Do House Rebates Create a Player Edge?

Ok, now hear me out on this. I have been reading on this site that house rebates are usually just ploys to keep you gambling, but I am curious as to how the math works out on this.

In my example, the house offers 10% back on losses over X dollars over a set amount of time. In my case, for a three day period, if I lose X dollars or more, I get 10% back. For simplicity, let's say you get 10% back on losses over $1,000.

The reason I am even thinking it might give an edge is because of the style of play I use. I play the pass line with full odds (3-4-5 casinos), with come bets, and full come odds when come is established. Also, when come is established, I put another come bet out there, trying to catch the next one if the point doesn't hit. This strategy tends to give less than one half of one percent house edge, according to the Wizard. I tend to press after a few collections, but for this argument's sake I will keep it to the basic strategy I just mentioned.

Now, my question is, can I integrate the 10% back into one of the odds calculations to get an EV? If so, where/when?

No!

Since the rebate only kicks in if you've lost money, it doesn't affect the expected value. It reduces the variance somewhat.

Comps (not rebates) can be seen as changing the house edge to a negative number. If you are playing 6/9 video poker with a very small HA, the free drinks and food can be seen as giving you a positive expectation. But casinos are usually pretty careful about comping you on machines or games with no little or no HA. For instance if you get "comp dollars" in most casinos they only work in the slot machines with a larger HA.

Quote: pacomartin

Since the rebate only kicks in if you've lost money, it doesn't affect the expected value. It reduces the variance somewhat.

Changing variance makes sense, but how would you articulate that?

This is where I am trying to wrap my head around this: if they gave you 100% back, then you can't lose and might as well keep playing. Sure, 10% may not tip the scales in your favor, but there has to be some point in between 10 and 100 where the casino won't cross. Is there a way to calculate that? Thanks!

if you can bet a large amount of money on one hand its easy to figure out the breakeven point.

assuming a 2% HE for easy calculation.

(.49)(1000) + .49X - (.51)(1000) - (.51)(X) + (.51)(.10)(X) = 0

490 + .49X - 510 - .51X +.051X = 0

-20 +.031X = 0

.031X = 20

X = $645

so if you bet > $1645 on one hand assuming a 2% HE, its +EV.

It is not 10% on each bet that goes over $1000. It is 10% back at the end of the night. In theory that calculation would be correct if you could make one bet, settle, and then return and make another bet.

Quote: phonebooth

Ok, now hear me out on this. I have been reading on this site that house rebates are usually just ploys to keep you gambling, but I am curious as to how the math works out on this.

Let us try to be consistent in our terminology.

Comps are different than what is termed a rebate. A rebate on your losses is, I'm told, usually limited to those who are at the two grand a hand and up levels. That is territory I know nothing about.

Comps are usually based on your Action, irrespective of actual win/loss results though actual results may be the basis for a very slight adjustment to your Comps.
Rebates are only based on your Losses. As far as I know, such rebates are High Roller perks.

I look at rebates the same way I look at making a tax-deductible purchase.


It amazes me that people will go ahead and purchase something, and select the more expensive choice with a comment like "It's deductible" as if that makes the item free, or at least irresistably cheap.

Me? I get my value and probably go for the cheaper item. The fact that it's deductible only make the purchase sting a little less.


Unless you know that you play so badly that you're going to hit that rebate minimum, it shouldn't matter.

Here's the thing: If you're playing with a rebate program as described, such as 10% of losses if you lose $1,000 or more, then the only time you should even be thinking about it is if you've lost between $900 and $1,000.

In that range, then, and only then, are you playing with "free money". If you should happen to win, GREAT! If you lose, stop once you've lost $1,000.


Continuing to play, with the knowledge that it's only costing you 90% of what you put in, is like making the big purchase, but saying it's deductible.

I have had some difficulty understanding the concept of this question from the very beginning. If there were some scenario by which you had a player advantage or negative HA, that would seem to imply that you expected to come out ahead. If a rebate implies that you lost money, I don't see how any rebate not greater than 100% could ever lead to a player advantage. I don't even comprehend why anyone would ever think it would. (Don't mean that as a slight to anyone else, it's just that I don't follow that thinking.) I think DJTeddyBear explained it best -- for a 10% rebate with a $1000 threshold, you have to already be down more than $900 for the rebate to have any chance of putting you better off than you already are, and you will still be negative at least $900. Not what I would call a case of player advantage.

