Cumulative WR versus zeroing out - pros and cons (rep friendly)

[jstrike]

Point well taken. But that's what I've been trying to avoid. The things I've done have had real value and have been massively EV+... almost too much so. e.g. win the most on the weekend and have the winnings doubled, 2nd place gets 50% extra, 3rd place gets 25% extra, with no WR... and I only ever had 10 players or so in real competition for that at any time. Since there were freerolls too, one player (a pretty extreme bonus hunter, and blackjack card counter) got $1100 or so out of this without ever depositing.

In total, I've put more than 100% of the site's income back into promotions to players every month, for a year and a half. Another way of looking at it is, I've basically created a system that overall wins money from players who are there to have fun and don't care much about promotions (i.e. the players I want), and redistributes it to EV+ bonus hunters who never play in an EV- situation, while not making any profit myself. The worst part is, the only people who recognize that these things are EV+ or care are exactly the wrong people...they have no loyalty, they don't even like to gamble, and they only come if they think it's a sure thing. They won't bet a penny on chance after they clear it or win it.

Realistically, no site can create an EV+ promotion and make it available to everyone on the internet. It's gotta be available just to regular players. So I guess what I need to figure out is something that's meaty enough to be real but only applies to players willing to risk something for it, and players who've already shown loyalty. In that way, a points systems and Chris's after-WR bonus are both pretty good ideas I think.

The only problem with that is, it doesn't bring any new players in. So how do you craft a promo that's interesting to new players without attracting only sharks? I guess the solution for a lot of casinos is just to throw out these "meatless" promos you're talking about. OTOH it seems kind of silly and possibly unethical. It's a catch-22.

 

[binary128]

@Jufo,

Re: hallucinogens and backpacking. You should google the source of my signature line.

"I hate to advocate drugs, alcohol, violence, or insanity to anyone, but they've always worked for me".

hehe. Was it "geek site"? I know you said that in a moment of excitement, I remember you were trying to convince me of something

Wow, I just checked my email. That goes all the way back to September 2011 - almost a year and a half. Yeah, I was trying to convince you of something.

Chris

 

[jstrike]

So yeah, I decided yesterday I'm actually going to backpack around for the next few months. No joke. I'm working on a big project for a real video game company ("real" meaning one that actually pays for work), and I can do it from anywhere, but I need to get some time & space to think about things.

Somewhere in there I'm going to decide if this casino I built is worth continuing or if I should just cash all my players out, box up the games onto a hard drive and move on from it. I'm almost 33... I think it's time to get serious one way or another about things. I guess I could always come back to it in ten years and try to sell some of the slot math to Ballys or something. Doing this has been a lot of fun, but it's also been a lot of stress, and truth be known no one believed in it when it started ...and now that people are interested because it's in Bitcoin, it's getting some attention...I just did an interview with Business Week, lol, and I must be the lowest-earning CEO ever to get a quote in that magazine. But if it gets any bigger or becomes "mainstream", none of the people currently intrigued by it will be interested anymore, and the thankless work of running it just goes on and on.

I've never felt that good about taking a profit in this business by having a house edge on people, and I guess that's why I never made a profit. My reticence to earn money and promote it by doing things other casinos do, which I think amount to screwing people quietly, has made it mostly a business failure. I'm lucky it's broken even, but I seem to have this suicidal wish to give away everything it earns, and I don't know if I can get past that. I almost feel like it's just what was meant to be.

Anyway, there's the honest facts and I guess we'll see what happens, I guess for now I'll just get a mobile card and run it from the road...

 

[binary128]

Josh,

I read through a few of those old emails from Sep of 2011. I suggested an "option 1", which didn't work for you at all. But I then suggested:

"Then I think that you should do whatever it takes to grow your application into as much as you can do, take it as far as it can be taken. And when your app “goes viral” and takes over the world, then I can be one of the people who say “I knew him when …”

and

"I think that you should do everything that you know how to do to grow “your baby” to the fullest extent possible. Go for it. “Viral or Bust!”


