|
While people play the picks IQ game we analyze their picks and pick history to arrive at a collective intelligence Picks IQ prediction for each match-up.
As another example involving the old "Inside the NFL" crew, consider the case where Bob had a picks IQ of 120, Dan had a picks IQ of 110, and Chris had a picks IQ of 90. Say they all made predictions on a Dolphins/Bengals matchup and Dan picked the Dolphins to win by 14, Chris picked the Bengals to win by 3, and Bob picked the Dolphins to win by 3.
A simple average prediction of all three picks would be as follows.
Average Spread Predicted = (14 - 3 + 3) / 3 = 4.7 (i.e. Dolphins win by 4.7 points)
However, the picks IQ prediction is actually a weighted average based on the Picks IQs of the players such that every 10 IQ points corresponds to a doubling of the weighting used. For example, if a picks IQ of 100 was assigned a weight of 1 then a picks IQ of 110 would have a weight of 2 (i.e. Dan), a picks IQ of 90 would have a weight of 0.5 (i.e. Chris), and a picks IQ of 120 would have a weight of 4 (i.e. Bob).
The Picks IQ weighted average predicted spread calculation would be as follows:
Picks IQ Prediction = (14*2 - 3*0.5 + 3*4) / (2 + 0.5 + 4) = 5.9 (i.e. Dolphins win by 5.9 points)
The Average Spread Predicted and the Picks IQ Prediction are shown at the top of every match-up. Note: These collective intelligence predictions are hidden until one day before game time, the time when the prediction window is closed.
|
