Prof Tomas Sandholm claims the challenge was a statistical tie. BrainsvsAI

On the twoplustwothread, one of the developers, Noam Brown claimed they calculated the 95% confidence interval based on the 80,000 mirrored hands that were played and it was +/- 10.35bb/100. The pros won by 9.16bb/100.

But isn't Noam wrong as they should have used a one-sided test based on the apriori hypothesis that the Humans are better than the AI. Hence Humans won at 95% confidence interval.

Secondly how does one statistically account for the mirrored hands. The 80,000 hands are not an independent sample. There are 40,000 independent hands.