I didn't realize that this much time has passed. I'd have to agree with you though - 2 years of operation and you've gone from "losing money" (your Blog entry "How to start a casino and lose money." from Jan 18 2012) to "breaking even". It may be the time to do a core re-think, but maybe its too early as well. Tough call, but it's yours to make.

Galewind is my 4th start-up - the other 3 went belly up. The second one took my house (as well as the house of my wife and 2 kids) and both cars - bankruptcy. That one was really horrible - one step short of living on the street.

For you the world, even in its current state of economic depression, remains your oyster. You've got a lot of options, and most of them seem to me to be good ones.

Chris

 

[clees42]

I am pretty sure that there is a solution to making the casino attractive to players without giving them straight +ev to all their bets.
The usual deposit bonuses are very problematic because either they are good for the players or they are good for the casino.
What this means is that either the casino is going to be unhappy or the players are unhappy.

What i am proposing is to stop with the usual deposit bonuses that apply to the all players, and concentrate on the losing players.

For example we have 2 players: player A and player B. Both players deposit 100 euro in the casino(without bonus),
and player A wins 150 euro and player B loses whole 100 euro.
Clearly player A is happy with his winnings even if he doesnt receive any special attention.
now for example, If player B got 20 free chip with 10XWR for a second change, he would be happy too.

I am pretty sure that the casino would be expected to make profit,
if players were to get free chip everytime they fall to zero with their deposit(assuming ofcos no withdraws)
The free chip could even be fixed % of the original deposit.

Someone could work out the numbers to make it -ev for the players in general,
but the players would still be happy.

 

[jstrike]

I am pretty sure that there is a solution to making the casino attractive to players without giving them straight +ev to all their bets.
The usual deposit bonuses are very problematic because either they are good for the players or they are good for the casino.
What this means is that either the casino is going to be unhappy or the players are unhappy.

What i am proposing is to stop with the usual deposit bonuses that apply to the all players, and concentrate on the losing players.

For example we have 2 players: player A and player B. Both players deposit 100 euro in the casino(without bonus),
and player A wins 150 euro and player B loses whole 100 euro.
Clearly player A is happy with his winnings even if he doesnt receive any special attention.
now for example, If player B got 20 free chip with 10XWR for a second change, he would be happy too.

I am pretty sure that the casino would be expected to make profit,
if players were to get free chip everytime they fall to zero with their deposit(assuming ofcos no withdraws)
The free chip could even be fixed % of the original deposit.

Someone could work out the numbers to make it -ev for the players in general,
but the players would still be happy.

I think this is where personal attention comes into play, and that's been good in the past. If I'm on and I see someone take a bad beat at blackjack, I tend to comp them afterwards. And that does do some good, players appreciate it. It's better when no one sees it coming. The 10% cashback is kind of like that, too... and in a way it sort of works for a very small group of players who've figured out how to extend their play time with it (by depositing middle-sized amounts, getting the cashback, getting back up and withdrawing...) But you make a good point, the bonuses should go more to the people who lose than the people who win. I'm not sure if it's better to codify that into a general rule, or just be there to give people something back when they're down...

 

[binshakindown]

I am pretty sure that there is a solution to making the casino attractive to players without giving them straight +ev to all their bets.
The usual deposit bonuses are very problematic because either they are good for the players or they are good for the casino.
What this means is that either the casino is going to be unhappy or the players are unhappy.

What i am proposing is to stop with the usual deposit bonuses that apply to the all players, and concentrate on the losing players.

For example we have 2 players: player A and player B. Both players deposit 100 euro in the casino(without bonus),
and player A wins 150 euro and player B loses whole 100 euro.
Clearly player A is happy with his winnings even if he doesnt receive any special attention.
now for example, If player B got 20 free chip with 10XWR for a second change, he would be happy too.

I am pretty sure that the casino would be expected to make profit,
if players were to get free chip everytime they fall to zero with their deposit(assuming ofcos no withdraws)
The free chip could even be fixed % of the original deposit.

