Gambling is a little purer, thanks to Stanford. Statistics professors Persi Diaconis and Susan Holmes analyzed the prototype for a casino card-shuffling machine and concluded that an experienced player could correctly guess a forthcoming card more easily than a well-randomized deal should allow.
"We are not pleased with your conclusions," the professors report the machine manufacturer as replying, "but we believe them and that's what we hired you for."
Diaconis and Holmes, who partnered with USC math professor Jason Fulman for the analysis, advised that better randomization would occur each time by having the machine shuffle twice. But Holmes said by email that casinos are using other machines because "time is money in these situations and they wanted the fastest way to shuffle properly."
