Reporting on market-generated predictions *before* the event outcome, and reporting on expired prediction markets *after* the event outcome are 2 difficult things, and neither Michael Robb not Panos Iperiotis have the first clue. Not saying that I hold all the keys, though —but I am learning.
P.S.: Your word is insulting. Mike is a young Padawan.
My blog articles that you are linking to are about the “failure” of the PM for the NH for Clinton. I also blogged about the Oscar’s markets and why “failing to fail is a failure”. No historic averages (in fact, I never wrote about historic averages, perhaps I should)
In both posts, you did *not* report about one expired prediction market. You wrote up an analysis about accuracy.
That is what I am trying to explain to you. Somebody has to *report* on whether one expired prediction market was spot on or not —and that can’t be scientific. It is done in a binary way (success/failure), because that is what people are accustomed to.
The result merely indicates that Michael Robb is an idiot, who does not understand how markets work.
Niall,
Reporting on market-generated predictions *before* the event outcome, and reporting on expired prediction markets *after* the event outcome are 2 difficult things, and neither Michael Robb not Panos Iperiotis have the first clue. Not saying that I hold all the keys, though —but I am learning.
P.S.: Your word is insulting. Mike is a young Padawan.
Someone does not have a clue on spelling my last name
I have written about outcomes of prediction markets and these posts happened to be some of my most popular ones.
You have written about “the historic average of similar markets”. Have you blogged about single, individual expired prediction markets?
My blog articles that you are linking to are about the “failure” of the PM for the NH for Clinton. I also blogged about the Oscar’s markets and why “failing to fail is a failure”. No historic averages (in fact, I never wrote about historic averages, perhaps I should)
http://www.midasoracle.org/2008/01/13/defining-probability-in-prediction-markets/
http://www.midasoracle.org/2008/01/14/prediction-market-efficiency-vs-prediction-market-accuracy/
In both posts, you did *not* report about one expired prediction market. You wrote up an analysis about accuracy.
That is what I am trying to explain to you. Somebody has to *report* on whether one expired prediction market was spot on or not —and that can’t be scientific. It is done in a binary way (success/failure), because that is what people are accustomed to.
After that, a deeper analysis can be published.