Here’s my thinking…
With slots, most players lose at a rate higher than the actual overall house edge, because somebody has to pay for those big occasional jackpots. So, most of your players will actually lose substantially more than the 3.71% margin they are predicting, even assuming that figure is correct overall.
Since this program wipes out negatives at the end of the month, when you do have a big winner, the effect is limited to one month. When negatives are wiped out, you can consider that a big boost in your commission rate.
The combination of those two things makes it more profitable to go with the old rev-share plan for slots players.
Now, for me things are different. They attribute a house edge of 1.26% to blackjack, while many of my players will do better than that because they’ll play accurate basic strategy. (Well, a lot of my players anyway. Thank goodness for those guys who ignore my advice and contribute a lot more to the bottom line!) If all my players played the best blackjack game offered at Microgaming, and played correct basic strategy, their real theoretical loss is only 0.36%, and RewardsAffiliates is paying me 0.44%. That’s a nice commission rate of 122%!
Now, in reality, my players will play a mix of games, and they won’t play perfectly. Still, for my particular situation, this sounds like a good deal, at least for my blackjack site.