$1000 reward to find exploits in my betting system challenges

As many of you know, I offer a $30,000 betting system challenge on my website, Easy Vegas. I recently overhauled the rules to make them more palatable to laypersons (e.g., instead of requiring a billion spins, it now requires only 200,000). But I'm worried that there's some loophole in my rules that someone could exploit to win my $30k. So I'm offering a $1000 reward to anyone who can come up with a system that would beat my revised challenge rules. That means you send me your idea, without having to put any money up, and if your idea beats my challenge rules then you win $1000. The caveat is that if it's clear to me that a system has no chance of exploiting my rules then I'm free to reject it on sight without wasting my time coding a simulation for it.

I should point out that the first time I offered such a $1000 reward here, weaselman won it handily -- though he graciously declined the prize! I've also gotten great advice on this forum from MathExtremist, and also good suggestions from mkl654321, AyeCarumba, atlarson, thecesspit, DJTeddyBear, Doc, SOOPOO, and RiverTern -- thanks to all!

In addition, for a long time I've wanted to offer an alternate, additional "live challenge" that takes place in a casino, to placate those who think that a computer simulation isn't a real-world test (because of their superstitious belief that the results are somehow different). The main problem in doing so is that it's hard to make a real-world test sufficiently long enough to always see it fail. But I'm making a stab at it. As you'll see in my proposed rules below, I've made a number of changes to try to account for the short-term nature of the test. However, it could still be vulnerable to exploit, because of something I didn't think of, so to motivate folks to help me find the flaw, I'm offering another $1000 reward to anyone who can come up with a way to beat the proposed Live Challenge rules (with the same caveat as above). Here are those proposed rules in full:

Quote: Proposed "Live Challenge" rules

The challenge is the same as the regular challenge, except:

1. It takes place in a Vegas (or other U.S.) casino, with you and me (or my representative) present.
2. It's an even-money bet between you and me (not 10-1 odds), for any amount up to $30,000, with a minimum of wager of $50 per hour of play. (Time you spend not betting while waiting for "streaks" or other conditions counts as playing time.)
3. You will choose the number of rounds to wager on before we begin (1000 to 10,000 rounds).
4. All play must be completed within one month.
5. Starting bankroll is half the number of total rounds, in dollars. (e.g., a 1000-round challenge means a starting bankroll of $500.)
6. You win if you achieve a profit of at least $15 for every 50 rounds, on average (roughly equivalent to $15 per hour of play).

Ideas for beating the live challenge will be tested by computer simulation, not in the casino, to save time.

Thoughts?

6. You win if you achieve a profit of at least $15 for every 50 rounds, on average (roughly equivalent to $15 per hour of play).>>>

So if I'm betting $15 on red or black, after 50 bets I need to be ahead by just one unit?

Bob, read the stipulation that the minimum number of rounds is 1,000 with a max of 10,000. That means you'll have to show a profit of at least $300 after 1,000 rounds ($15 per every 50 rounds which would mean 20 X 50 rounds to complete the 1,000 round minimum). The minimum profit at 1,000 rounds with a $500 bankroll would be $300, which is a 60% increase over your starting bankroll. So yes, you're right, you'd only need to be ahead by one unit after 50 rounds BUT you'd have to do it consistently for 20 times that many rounds.

I've been aware that some "rules for a challenge" thread existed, but I've not read it. Partly due to laziness and partly due to the level of senseless bickering that was appearing on the forum at the time. (I think everyone realized that No Stinking Badges was amongst us, despite his having been repeatedly banned elsewhere and the subject of much discussion elsewhere for his incessant off-topic bickering and baiting).

Your Rule 6 should probably worded a bit more clearly to emphasize that it is a final figure and not that every fifty rounds must produce that figure. Obviously the system being tested must not allow for some final ridiculously high bet that just might wind up with a few trials that do meet the fifteen dollar per hour test.

I would suggest you simply provide a hurdle such as a fee for testing that will keep the nutcases away.

As for the "digital simulation versus live testing in a casino", there have already been quite a few posts on "the buzz" in a casino having an effect on players. We would hope that a formal test such as this 30k challenge would be immune to such emotional influences but I fear our hopes might be misplaced. Quite frankly, I would suggest a rule that requires a session or two of digital simulation to have been successful before a "live test" would even be considered.

Your Rule 1 obviously has a need to be tightened up a bit since most strip casinos are not in "Vegas" at all and you want to be able to veto a particular selection of a casino even if you contemplate never really exercising that veto power.

I think you might give some thought to the fact that although some Live Session is unlikely to take place due to any system being similar to all other systems (an absurd concoction of a positive figure out of a series of negative ones), such an event would be crucial from both a financial and publicity viewpoint. Therefore some thought should be given to a casino's willingness to allow photography and to provide a suitable roped-off area free of other players who might be acting in collusion with the system-owner.

