Hacker News Reader

I've also concluded that the lowest chance (and highest variance) is the best strategy. Even though the expected value is the same for all options, as another comment says.

This means that in a real life casino roulette one should bet (the minimum bet) on a single number and stop after the first win (or after running out of money).

Did simulate this stuff https://github.com/sashabaranov/betting/blob/master/Betting....

2% looks like good short-term strategy

Cool simulation, I guess it could be extended to generate more than 2 lines per config and aggregate all lines, and then it would likely converge to the respective binomial distribution stats on https://en.wikipedia.org/wiki/Binomial_distribution.

Yep very interesting to let it run for a while on different choices even though the expected value of each one is 0.95