Someone could work out the numbers to make it -ev for the players in general,
but the players would still be happy.

My thoughts exactly. I am very happy to spend my money at a casino with no bonus, if I know they will throw something back (around 10%). I don't even mind a maximum cash out cap on the freebie, but love it when there isn't one. I rarely, if ever, cash out and I know the casinos like it. What they won't like is when I go for the $100k cash out, since that is what I am after.

 

[clees42]

another idea is to stop with the all other promotions and concentrate solely on the cashback.
If you think about it cashback is equivalent to upping the rtp. The following is assuming a cashback that is calculated on each bet separately(like rake in poker).
Lets say you tie up the cashback based on the machine used and the lifetime(or last 30 days) deposit/withdraw ratio of the player.

For example we have player A, player B and player C.

Player A has lost 20k to the casino in his lifetime and he could get 0,003% cashback in blackjack(99,9% rtp) and 2% cashback in slots(99% rtp).
Player B is up 10k in winnings and he could get 0,001 cashback in blackjack(99,88% rtp) and 0,5% cashback in slots (97,5% rtp)
Player C is new player and he could get temporary 0,002 cashback in blackjack(99,89% rtp) and 1% cashback in slots (98% rtp).

In my opinion it would be attractive to play in a casino that offers varied rtp based on my performance.
If i am winning i dont care about lower cashback, because i think the games are hot.
If losing it would be nice to get improved odds.

The point is to keep the house edge in all circumstances while still rewarding the losing players most.
Also if the house edge is alway present there is no need to tie up the cashback to WR.
The cashback could be freely withdrawable or have just a small nominal 1-5xWR.

 

[jstrike]

Holy cr** that's brilliant. Stepped cashback bonuses with no WR, based on player points level. In a weird way I feel like it's so obvious but... I've never heard of anyone doing it like that online before. It's kind of like what Vegas does but they never put it in writing, and they do it in a very casual way. Did you come up with this, or are there other online casinos doing this right now?

It might be the perfect balance. Gives back more to players who are more loyal, but still gives new players a reason to join. Stops the bonus hunters and lets the longer-term players get a better RTP.

This is a really solid idea. I'm going to draw up the numbers on it...

 

[clees42]

Holy cr** that's brilliant. Stepped cashback bonuses with no WR, based on player points level. In a weird way I feel like it's so obvious but... I've never heard of anyone doing it like that online before. It's kind of like what Vegas does but they never put it in writing, and they do it in a very casual way. Did you come up with this, or are there other online casinos doing this right now?

I came up with it after seeing this thread here. I have never seen casino with system like that.
But, Its quite obvious if you think about it, in a way it is very similar to what is used in poker.
In poker player knows that the more he plays the more he gets rewarded by the end of the month as rakeback.
And when he gets the rakeback he is free to do whatever he wants with it.

I think that casino focusing primarily on the cashback, could build their promotions around it too.

For example: limited happy hour with 2x cashback that lasts for 1h after the user initiates it(new users, losing players, holidays, birthdays etc)
or instead of happy hour it could be 2x cashback until total 500 wagered

 

[Jufo]

Hi clees42, a fellow Finn Is it cold in Oulu at the moment?

A couple of remarks about what has been posted:

Those "sharks" jstrike mentioned (bonus hunters or +EV position hunters) sometimes pretend to be regular gamblers to get access to better bonus offers. At Beatingbonuses forums (where many of these sharks hang out) there is a lot of discussion about when coverplay (ie. pretending to play like a degenerate gambler) is profitable. Not to make you paranoid jstrike, but a player which you consider as a recreational player might be a shark in disguise.

10% and 20% cashback bonuses have been known to cause losses to the casino. Some time ago there was a news article about a Live Blackjack player who was comped with 20% cashback every time he lost and eventually he ended up winning millions and cleaning the house. That's because the cumulative house edge of the game was less than the cumulative cashback he received. So cashback bonuses are not always -EV and sharks are after them as well. However if the cashback is discretionary then it probably won't get shark's attention as they have better things to focus on.

