Calculating Loss Rebate Advantages

[tomf23]

Simply bet $25K on the first hand. With B.S. splits and DD's, your average wager will be roughly 10% larger, so $27.5K.

If you win, quit.

If you lose, quit and collect, on average, $9625, for a loss of $17875.

If you push, bet $25K once again.

So, ignoring pushes, you'll win roughly 48% of you "sessions", with an average win of about $28K (more than $27.5 due to BJ's). You'll lose the other 52% of your "sessions" with an average loss after rebate of $17875.

This means that you'll win an average of $4145 per session.

Hope this helps!

Dog Hand

isn't this method just begging for them to cut the rebate promotion?

[Norm]

Eliot Jacobson sent me the following link:

http://www.ijmp.org/AAP/Loss_Rebate_Theorem.xlsx

He also adds this:

Here are approximate numbers for single deck blackjack with a 35% discount.

The person should quit after winning 66 units or losing 76 units. Scale this so that 76 units = $25000.

[Dog Hand]

isn't this method just begging for them to cut the rebate promotion?

tomf23,

Undoubtedly! However, the Ice_man simply asked if the rebate plus B.S. gives +EV: I was only confirming that the situation is indeed +EV.

Dog Hand

[randooo_1234]

Norm might have already given you the right Numbers, but if you need more details like RiskofRuin, IBA,... feel free to contact me, i just wrote a sim for other Rebate Games Last week!

[James989]

Eliot Jacobson sent me the following link:

http://www.ijmp.org/AAP/Loss_Rebate_Theorem.xlsx

He also adds this:

Here are approximate numbers for single deck blackjack with a 35% discount.

The person should quit after winning 66 units or losing 76 units. Scale this so that 76 units = $25000.

Based on BJ HE = -0.5% and Var = 1.30, my both simulation results and iteration results matched, as below :-

1) loss exit = 29.6 units
2) win exit = 25.5 units
3) average win = 2.442 units
4) expected rounds = 604
5) loss likelihood(%) = 51.45


Meaning you will win 2.44 units after flat bet 604 rounds, average ev(%) = 2.44/604 = +0.404%

Please correct me if anyone have different answers.

Thanks

[James989]

tomf23,

Undoubtedly! However, the Ice_man simply asked if the rebate plus B.S. gives +EV: I was only confirming that the situation is indeed +EV.

Dog Hand

Of course, if without any constraint, just bet ONE hand and quit ! LOL

Dog Hand, could you please check my results ? I just want to know what is my mistake, if any. Thanks

[Eliot]

Based on BJ HE = -0.5% and Var = 1.30, my both simulation results and iteration results matched, as below :-

1) loss exit = 29.6 units
2) win exit = 25.5 units
3) average win = 2.442 units
4) expected rounds = 604
5) loss likelihood(%) = 51.45


Meaning you will win 2.44 units after flat bet 604 rounds, average ev(%) = 2.44/604 = +0.404%

Please correct me if anyone have different answers.

Thanks

My spreadsheet gives these results (with h/a = -0.5% and var = 1.3):

Win-quit point	25.9967
Loss-quit point	30.0051
Expected net win	2.4344
Probability of win	48.215%
Effective edge	0.4052%
Expected rounds	600.8

The closed solution assumes differentiable IID random variables. A simulation will always give more accurate results for a discrete game, but of course takes a lot more time and effort.

[James989]

My spreadsheet gives these results (with h/a = -0.5% and var = 1.3):

Win-quit point	25.9967
Loss-quit point	30.0051
Expected net win	2.4344
Probability of win	48.215%
Effective edge	0.4052%
Expected rounds	600.8

The closed solution assumes differentiable IID random variables. A simulation will always give more accurate results for a discrete game, but of course takes a lot more time and effort.

Thanks for your verification.

[James989]

My spreadsheet gives these results (with h/a = -0.5% and var = 1.3):

Win-quit point	25.9967
Loss-quit point	30.0051
Expected net win	2.4344
Probability of win	48.215%
Effective edge	0.4052%
Expected rounds	600.8

The closed solution assumes differentiable IID random variables. A simulation will always give more accurate results for a discrete game, but of course takes a lot more time and effort.

Possible to find win and loss exit for game with player edge off the top ?

[Eliot]

With a player edge off the top, intuitively there should be no win-exit point. I think my sheet gives nonsense in this case.

[randooo_1234]

Possible to find win and loss exit for game with player edge off the top ?

If you give me the exact rules I can do a simulation right now for you!

Of course the longer you play the more you will win, but if you want to have a target EV of at least x%, I can give you the stop points for it!

[James989]

If you give me the exact rules I can do a simulation right now for you!

Of course the longer you play the more you will win, but if you want to have a target EV of at least x%, I can give you the stop points for it!

If I set a target x%, I can find win exit point with simulation as well. Thanks anyway.

[randooo_1234]

But what's your question then?