Um... error! You CAN gain a player advantage when considering the expectation of a session/trip against the amount bet over that trip. Concept explained in askthewizard column:

On a recent Travel Channel show about Las Vegas whales, they surprised me with the statement that casinos sometimes offer loss discounts to whales. In other words, the whale plays on credit, and are charged only a percentage of their total losses at the end of a visit. Does this make it possible to set up a positive-expectation game? Would betting systems start to make sense in this context?

Yes, this is true. It is not unusual for "whales" to get a 10% rebate on losses. In my opinion, this is a very risky offer to make and a sharp player could easily abuse it and gain an advantage. The kind of player who would make an ideal recipient of this offer is one who grinds out a lot of play in a high house edge game. The kind of player who could best exploit this offer is one who plays a low house edge game, for a short time, and with a wide range of bet size. It sounds paradoxical but under this deal the player must lose to gain any benefit. Thus, the player should set a high winning goal and relatively low losing goal. Ignoring the house edge for the sake of example, if the winning marker was $1,000,000 and the losing marker was $100,000 then the probability of success is 1/11, as I show in a later question. The expected value after the 10% rebate is (1/11)*$1,000,000 + (10/11)*(0.9*-100,000) = +$9,091. A good strategy to achieve a high winning goal quickly would be something like an anti-martingale, or anything where you bet more after you win. Nov. 28, 2004

Quote: Doc

Not what I would call a case of player advantage.

The only player advantage would be to have the rebate be at a rate of 110 percent of losses and no casino is stupid enough to do that or anywhere near that. I think this would have to be a program wherein the gambler places great psychological value on a comp such as a free drink or a rebate. Sort of "I lost, but the casino loves and respects me enough to rebate ten percent of my loss". This would be utter foolishness on the gamblers part but its probably something that happens from time to time.

Quote: pacomartin

It is not 10% on each bet that goes over $1000. It is 10% back at the end of the night. In theory that calculation would be correct if you could make one bet, settle, and then return and make another bet.

hence why i said if you could make a large wager on one hand.

Am I missing something here? This can create a pretty clearly illustrated PA, can't it?

Assume I bet $1000 on black one time each day on a single 0 roulette wheel (2.7% HA)

Without the rebate my EV is 1000*.973=$973


To see the effect of the rebate, assume I play 1000 days in a row with perfectly predictable results. black hits 486 times, red hits 486 times, and green hits 28 times.

I win 486*$1000=486000
I lose 514*($1000-($1000*10% rebate))=$462600
My net is $23400, which represents a PA of 2.34%

My EV with the rebate under this scenario is $1000*1.0234=$1023.40

if i interpreted it correctly.

OP gets a rebate on losses over $1000.

so if you lose $2000, you get a rebate of 10% on $1000 not $2000.

so a $2000 loss has a rebate of $100. a $3000 loss has a rebate of $200. a $4000 loss has a rebate of $300 and so forth.

Quote: rudeboyoi

OP gets a rebate on losses over $1000.

so if you lose $2000, you get a rebate of 10% on $1000 not $2000.

so a $2000 loss has a rebate of $100. a $3000 loss has a rebate of $200. a $4000 loss has a rebate of $300 and so forth.

You're right.... for the wording in the original post.

Except other posters, as well as the AskTheWiz that dwheatley copied, has it worded that if losses exceed $1,000 then the rebate is 10% of ALL losses.

Quote: dwheatley

Um... error! You CAN gain a player advantage when considering the expectation of a session/trip against the amount bet over that trip. Concept explained in askthewizard column

dwheatley: Thanks for your comments -- perhaps I am beginning either to see the light or to be blinded by all the flash. I think it relates to the Wizard’s comment, “the player must lose to gain any benefit.”

I suppose this is just a way of exploiting variance, accepting it if it is good and having it partially canceled if it is bad, perhaps gaining on average. My thinking on this had been that if your performance for the session has been right along the expectation for the game (very low variance), then you are still going to lose even with the rebate.

In the Wizard’s example, I think he ignores house advantage, assumes the player keeps playing until reaching one extreme or the other (with a 90.9% probability of losing the max), and suggests the player should only require a short time to conclude his play. I rarely if ever play with that kind of approach, which probably explains why I didn’t anticipate that line of thinking.

Quote: DJTeddyBear

You're right.... for the wording in the original post.

Except other posters, as well as the AskTheWiz that dwheatley copied, has it worded that if losses exceed $1,000 then the rebate is 10% of ALL losses.

yah im reading it again and its a little difficult to interpret. now i can see it going either way.

