I just reviewed all 50 responses to my original Ukraine prediction challenge. Seven comments did not make unconditional predictions, and two were obvious jokes. Here’s how I scored the remaining 41 sets of predictions:
1. My last post named six major facts about what happened during the last year:
2. Responses got one point for correctly predicting each of these facts.
3. Response lost one point for predicting anything contradictory to each of these facts.
4. Responses were neither rewarded nor penalized for failing to address these six points, or for making other predictions. My apologies to everyone who made correct predictions I didn’t score; many made good points, but I wanted to keep scoring simple and consistent.
5. I tried to grade the responses blind, but only via deliberate inattention rather than a serious blinding protocol.
The resulting histogram has a mean of -.29 and a standard deviation of 1.08, with a maximum of +2 and a minimum of -4.
The winner of the Ukraine Prediction Challenge was Hansjörg Walther, with a score of +2 out of a possible +6. His original forecast:
I gave Walther one point for calling the formal annexation of Crimea, and another point for predicting no action by Western governments serious enough to lead to NATO fatalities, with zero lost points for contradicting any of the the other four major facts.
I suspect that Tetlock will not be surprised by the results. If he’d like to comment, you’ll be the first to know.