Can technology predict who will win a poker tournament?
May 9, 2016
Not everybody can be a Professional Poker player. You need to be calm, cool and collected. Strong emotional stability, discipline, and an aggressive yet patient ambitious nature are critical. If you can score a person on each of these traits, would you be able to predict if he is capable of winning a poker tournament? Probably, but how can this be done?
According to Social and Life Science research personalities are affected by genes. Facial features reflect our DNA, therefore facial analysis may be the key to analyzing a person’s personality.
If we go back to the professional poker player, when we analyze the faces of leading poker players, we find commonality in their facial features. Even if we can’t yet correlate each of these common features to various personality traits, we can with a high level of accuracy detect the ever powerful “Poker Face.”
Faception, a 500 startups company, can do this and more. To qualify for this batch, we were challenged by the 500 Startups team, with a blind test of 27 quality facial images of poker and non poker players and were able to accurately classify 25 of them (93% accuracy).
On March 14, YouStake, another 500 Startups company hosted a poker tournament at the SF 500 office. We took pictures of 50 participants and analyzed them using our machine learning technology. We than ranked them according to how they scored against matching the profile of a professional poker player. The results were sealed and to be revealed before the start of the final table. When the results were revealed, it turned out that our technology correctly picked 4 out of the remaining 7 players at the final table. Furthermore, 2 of the top 3 were winners. (Top pros Jamie Gold and Anthony Zinno were excluded).