> 1/184 = .000000952 chance of that occurring. Probably not confirmation bias.
On any one set of four throws. Even if you used the smaller probability of the exact result (rather than grouping the two sets of green zeroes), its exactly the same as the chance of any other result of the four throws.
Sure, any four (or 4 groups of two in this case) has the same chance of appearing. But... these were a "special" 4 under special circumstances. Coincidence? No way!! I was watching her practice! Context is a big part here. I would love to have a big data set of wheel rolls from all over broken into individual dealers and wheels. Collected without knowledge of the dealer of course (which would likely change that which is observed).
On any one set of four throws. Even if you used the smaller probability of the exact result (rather than grouping the two sets of green zeroes), its exactly the same as the chance of any other result of the four throws.