Rebating losses = Player advantage?

Just mulling the story about the "the blackjack player who beat the Trop for $6 million, Borgata for $5 million and Caesars for $4 million" over , it has occurred to me that getting rebated for your gambling losses must equate to Player Edge if the HE is low, AND they go about it in a liberal way.

Let's say a casino agreed to this, and you played a game where you might go ten games winning $1000 each time, and this would typically be matched by 11 losing sessions of $1000 each. $10k up and $11k down, that can't be too unusual. If you got rebated for 20% of your losses *each time* immediately upon losing, for each session, you would indeed have the Golden Goose on your hands. Your losses would be cut to $8800 vs your $10k wins every 21 sessions. I'm sure somebody can make a better math example than that, but...

So I guess they don't do it that way? But the story seems to suggest they do.

edit: PS, it occurs to me the casino would only want to rebate theoretical expected losses unless they were crazy.

Yes, if you are smart, disciplined player you can exploit loss rebates very efficiently. That's probably why Mr. Tropicana has found himself banned/not welcome at many properties. He's a good player.

By the way, Terrible's in Las Vegas will give you a 5-10% loss rebate on $500-$1000 losses.

Quote: odiousgambit

Just mulling the story about the "the blackjack player who beat the Trop for $6 million, Borgata for $5 million and Caesars for $4 million" over , it has occurred to me that getting rebated for your gambling losses must equate to Player Edge if the HE is low, AND they go about it in a liberal way.

Let's say a casino agreed to this, and you played a game where you might go ten games winning $1000 each time, and this would typically be matched by 11 losing sessions of $1000 each. $10k up and $11k down, that can't be too unusual. If you got rebated for 20% of your losses *each time* immediately upon losing, for each session, you would indeed have the Golden Goose on your hands. Your losses would be cut to $8800 vs your $10k wins every 21 sessions. I'm sure somebody can make a better math example than that, but...

So I guess they don't do it that way? But the story seems to suggest they do.

edit: PS, it occurs to me the casino would only want to rebate theoretical expected losses unless they were crazy.

I'm pretty sure at that level, it is done on a "per trip" basis, and on actual losses, not theoretical. However, we do have some former hosts who participate on this board who should be able to confirm the details.

.....

too bad some hosts didnt pipe in and give us the lowdown on this.

I think it is too easy to think that actual losses are better than theoretical losses, but that can't be right can it?

It's fascinating to think that these AC casinos might have really gotten stupid about this! IMHO you would have to rebate only theoretical losses based on the Expected Value of your bets, otherwise the casino could find itself rebating expected negative variance on, well, a game like blackjack where the HE is too low for that [so I disagree with some other posts]. See the first post to see what I mean, and feel free to point out where I am all wet.

Quote: odiousgambit

I think it is too easy to think that actual losses are better than theoretical losses, but that can't be right can it?

Rebating actual losses is better to the casino, because they still get to keep a portion of your money (they won't rebate your entire loss, so, you will still lose, just less than you otherwise would), while if they rebated theoretical losses, they would have to pay you more when you win.
In the long run, it would all even out, of course, because the total actual losses would converge to theoretical, but short term, the casino could easily find itself paying out way more than getting in if it rebated theoretical losses.


Pretty much any time I heard about loss rebate, it was about total losses per day or something like that, not immediate rebates after a loss. If they wanted to do the latter, I think, a better way would be to simply lower the house edge (like that promotion in Connecticut when they'd let you triple down in black jack) ... this would btw, be equivalent to rebating theoretical losses.

Quote: weaselman

In the long run, it would all even out, of course, because the total actual losses would converge to theoretical, but short term, the casino could easily find itself paying out way more than getting in if it rebated theoretical losses.

No no no.

Actual loss rebates don't even out to theoretical loss rebates - because you cannot "rebate" winnings.
Losing $800 instead of $1000 is not equalized by winning $800 instead of $1000.

Aren't normal comps effectively a small rebate of theoretical loss?

Quote: thecesspit

Aren't normal comps effectively a small rebate of theoretical loss?

Exactly. This is why expecting the casino to give you better comping when you lose big is so disappointing. Whether you win or lose, it is the theoretical that matters. In the case of carefully only rebating losses, I can see that being done in a wise way that makes sense... or not. Surely the EV of your betting, always theoretical, has to play a role.

.....