The point is to keep the house edge in all circumstances while still rewarding the losing players most.

Personally I disagree with this train of thought. A casino who is so cheap that they are NEVER willing to give up on their mathematical edge, even for a short while, tells me that they are just cheap and probably under-bankrolled. I believe in the phrase: "You have to give action to get action". It means that I expect the casino to voluntarily put themselves in a losing position once in a while. It shows that they are willing to have a gamble against me, a fair fight so to speak where they don't have the advantage, and they earn my respect and loyalty that way.

 

[clees42]

Personally I disagree with this train of thought. A casino who is so cheap that they are NEVER willing to give up on their mathematical edge, even for a short while, tells me that they are just cheap and probably under-bankrolled. I believe in the phrase: "You have to give action to get action". It means that I expect the casino to voluntarily put themselves in a losing position once in a while. It shows that they are willing to have a gamble against me, a fair fight so to speak where they don't have the advantage, and they earn my respect and loyalty that way.

This could be solved by giving player temporary boost to cashback resulting over 100% rtp under some circumstances(reward for loyalty, cheer up player that lost etc)
It could be time limited, or until certain wagering target is reached.
But it would still need to be implemented in a way that the casino will be in control of their losses.

 

[jstrike]

10% and 20% cashback bonuses have been known to cause losses to the casino. Some time ago there was a news article about a Live Blackjack player who was comped with 20% cashback every time he lost and eventually he ended up winning millions and cleaning the house.

This gets really interesting. And it has to do with the churn or playthroughs on that player's chips. If the bet size as a fraction of his bankroll is small enough that he has to play through 10x to double up or zero out, then the house can still have a tiny edge giving him a 10% cashback. On the other hand, if the player can bet his whole bankroll on one hand and then get 10% back on a loss every time, the house is going to be playing against its own money within 20 hands. Which would be an argument for putting WR on the cashback or only giving the cashback if a certain WR had been reached on the deposit before the zero-out.

I know about the player you're talking about, but I don't know when he was cashed back or for what, exactly. It sounds like any time he left the casino down he was getting 20% back. That's different, at least psychologically, from getting a cashback on an online casino only when you zero out your deposit.

It means that I expect the casino to voluntarily put themselves in a losing position once in a while. It shows that they are willing to have a gamble against me, a fair fight so to speak where they don't have the advantage, and they earn my respect and loyalty that way.

I agree with this. But players can get used to certain things and be pushy about them. Many players, for some strange reason, seem to assume that a casino should put itself in that position all the time. It takes a light touch to manage expectations so they don't get unreasonable.

 

[jstrike]

Okay, so here's the idea I've put together, based on this great discussion. Tell me what you think!

What I would do is automate cashbacks, so anyone can claim one after zeroing out, with no WR attached. This would be a cashback based on the amount deposited (in one or more deposits) since the player's last withdrawal or last cashback, and no more than 1 week in the past.

The way it would work is this. I'd take the player's total bets since their last withdrawal, divided by the total deposits and bonuses in the same time period. That would give a certain maximum cashback rate based on the lowest house edge being around 1%. So let's say a player had deposited $100 and placed $1100 in bets before zeroing out. The maximum cashback I could give her and break even theoretically is 11%. This wouldn't be restricted by what games she had played (roulette, craps, blackjack would all count the same).

Now, of that theoretical maximum, a player's actual cashback eligibility is determined by how many player points they have. For example someone with more than 1 million points could get 90% of that 11% (a 9.9% cashback), whereas someone with under 1,000 points would only get 10% of it (an 0.99% cashback). (One point is awarded for every $0.25 bet).

Does this sound too complicated? Or does it make sense? I would set it up as a separate window in the software that would show what the player was eligible for if they just zeroed out, and let them take the cashback immediately. I think I'd have to put some maximum on the playthrough count, like 20%, so players in the highest class with the most points could get up to an 18% cashback if they had played through their last deposit 20x or more.

