- They focus on the predicted number of electoral votes for each presidential candidate. That’s the wrong method. They should look state by state.
- They don’t tell the public that InTrade just aggregated the state polls, which were quite accurate —giving a little plus to Obama in Missouri, where the polls were saying that “Missouri is too close to call”.
-
Note that the way they over-sold InTrade’s predictive power in 2004 was different —they, then, were trumpeting that InTrade got all the 50 states right. What explains that they don’t look state by state, this time?
They have changed their method because it better fits their ideological agenda of over-selling the prediction markets.
-
Pingback: Are prediction markets useful? | Midas Oracle .ORG
Pingback: Are prediction markets useful to you? | Midas Oracle .ORG