As for Midas Oracle not believing in prediction market accuracy, he’s got some support. If you look at the HP sales forecast prediction markets, (one of the few that has published its results), you’ll see that in most cases their predictions were not significantly better than the official HP forecasts. While the PM predictions were better than the HP forecasts in 6 of 8 markets, they weren’t “better” enough to have influenced decision-making (at least not in my opinion), and many of the predictions had very large errors (>25%), when comparing with the actual outcomes.
Personally, I believe that prediction markets CAN be more accurate than other means of forecasting, but, for a variety of reasons, they haven’t been significantly more accurate, so far. I do agree that prediction markets can be more efficient and provide other benefits that traditional forecasting methods are unable to provide. LOTS more work needs to be done to improve prediction market results and usefulness.
I’m in the middle of reviewing the paper on the HP forecasts and will publish my comments soon.
>> “Personally, I believe that prediction markets CAN be more accurate than other means of forecasting, but, for a variety of reasons, they haven’t been significantly more accurate, so far.”
I know what you mean. You can also throw ignorance in there as one of the reasons.
I wasn’t suggesting that PM’s are (significantly) more accurate by definition even with the right conditions.
Here is a link to my review of the “research”.
http://torontopm.wordpress.com…..n-markets/
Bentley,
I will contact you soon by e-mail.
Thanks.
Bentley,
you should know that Midas Oracle does not believe in prediction market accuracy, only efficiency.
Sorry, have a nice life.
Thanks for caring. So far it’s not too bad.
As for Midas Oracle not believing in prediction market accuracy, he’s got some support. If you look at the HP sales forecast prediction markets, (one of the few that has published its results), you’ll see that in most cases their predictions were not significantly better than the official HP forecasts. While the PM predictions were better than the HP forecasts in 6 of 8 markets, they weren’t “better” enough to have influenced decision-making (at least not in my opinion), and many of the predictions had very large errors (>25%), when comparing with the actual outcomes.
Personally, I believe that prediction markets CAN be more accurate than other means of forecasting, but, for a variety of reasons, they haven’t been significantly more accurate, so far. I do agree that prediction markets can be more efficient and provide other benefits that traditional forecasting methods are unable to provide. LOTS more work needs to be done to improve prediction market results and usefulness.
I’m in the middle of reviewing the paper on the HP forecasts and will publish my comments soon.
>> “Personally, I believe that prediction markets CAN be more accurate than other means of forecasting, but, for a variety of reasons, they haven’t been significantly more accurate, so far.”
I know what you mean. You can also throw ignorance in there as one of the reasons.
I wasn’t suggesting that PM’s are (significantly) more accurate by definition even with the right conditions.
Bentley.
don’t let this Masse guy manipulate you.