 

[clees42]

Does this sound too complicated? Or does it make sense? I would set it up as a separate window in the software that would show what the player was eligible for if they just zeroed out, and let them take the cashback immediately. I think I'd have to put some maximum on the playthrough count, like 20%, so players in the highest class with the most points could get up to an 18% cashback if they had played through their last deposit 20x or more.

I am pretty sure that most player are only interested too see what the potential cashback window is displaying at the moment.

 

[clees42]

Does this sound too complicated? Or does it make sense? I would set it up as a separate window in the software that would show what the player was eligible for if they just zeroed out, and let them take the cashback immediately. I think I'd have to put some maximum on the playthrough count, like 20%, so players in the highest class with the most points could get up to an 18% cashback if they had played through their last deposit 20x or more.

If the cashback is extented to previous deposits, then why is cashback % only defined on the playtrough of last deposit?

Lets say i deposit 1000 euro first and lose it. then i deposit 20 euro again and again hoping to reach high enough playtrough to reach best cashback % for all the deposits including the 1000 euro.

 

[jstrike]

If the cashback is extented to previous deposits, then why is cashback % only defined on the playtrough of last deposit?

Lets say i deposit 1000 euro first and lose it. then i deposit 20 euro again and again hoping to reach high enough playtrough to reach best cashback % for all the deposits including the 1000 euro.

Maybe I said it wrong... the playthrough would apply to all the deposits and any bonuses (including other cashbacks) received since the last withdrawal.

So if you deposit $1000, take a $100 cashback, and then deposit another $20, your potential cashback on that $20 if you zero out would be:

(totalbets since last withdrawal / 1120) * (playerpoints level)

if you waited to take the cashback until you had lost both deposits, it would be:

(totalbets since last withdrawal / 1020) * (playerpoints level)

Now... if you had put in that $1k more than a week ago, you would lose the right to a cashback on it, but it would still be factored as part of your playthrough, as would all the bets you made with it, since the playthrough is just concerned with what happened since the last withdrawal. Meaning even if you missed the cashback on that, if you had high playthrough on it, that would still help you qualify for higher cashback on your next deposit.

One reason I'm thinking about setting it up this way is because there are a lot of players in my system who have only ever played freerolls, and never deposited, or maybe only deposited $1 once. And they have an enormously high "playthrough" by this definition. However they don't have a lot of player points, because by and large they've spent them already to enter higher-paying tournaments. So this encourages them to deposit and reap the rewards of a relatively larger -- but not huge, maybe only 5% -- cashback, or else withdraw and start from scratch instead of letting their freeroll wins heap up in my system, which I prefer not to have on my books.

 

[clees42]

Now... if you had put in that $1k more than a week ago, you would lose the right to a cashback on it, but it would still be factored as part of your playthrough, as would all the bets you made with it, since the playthrough is just concerned with what happened since the last withdrawal. Meaning even if you missed the cashback on that, if you had high playthrough on it, that would still help you qualify for higher cashback on your next deposit.

ok that sounds better. But, however i think that starting from scratch after each withdraw is a bit harsh in my opinion.
The size of the withdraw should be factored in also.
Lets say that i have huge playtrough and 2000 euro in deposits, then i withdraw 50 euro and end up with 0 playtrough.
Would sound like too big penalty. Maybe the playtrough should be reduced by this type of formula: (withdraw*X*total playtrough)/total deposits.
For example: assuming X=2 total deposits at 1000 euro and you withdraw 250 you would lose 50% of your playtrough value.

 

[jstrike]

ok that sounds better. But, however i think that starting from scratch after each withdraw is a bit harsh in my opinion.
The size of the withdraw should be factored in also.
Lets say that i have huge playtrough and 2000 euro in deposits, then i withdraw 50 euro and end up with 0 playtrough.
Would sound like too big penalty. Maybe the playtrough should be reduced by this type of formula: (withdraw*X*total playtrough)/total deposits.
For example: assuming X=2 total deposits at 1000 euro and you withdraw 250 you would lose 50% of your playtrough value.

