I hope you like math!
Ex. Lets say a player has signed up through an affiliate whose going to be getting 35% of this players losses each month. This player is going to play just a single round of European Roullete each month for 1000$ by betting on either RED or BLACK. From the casino's standpoint, their either going to be recieving 650$ from this player in the event that he loses, which has about a 51.4% (=19/37) chance of happening..........Or their going to lose 1000$, which has a 48.6% (= 18/37) chance of happening. This is hugely negative EV for the casino as EV = -134$ = (650$ x 51.4%) - (1000$ x 48.6%).
FYI; The more wagers a player makes, the better it is for a casino and I suspect most online gamblers don't make JUST A SINGLE WAGER per month, unless they are colluding with an affiliate for joint profit.
On the other hand, scenarios like the example listed below are very frequent. In most instances, the casino still makes a slight profit in the presence of lots of wagering however some games can still net them a loss via their affiliates.
................................................................................
Suppose an online player signed up through an affiliate whose going to get 35% of his/her losses each month. This player is also going to be sticking with playing something like full-pay JoB video poker say (though most vp would suffice in this example)............Now on most months, this player will be donating his/her money to the casino and the affiliate he/she signed up through, however on the months that this player comes out ahead (sometimes he's going to hit the RoyalFlush and likely be in the money for that particular month!), only the casino suffers the loss........
Calculating the EV for the casino in this example is MUCH more difficult, but here goes!
Lets say the player wagers 10,000$ each and every month on 9/6 JoB video poker and lets consider his play over one year under reasonable assumptions;
First off, lets assume he plays 9/6 JoB at 25cent denomination at full-pay, thus 1.25$ bets.
He will be making 8000 wagers each month in order to complete the 10,000$ wagering quota. Over one year, this will be 12 x 8000 = 96,000 wagers and from that it's reasonable to assume that this player could get two RoyalFlush wins. He will make 1000$ on each of these wins.
Now because we are assuming he hasn't gotten any more RoyalFlushs for the remainder of the year, then throughout this time, the player is actually going to be playing a 97.54% return game (as the RoyalFlush win accounts for 2% of the pay-out and we've just assumed he isn't going to win for the other 10 months).
Hence for 10 months of the year, the player's losses will be EV = 10,000$ x 10 months x 2.46% = 2460$. From this 2460$, the affiliate is going to take 861$. Thus casino profit during this time will be 1599$.
On the two months in which the player recieved a RoyalFlush, the loss to the casino is going to be 2000$ - (9,998.75$ x 2.46% x 2 months) = 1508$
Over the course of one year, the casino makes a slight profit of 91$(=1599 - 1508) off of this player.
.......................................................
Overall conclusion; The first example has shown that it's very possible for an online casino to lose money via its affiliate program if the player that has signed up through them is only making very few wagers each month and plays a game with a low(ish) house edge.........In the second example the casino makes only a slight profit over a full year (< 100$), but this profit can only be had if the player wagers frequently enough per month. Mind you, even if the player decides to wager frequently (ie > 8,000 wagers per month), there are video poker games out there that return more than 9/6 JoB. Many microgaming online casinos have the game All-Aces which has a return table north of 99.9%. If I redo the math in example two for this game, then the casino ends up losing 215$ over the course of a year. I suspect this loss would exist regardless of how often a player wagers each month.
Ex. Lets say a player has signed up through an affiliate whose going to be getting 35% of this players losses each month. This player is going to play just a single round of European Roullete each month for 1000$ by betting on either RED or BLACK. From the casino's standpoint, their either going to be recieving 650$ from this player in the event that he loses, which has about a 51.4% (=19/37) chance of happening..........Or their going to lose 1000$, which has a 48.6% (= 18/37) chance of happening. This is hugely negative EV for the casino as EV = -134$ = (650$ x 51.4%) - (1000$ x 48.6%).
FYI; The more wagers a player makes, the better it is for a casino and I suspect most online gamblers don't make JUST A SINGLE WAGER per month, unless they are colluding with an affiliate for joint profit.
On the other hand, scenarios like the example listed below are very frequent. In most instances, the casino still makes a slight profit in the presence of lots of wagering however some games can still net them a loss via their affiliates.
................................................................................
Suppose an online player signed up through an affiliate whose going to get 35% of his/her losses each month. This player is also going to be sticking with playing something like full-pay JoB video poker say (though most vp would suffice in this example)............Now on most months, this player will be donating his/her money to the casino and the affiliate he/she signed up through, however on the months that this player comes out ahead (sometimes he's going to hit the RoyalFlush and likely be in the money for that particular month!), only the casino suffers the loss........
Calculating the EV for the casino in this example is MUCH more difficult, but here goes!
Lets say the player wagers 10,000$ each and every month on 9/6 JoB video poker and lets consider his play over one year under reasonable assumptions;
First off, lets assume he plays 9/6 JoB at 25cent denomination at full-pay, thus 1.25$ bets.
He will be making 8000 wagers each month in order to complete the 10,000$ wagering quota. Over one year, this will be 12 x 8000 = 96,000 wagers and from that it's reasonable to assume that this player could get two RoyalFlush wins. He will make 1000$ on each of these wins.
Now because we are assuming he hasn't gotten any more RoyalFlushs for the remainder of the year, then throughout this time, the player is actually going to be playing a 97.54% return game (as the RoyalFlush win accounts for 2% of the pay-out and we've just assumed he isn't going to win for the other 10 months).
Hence for 10 months of the year, the player's losses will be EV = 10,000$ x 10 months x 2.46% = 2460$. From this 2460$, the affiliate is going to take 861$. Thus casino profit during this time will be 1599$.
On the two months in which the player recieved a RoyalFlush, the loss to the casino is going to be 2000$ - (9,998.75$ x 2.46% x 2 months) = 1508$
Over the course of one year, the casino makes a slight profit of 91$(=1599 - 1508) off of this player.
.......................................................
Overall conclusion; The first example has shown that it's very possible for an online casino to lose money via its affiliate program if the player that has signed up through them is only making very few wagers each month and plays a game with a low(ish) house edge.........In the second example the casino makes only a slight profit over a full year (< 100$), but this profit can only be had if the player wagers frequently enough per month. Mind you, even if the player decides to wager frequently (ie > 8,000 wagers per month), there are video poker games out there that return more than 9/6 JoB. Many microgaming online casinos have the game All-Aces which has a return table north of 99.9%. If I redo the math in example two for this game, then the casino ends up losing 215$ over the course of a year. I suspect this loss would exist regardless of how often a player wagers each month.