No doubt you've been paying attention to the oddity that is the Revel casino in Atlantic City. In its brief casino-life, they have repeatedly re-branded themselves, making 180-degree turns in their marketing strategy several times. The most recent turn was an admission that they are fundamentally a casino, not a resort. This turn included a name change (from "Revel" to "Revel Hotel-Casino") that was designed to show a serious attitude shift. Revel's most recent marketing theme, designed to show off their new attitude, is "Gambler's Wanted"

Where this story takes a weird turn is with a current marketing promotion that is beyond baffling as far as advantage play is concerned. Here is the press release that Revel put out, in part:

Revel Announces "Gamblers Wanted" Campaign: Casino To Refund All Slot Losses In July And Match All Competitor Slot Offers.

ATLANTIC CITY, N.J., June 20, 2013 /PRNewswire/ -- Today, Revel Casino-Hotel, Atlantic City's newest and largest casino-hotel, announced its new marketing strategy: "Gamblers Wanted."   Along with the introduction of the city's largest contiguous smoking section and new affordable dining options, Revel is announcing the most ambitious promotion in the history of Atlantic City:

You Can't Lose.

From July 1 to 31, all rated slot losses greater than $100 will be refunded to gamblers using a Revel Card.  Gamblers who win, win, but a gambler that loses over the month of July will get that loss refunded. 

In addition, on July 1 Revel also will begin matching slot offers players receive from all other Atlantic City casinos.  Customers need only bring their original mail piece to the Revel Card Desk and an identical amount of Free Slot Play will be loaded to their Revel Card.

This is, essentially, a 100% Loss Rebate program for all players. The fine print shows that there are some restrictions. From what I understand, the two most important restrictions are:

  1. The maximum loss a player can get a 100% rebate on is $100,000. No losses above $100,000 will be covered.
  2. The rebate is given in free play, with a maximum of $5000 in free play available per week, over a period of 20 weeks.

Some players who contacted Revel verified that this promotion covered not only slots, but also video poker (VP) as well.

If this promotion were not so public, APs may have quietly crushed it. Their strategy would first be to find the highest denomination, lowest house edge video poker available at Revel. Next, they would determine optimal quit strategy against this video poker. Then, they would play like crazy, as individuals or teams, either leaving  at a quit-win point, or playing until hitting that magic $100k loss threshold.

All of this may have quietly taken place, but the sheer publicity of this promotion brought it to the attention of gambling message boards and media of all types. In particular, the thread "Revel promotion Sounds too good to be true" at Michael Shackleford's WizardOfVegas.com quickly got to the heart of the matter for APs.

The best video poker was Revel's $25 denomination 9/6 Double Double Bonus Poker (DDB). By wagering a maximum of 5 coins ($125) the player could play a game with a low -1.019217% house edge and a huge standard deviation of 6.479582. A $100,000 loss rebate and a $125 betting unit gives a bankroll of 800 units.

This situation calls for an application of the Loss Rebate Theorem. Here are the results, using the parameters above:

loss rebate calculation

At a rate of 500 spins per hour, the expected playing time is about 47.4 hours of play. The win exit point is a win of 1326 units, with an expected average win of about 348.2 units. Breaking this down on an hourly basis, each $25 DDB machine will lose about $918 per hour to APs playing this strategy. For the month, each DDB machine will lose about $683,000 to APs. This is a stunning result. Revel could lose millions of dollars on its video poker alone.

I checked these results via simulation and found that these results are a fair estimate of optimal quit-win strategy. Here are the results of my simulation (one million simulated players per line):

revel $100K loss rebate promo 9/6 DDB


Note that the simulation gives an optimal expected win of about 356 units, with a quit win of about 1280 units. By comparison, the loss rebate theorem gives an expected win of 348 units. The loss rebate theorem works best with distributions that are "normal." With its large top jackpot, VP is far from normal.

For no other reason than that I did these simulations (in other words, it's purely academic), here are the results from simulating 9/6 Jacks or Better (JOB) video poker, with a $100,000 loss rebate, playing 5 x $25 ($125) per spin:

revel $100K loss rebate promo, 9/6 JOB


From this data, we see that the win goal is 1260 units, with an expected win of 365 units and an estimated playing time of 52008 spins. The win goal and expected win are commensurate with DDB. However, reaching this win goal requires more than double the time playing as DDB. At 500 spins per hour, the AP will play more than 104 hours, on average, to reach one of the stopping points. JOB has lower variance and a lower house edge than DDB. The lesson here is that variance plays a very big role in the viability of exploiting a loss rebate opportunity. Back to the matter at hand.

Days before this promotion began, Revel allegedly shut down all of its $25 video poker. As I write, they are allegedly backing off some players. They are allegedly refusing to rebate the losses of some players. Because of these last-minute actions and back-offs, there are now lawsuits threatened against Revel and general ill-will may have been experienced by some of the gamblers that Revel claims it now wants.

If you are interested in one of the most intriguing threads ever on a loss rebate advantage play opportunity, its evolution and fallout, I encourage you to invest yourself in this thread at the WizardofVegas message board:

"Revel promotion Sounds too good to be true"

The following post (SicBoPro, July 3rd, 2013 at 1:16:02 AM) is not atypical:

I am another player who has been given the proverbial middle finger by Revel. My losses cracked five figures before I was shut off. Specifically, this is what I was told by the people who run the Revel Card program ... They were provided a list of names that included my name; they don't know why I am on it, but they deactivated my card because of that .... I no longer permitted to participate in the slot refund promotion, but my losses up to that point will not be refunded, either; this is arguably the most troubling aspect of the whole ordeal....

There is another side to every story, and no doubt some APs have pulled some stunts. However, Revel appears to have been caught completely off guard by the sheer breadth of advantage play opportunities made possible by means of a 100% loss rebate. The avalanche of challenges and opportunities this promotion set in motion has not yet found its bottom.

Received his Ph.D. in Mathematics from the University of Arizona in 1983. Eliot has been a Professor of both Mathematics and Computer Science. Eliot retired from academia in 2009. Eliot Jacobson