I was just thinking about this, but the problem I see is -- imagine someone deposits $100 and right away withdraws $5. Then they deposit another $100. They now have $195 to play with, but only $100 to "clear". Now, in order to get a cashback they have to zero out, so they do have to lose all $195. But if they play it through 10x before losing it, they will already get a much larger cashback -- because they played $1950 on a $100 deposit, so they would get 19.5% (times their player points modifier).

This means that they are actually getting cashback based on whatever portion of their previous amount they did not withdraw (which is another reason to cap it at 20%). So if a player is up against us by $2000, takes out $1900 and deposits $100, they will almost definitely get a 20% cashback on their next zero-out. But this is OK because in this case they have chosen to keep $100 in their account, which they then lost -- basically the same as putting it into a new deposit. But if I gave them credit for their playthrough before that, I think everybody would always be at the highest cashback level after a little while... right?

 

[binshakindown]

Does this sound too complicated? Or does it make sense?

Yes is does sound too complicated, but it makes sense. Why not incorporate the cash back formula on the player's RTP (not sure if I used correct word here), deposits and withdraws over an interval of time, say 30 days for example. This way, the player who has withdrawn more money than deposited doesn't receive a cashback bonus on subsequent deposits until the player has actually lost some of their own money.

 

[clees42]

I was just thinking about this, but the problem I see is -- imagine someone deposits $100 and right away withdraws $5. Then they deposit another $100. They now have $195 to play with, but only $100 to "clear". Now, in order to get a cashback they have to zero out, so they do have to lose all $195. But if they play it through 10x before losing it, they will already get a much larger cashback -- because they played $1950 on a $100 deposit, so they would get 19.5% (times their player points modifier).

This is easily solved if any remaining balance in the time of new deposit is added to the value.
Lets consider 3 situations.

A
player deposits 200 and loses it with 2000 playtrough so he gets 10% cashback of 200

B
player deposits 100, loses it, then redeposits another 100 and loses it. The playtrough on the last deposit is 1000 so he gets 10% cashback of 200

C
player deposits 100, withdraws 5 and deposits another 100. The playtrough is 1950 so he should be eligible to 10% cashback of 195.


Case C should be equivalent to someone first withdrawing 5 and then depositing 195.

 

[jstrike]

C
player deposits 100, withdraws 5 and deposits another 100. The playtrough is 1950 so he should be eligible to 10% cashback of 195.


Case C should be equivalent to someone first withdrawing 5 and then depositing 195.

But, I think it is the same. It doesn't matter what the playthrough was before the withdrawal, since either way there's now 195 in the account and you'd expect there to be about $1950 in playthrough from that point before the player zeroed out. So instead of the player getting 10% of $195, in Case C she gets 19.5% of $100. Same thing, but I think it's important to keep the rule that only deposits since the last withdrawal are eligible... otherwise I think this could turn into a mess...

 

[clees42]

But, I think it is the same. It doesn't matter what the playthrough was before the withdrawal, since either way there's now 195 in the account and you'd expect there to be about $1950 in playthrough from that point before the player zeroed out. So instead of the player getting 10% of $195, in Case C she gets 19.5% of $100. Same thing, but I think it's important to keep the rule that only deposits since the last withdrawal are eligible... otherwise I think this could turn into a mess...

i got an idea how to improve on the cashback idea, but im bit busy right now so ill have to formulate it bit later.

 

[clees42]

EDIT: i think it should be fixed now.

Maybe the cashback should be calculated with this formula.

(((depositA-withdrawA)*playtroughA/depositA)+((depositB-withdrawB)*playtroughB/depositB)+...+((depositX-withdrawX)*playtroughX/depositX))/total deposits

In this formulate any cashback should be counted as a deposit. cashback on cashback should be added to the next deposit.
WithdrawA is a withdraw made from balance generated with depositA. if withdrawA is larger than deposit A then depositA-withdrawA=0
however if you feel like winning player gets too high cashback with this depositA-withdrawA can be calculated as negative to lower the cashback.
The idea is to calculate the cashback for the next deposit by weighting the previous cashbacks relative to deposit sizes.

example1:
Player has following deposits and withdraws.