"In my example, the house offers 10% back on losses over X dollars over a set amount of time. In my case, for a three day period, if I lose X dollars or more, I get 10% back. For simplicity, let's say you get 10% back on losses over $1,000."

if we are unsure, we might as well choose the conservative option of 10% of only the losses over $1000 because it would still be applicable to ALL losses and any strategies set in place would still work for both situations just a little more effectively if its 10% of ALL losses.

I should probably clarify, my original post might have been poorly worded...

It's 10% on ALL losses, so long as you PASS $1000. Meaning $100 back on $1000, $200 on $2000, etc. But no money back on $999. It's not a comp, it's money back once you close out for the three day period.

Now, taking this into account, if you find yourself in a situation where you have lost $1100 (let's say), would it be fair to assume that at that point, you are playing in a range where losses are 90% and not 100%? And if so, how would you formulate this? I'm actually interested in how the math is calculated for this situation and how it may apply to an overall strategy.

For example, if I win money on a trip, I will probably wait three days to play there again, because if I lose it, it doesn't count towards the rebate. However, if I have a bad run and cross the threshold, could 90% expected loss (again this is where the question comes in on whether this is an appropriate distinction) change my desire to leave? If not 10%, then how much?

Thanks for all your help!

Quote: phonebooth

It's 10% on ALL losses, so long as you PASS $1000. Meaning $100 back on $1000, $200 on $2000, etc. But no money back on $999. It's not a comp, it's money back once you close out for the three day period.

Now, taking this into account, if you find yourself in a situation where you have lost $1100 (let's say), would it be fair to assume that at that point, you are playing in a range where losses are 90% and not 100%? And if so, how would you formulate this? I'm actually interested in how the math is calculated for this situation and how it may apply to an overall strategy.

Once you have crossed the threshold, you are in an interesting situation where you really do have positive expectation on every bet you make, unless that bet (or a future bet) would take you above -$1000.

Simple example: You are at -$1500, you place a $100 bet on a game with 1% HA. With 49.5% chance you win $100, while with 50.5% chance you lose $90 (so long as you don't finish above -$1000 this session!). So each bet has a +EV of $4.05. That's pretty good... However, for this simple math to hold, you MUST end up below the -$1000 threshold! The simple strategy would be to play aggressively like the wizard suggests early in your trip. If you ever drop below -$1000, then switch to a low grind until you are at -$1001 and cash out for the session.

There may still be an advantage to betting once you cross above the threshold, but it is not as simple to calculate as above.

Quote: dwheatley

Um... error! You CAN gain a player advantage when considering the expectation of a session/trip against the amount bet over that trip. Concept explained in askthewizard column:

On a recent Travel Channel show about Las Vegas whales, they surprised me with the statement that casinos sometimes offer loss discounts to whales. In other words, the whale plays on credit, and are charged only a percentage of their total losses at the end of a visit. Does this make it possible to set up a positive-expectation game? Would betting systems start to make sense in this context?

Yes, this is true. It is not unusual for "whales" to get a 10% rebate on losses. In my opinion, this is a very risky offer to make and a sharp player could easily abuse it and gain an advantage. The kind of player who would make an ideal recipient of this offer is one who grinds out a lot of play in a high house edge game. The kind of player who could best exploit this offer is one who plays a low house edge game, for a short time, and with a wide range of bet size. It sounds paradoxical but under this deal the player must lose to gain any benefit. Thus, the player should set a high winning goal and relatively low losing goal. Ignoring the house edge for the sake of example, if the winning marker was $1,000,000 and the losing marker was $100,000 then the probability of success is 1/11, as I show in a later question. The expected value after the 10% rebate is (1/11)*$1,000,000 + (10/11)*(0.9*-100,000) = +$9,091. A good strategy to achieve a high winning goal quickly would be something like an anti-martingale, or anything where you bet more after you win. Nov. 28, 2004

Thanks dwheatley, this is pretty much what I am looking for, however, in the Wizard's answer above, he notes that the probability of success is 1/11, which he says he will "show in a later question". Do you know where he follows up on this? Because I can't find it, and I would want to take that into account when making a strategy. I saw your other post as well, where you seemingly understood my dilemma and you said that it was not as simple as the calculation you had made. I agree, I made the same calculation, but I also understand that this positive EV is over an infinite number of trials and not always best for short term gain, especially since I obviously don't have an infinite bankroll. Any idea though on where the Wizard follows up on this?

Thanks!

At a casino in Colorado, you get $10 instantly when you buy in for $100 in BJ. Min bet $5. You must play two hands. They've been doing this every day this year... You can do it every 24 hours.