If we went to a roulette table for the challenge, the first thing I would say is you're allowed to write down my wins or losses but not the numbers as they appear OR the bet I make. You can deduce if I'm making $15 an hour from my wins and losses and the chip count. There is no where in the rules where it says you get to record anything else.

Betting $1 per round for 50 rounds an hour would satisfy your rules and one would only expect to lose $1.35 per hour on a roulette at this rate, or $27 for the whole challenge. Betting on Baccarat would have an expected loss of half that amount! If one were to bet a lot the first few rounds and bet very little until the end they would have more than 50% chance of winning your challenge. Another way to win would be to make to make small but high variance bets like inside bets on roulette, high payout bets on craps, tie/pair on baccarat, etc. This would make it very likely to be up 60% despite the high house edge. I think you should either forbid flat betting small units, create a maximum spread between minimum and maximum bets, and ask for at least a 100% increase in the bankroll: increasing your bankroll by 60% will happen more than half of the time with the right strategy and your even bet would have a negative EV for you even though the betting system would still lose in the long run. You should also define what you mean by wager. Doubling in games like video poker or placing certain bets in craps carries no house edge. If you consider doubling up as a wager, then one could simply double several times since the bet has no house edge and has a very low variance (only 6 chance out of 13 to lose). This would also make the chances of being up 60% after 1000 rounds rather high!

Here are some ideas that might beat your challenge more than half of the time: Betting $250 on Baccarat until I am up $250 (about 67% chance of happening), then bet $50 and martingale until I am up another $50 (about 96% chance of happening), and then bet $1 a hand, after which I would only be losing only 44% of my hands since baccarat gets many ties. The chances of being up $300 before being down $500 are rather high! Some casinos pay 9:1 on ties, which happen 9.5% of the time. Although the HE is higher the variance would make it very likely for someone betting $25-$30 a hand to be up $30 quite fast and then betting $1 a hand would minimize the loses. Betting very small units in video poker and doubling up as much as possible would also make it rather likely to be up $300. Making any long shot bet would also have a high chance of defeating your challenge since a single successful bet might win more than enough to cover all the loses combined.

Forbidding people to make small bets after a series of big bets would help prevent this kind of exploits but you would still be vulnerable to systems that win often at the cost of infrequent loses since 1000 rounds is not long enough to be absolutely confident a system would fail. Here is an example that would probably bet your 1000 hands challenge more than half of the time: Start by betting a unit on the banker in Baccarat. If you win and are up a unit, start over. If you lose, record the number of units you lose and bet the same number of units again. If you win, raise the next bet by a unit unless winning the next bet would make you up by more than a unit. If this is the case, bet enough to be up by exactly a unit. Here is how it would go: Bet 1, Win, start over. Bet 1, lose, (down 1), bet 1, lose, (down 2), bet 1, lose, (down 3), bet 1, win,(down 2), bet 2, win, (down 0), bet 1 since betting 3 is too much to be up 1, win, start over. Using this system on a high variance bet like tie would make it even more likely to be up after 1000 hands since a single hand could significantly influence your total. This system would have a high chance of beating your challenge since big bets only happen after you win so a big losing streak would only cost a few units but a big winning streak would get the player ahead. Since your challenge requires the player to be ahead 60% after 1,000 hands a system like this would be ideal since only winnings are 'reinvested', making it quite rare for the player to risk many units of his original bankroll. The only thing that would kill a player would be a win-lose-win-lose streak, and this would not be very likely in a game with many ties. I don't know if this would survive 10,000 hands but 1,000 hands should be a breeze.

To summarize my previous post: 1,000 hands is too little and your betting rules are too loose. Any system in which one can make a few long shot bets and then bet the minimum for a few hundred hands would beat it more than half of the time. Any system that increases bet size after winnings would also be very likely to beat it since the player would bet very little most of the time and be up a lot once he hits a winning streak, again making the probability of winning over 50% if one only plays 1,000 hands.

I agree with everyone here, although there are no successfull bet system in the longterm I believe that 1000 hands is too hard to risk. There could be 'lucky streaks' in this small gap.

Quote: aluisio

I agree with everyone here, although there are no successfull bet system in the longterm I believe that 1000 hands is too hard to risk. There could be 'lucky streaks' in this small gap.

There maybe ways of being up after a 1000 hands. However is it a 50/50 shot? Probably not. I also think people under estimate the difficulty of getting ahead and staying there, especially with a small bank roll.

Quote: aluisio

I agree with everyone here, although there are no successfull bet system in the longterm I believe that 1000 hands is too hard to risk. There could be 'lucky streaks' in this small gap.

There maybe ways of being up after a 1000 hands. However is it a 50/50 shot? Probably not. I also think people under estimate the difficulty of getting ahead and staying there, especially with a small bank roll.