1. deposit 1000 euro, playtrough 10000 and withdraw 2000
2. deposit 500, playtrough 5000 and withdraw 0
3. deposit 200, playtrough 2000 and withdraw 100 and lose rest.

Now we calculate how much cashback he should get when he deposits 100 and loses it with 1000 playtrough

((0+500*5000/500+(200-100)*2000/200+100*1000/100)/1700=(5000+1000+1000)/1700=4,1, -> 4,1 %cashback 4,1 euro OR
((1000-2000)*10000/1000+(200-100)*2000/200+100*1000/100)/1700=(-10000+1000+1000)/1700= -8000/1700 so no cashback.

example2.
player has following deposits and withdraws.

1, deposit 100 euro, playtrough 1000, withdraw 0
2. deposit 200 euro, playtrough 2000, withdraw 0

now cashback when deposit 300 and lose with 3000 playtrough.

(100*1000/100+200*2000/200+300*3000/300)/600=(1000+2000+3000)/600=10 -> 10% so 30 euro cashback.

example3 with cashback inbetween.

1.deposit 1000 euro, playtrough 10000, withdraw 0. cashback ((1000*1000)/10000)/1000=0,1=10% so cashback 100 euro
2. cashback 100 euro, playtrough 1000, withdraw 300, no cashback.
3. deposit 200, playtrough 2000, withdraw 0,

cashback: ((1000*10000)/1000+0+(200*2000)/200)/1300=(10000+2000)/1300=9,2 -> 9,2% cashback=18,4 euro OR
((1000*10000)/1000+(100-300)*1000/100+(200*2000)/200)/1300=(10000-2000+2000)/1300=7,7 -> 7,7% cashback=15,4 euro

 

[jstrike]

@clees42, thank you for taking the time to calculate all of that. I'm trying to wrap my head around it. I basically understand it, and in fact it seems like it's actually lowering cashbacks for players who recently withdrew winnings. Which is interesting. But I think this might be too complicated to communicate clearly, and kind of blur the goals for people who are trying to hit a certain playthrough mark.

Additionally, because the site's in Bitcoin, deposits and withdrawals tend to be a lot more frequent than on other sites. Players cash out as a way of managing their bankroll. Often times it will take 1-2 hours for the withdrawal to be approved, and in that time they'll deposit again and lose that deposit, then try to claim a cashback on the loss while the other withdrawal's still pending. What this means is that my code would have to spider through the player's entire transaction history and try to figure out which withdrawals matched which deposits. And most of the time, it's not a 1 to 1 relationship.

I think your system would also prevent people who already have a history of being net winners in the casino from ever claiming a cashback, if I understand it correctly...? And I'm not sure I want to do that.

In your first example, it's true, in the system where withdrawing clears the history, the player would forfeit the cashback just by withdrawing anything. But this is OK because I really only want the cashback to be if the player zeroes out the initial deposit without a withdrawal.

In exchange for that, the cashbacks are a little higher in the simplified system. To take your third example... with the simplified system it would look like this:

1.deposit 1000 euro, playtrough 10000, withdraw 0. cashback 10% = €100
2. cashback 100 euro, playtrough 1000, withdraw 300, no cashback.
3. deposit 200, playtrough 2000, withdraw 0. cashback 10% = €20

But more so, I think it's just easier to understand and show the player a target (I was thinking like a little meter that shows playthrough and cashback % since the last withdrawal) when it's from one withdrawal to the next.

Re-reading your post it's pretty mind-boggling, although I'll admit I'm not the best when it comes to math (or managing money, lol). But I can also see having a very hard time explaining to players in some situations (like a deposit while there was a pending withdrawal) why their cashback is exactly whatever it comes out to with that equation...