Size of bet and length of session are pretty standard in determining table games comps based on theoretical loss at most if not all casinos.

Where things start to vary is the other factor in the comp formula, house edge. It goes without saying that you will receive less comps playing the pass line at most casinos than if you were to play numbers all day long in roulette. That said, the house edge applied to the comp formula may vary greatly from one casino to the next. One casino may rate all poker variations with the same house edge, while others may rate each variation closer to theoretical edge based on perfect play within each specific game. Unless you have people working in each casino, you will have a tough time figuring out how each casino determines table games comps.

Most casinos will only comp based on actual loss if you are a known player that they have a good feeling of how much money you will risk in the future. Joe Schmo will have a tough time strolling into the MGM, getting a player's card, losing $5,000 in an hour and getting anywhere close to 30-40% of his actual loss back. He will may get a room and a meal out of the deal if he's lucky, but that's probably it.

You may be interested in these links over at WOO:

Ask the Wizard about Rebates
Academic Paper

.....

Quote: dwheatley

You may be interested in these links over at WOO:

Ask the Wizard about Rebates
Academic Paper

fixed links, thanks

PS: the WoO makes this statement I see, from the first link:

Quote:

In my opinion this is a very risky offer to make and a sharp player could easily abuse it and gain an advantage.

I'm going to make an "ask the wizard" question about of this. For now, I think he was milking the 20% rebate on losses the article mentioned, as odios speculated. A 20% rebate in a low house edge game like blackjack is very powerful, and not difficult to exploit.

Quote: Wizard

I'm going to make an "ask the wizard" question about of this. For now, I think he was milking the 20% rebate on losses the article mentioned, as odios speculated. A 20% rebate in a low house edge game like blackjack is very powerful, and not difficult to exploit.

When I read that article, frankly, I just didn't believe him. I think he was just lying. If the casinos were permitting $100K hands, and rebating losses up to 20% I would say they deserved to lose millions.

Under the assumption that Tropicana begin permitting the high stakes blackjack in November of 2010, it looks like the casino is still way ahead for the year. It is also possible that the same player lost several million at the Tropicana in previous months before his 12 hour $6 million winning spree.

Month	Blackjack	6 month average
Jan-10	$2,255,091
Feb-10	$1,422,192
Mar-10	$1,442,870
Apr-10	$2,077,529
May-10	$2,038,401
Jun-10	$1,827,979	$1,844,010
Jul-10	$2,567,605	$1,896,096
Aug-10	$3,055,275	$2,168,277
Sep-10	$2,344,475	$2,318,544
Oct-10	$1,333,374	$2,194,518
Nov-10	$6,429,130	$2,926,306
Dec-10	$4,044,500	$3,295,727
Jan-11	$3,848,590	$3,509,224
Feb-11	$2,045,580	$3,340,942
Mar-11	$2,691,070	$3,398,707
Apr-11	-$1,862,827	$2,866,007

If we assume he had a $6 million bankroll, and he was betting $50K per hand. He quits until he either loses $1 million (and takes the 20% rebate) or until he wins $6 million (for that session, not for overall). If he loses he returns another day. The probability of him hitting the $6 million before going bankrupt are fairly high. Remember we don't know how he did overall, just that he won $6 million that one day. The casino had some big wins, especially November to January. Some of that might have been our guy.

Here is a 2010 article on, "Superhost" Steve Cyr, and the changing environment to get high rollers to his casinos. Note that he mentions giving chips "on the house" to players just for showing up, paying what amounts to an "appearance fee".

There is a mention of rebates, and it is clear that this is based on actual, not theoretical losses.

Quote: Wizard

I'm going to make an "ask the wizard" question about of this. For now, I think he was milking the 20% rebate on losses the article mentioned, as odios speculated. A 20% rebate in a low house edge game like blackjack is very powerful, and not difficult to exploit.

excellent!

Quote: Ayecarumba

Here is a 2010 article on, "Superhost" Steve Cyr, and the changing environment to get high rollers to his casinos. Note that he mentions giving chips "on the house" to players just for showing up, paying what amounts to an "appearance fee". There is a mention of rebates, and it is clear that this is based on actual, not theoretical losses.

In a sense, that is like the cashback or match play or direct bets that the casinos send us ordinary schmendricks. It's just that the scale is different. But then, so are the luxury suites and the fantastic gourmet dinners and wines.