Quote: playpianoking

At a casino in Colorado, you get $10 instantly when you buy in for $100 in BJ. Min bet $5. You must play two hands. They've been doing this every day this year... You can do it every 24 hours.

Well if you live in that town and you go over and play 20 hands or less every day you will have a positive EV. But if you have to drive 10 miles each way to get there it will eat up your EV in gas money.

The advantage of the Terrible's casino 10% rebate for $1000 loss, is that Terribles casino is within a 10 minute drive of several hundred thousand people.

Quote: phonebooth

Thanks dwheatley, this is pretty much what I am looking for, however, in the Wizard's answer above, he notes that the probability of success is 1/11, which he says he will "show in a later question". Do you know where he follows up on this? Because I can't find it, and I would want to take that into account when making a strategy. I saw your other post as well, where you seemingly understood my dilemma and you said that it was not as simple as the calculation you had made. I agree, I made the same calculation, but I also understand that this positive EV is over an infinite number of trials and not always best for short term gain, especially since I obviously don't have an infinite bankroll. Any idea though on where the Wizard follows up on this?

Thanks!

The 1/11 figure is from the standard Risk of Ruin formula (with no house edge). The follow-up is the 3rd question in this column:

https://wizardofodds.com/ask-the-wizard/123

I am a bit confused. It seems to me that the current Terrible's promotion is beatable by players. Yet, I've seen very little about the advantage play of this offer on this forum, or any forums. Is it, long term, a profitable play?

I will basically summarize the promo T&C, and rules as follows:

10% rebate of NET loss on table games. May be claimed once every 24 hours. Max bet is $500.00. Players may play two hands of Blackjack. Player may take or lay full odds (3,4,5) on craps. Players are not hassled if making one or two plays and claiming rebate if qualified to return next 24 hour period. Qualifying for rebate is a net loss of $500.00 - $999.00 gets 5%, loss of $1,000.00 or more gets 10%. Players are tracked by pit for losses. For the sake of simplicity and to gain the maximum from this promotion, if a loss is incurred it will always be for a minimum of $1,000.00.

I have spoken to the "pit critters" and a shift manager at Terrible's, to answer some questions, verify terms and qualifiers, and was also told that this has been a promotion for many years, and is not going anywhere.

My instincts tell me this is a promotion where the player would have a significant advantage of much more than 1%, perhaps more than 3%+ depending upon how the player attempts to exploit the promotion. All table games are eligible.

Whether at craps, or blackjack, I just do not understand why there's been very little conversations about this. Is that because it is too good, players doing well and keeping it quiet and to themselves, or that long term it is a loser? It seems to me that it should be a profitable offer for players at either craps or Blackjack, providing a sufficient bankroll, and willingness to adhere to a strict strategy and goal for every session, whether that session is one hand, or one hour before a goal is achieved.

My approach would be betting the maximum, $500.00 (1 unit), at the "best" Blackjack game (less than 0.5%).

My strategy would be to quit a session when a win goal of $1,000 or more, or a stop loss of $1,000 or more is reached. (2 units+)

At Blackjack, it would be possible to win or lose 8 units with 4 splits and 4 double downs on the original hand. Perfect basic strategy would be used to play. A "session" is the amount of time it takes to reach a "goal, than waiting 24 hours, 1 minute or more before a new session is begun.

I wonder if, in the long term, this promotion is not going to provide the player an advantage after a certain number of "sessions".

And, is perfect basic strategy the correct strategy, or should it be adjusted for "special circumstance" such as: 8,8 vs dealer 9,10,A , or A,A vs dealer 10,A? Any other possible scenarios that would affect basic strategy? What adjustments if any, and how much will those adjustments affect the EV, since "always split Aces & Eights" is calculated into the basic EV to begin with.

Rules:Hit s17, DAT,Split 4x, DAS, Aces split once, no hits on Aces after split, no surrender, BJ pays 3-2.

So, is the player advantage long term or short term? Is this as good as it would seem? Am I not getting "answers" because it's so good that those playing want to keep it all to themselves for fear it will get burned out?

When I look at it using the 50-50 coin flip approach, it seems very lucrative. I'd really like to "know" exact expectation.

I'm thinking this is something that should be played often, and as long term as possible, if it is as good as it appears, and a player can has the bankroll and can withstand the potential deviation.

There is another thread dealing specifically with the blackjack math. But my other questions have not been raised.

Thanks

Quote: RoadTrip

I am a bit confused. It seems to me that the current Terrible's promotion is beatable by players. Yet, I've seen very little about the advantage play of this offer on this forum, or any forums.

Weren't there 28 posts about this on the thread you started a few days ago here?