise they would not have bothered to enter the market (play or real) in the first place.
Also the idea that:
"Theory suggests that real money may better motivate information discovery, while in play-money markets those with substantial wealth are those with a history of successful
prediction, suggesting potential for more efficient weighting of individual opinions."
Somehow, setting up a play money poll of NFL predictions and looking at those results to predict the outcomes does not weigh as heavily as listening to someone who would generate an educated (or even a reliable "insider") in depth analysis of that game, regardless of the prediction accuracy (or underlying rational) of the play money model. I imagine that that type of data is more appropriate towards stock market models, but the sports betting application would be more for line setting consultants, but people do buy from a variety of touts......Comment -
The obvious problem is that any market, even play money markets, is going to be part of the general football market. If there's a larger real money market then any smaller market will simply follow since there are people willing to bet it into line even for play money. If you wanted to REALLY test play money you would study a market with no parallels, so one trying to predict the outcome of something otherwise unavailable for wagering or buying/selling on a large scale and have a real money section and a play money section using randomly selected events.Comment
