> I'm surprised they aren't split testing the reward frequencies Of course this is done. It is just slower than in the web as other people has pointed out you

> I'm surprised they aren't split testing the reward frequencies Of course this is done. It is just slower than in the web as other people has pointed out you can't change things on the fly. I worked for the gambling industry and we had a statistics department that take care of this kind of things. You try different machines and you have different local maximums: - There is the machine for the slow player that is there all day. It gives back a lot of small rewards so people keep playing and playing. It doesn't makes so much money but makes sure that your section of machines (each section is from a different company) is never empty. - There are hard machines for places like Las Vegas. People bets hard and expects big colorful prices. - There are machines for the small town casinos. This machines are softer than the ones in Las Vegas. But also optimized for its target audience. Then there is the amount of noise, your section of machines should look like is giving prices all the time. The design of the cabinet (in Mexico they have a place to put plates as people eats there). Yes. It is serious business. And there are a million considerations that each company takes into account. A/B testing was there a lot before the web was a thing.