Ask The Wizard #211
Is the idea of a 'parity hedge' in craps a real concept?
No, I had to look this up to understand it. It seems to be a lighthearted urban myth involving some young scientists who allegedly found a way to win at craps. This anecdote is shared at Quatloos . I would categorize this alongside other fictional tales that are mistakenly believed to be true, like Joshua’s missing day . As I've pointed out numerous times, betting systems cannot outperform games such as craps; in fact, they can't even make a dent in the house's advantage.
One evening, during a poker league, a situation arose. The blinds were set at $300/$600, and the initial player wagered $2,000. The next two players folded, and it was my turn. Without realizing the original bet, I erroneously put in $3,000 without specifying whether it was a raise or a call, assuming I was the first to act. I started to retract $1,000 to just call but was informed that my chips must remain in play and that I would either need to add $1,000 to raise the initial bet or simply fold my hand. Can you clarify the correct ruling in this scenario? Thank you.
I believe both of you are misunderstanding the situation. You were right that by betting more than $2,000, you were suggesting a raise. The minimum increase should have been $1,400, bringing your total bet to $3,400, which contradicts the ruling at the table. This is due to the fact that the last bet of $2,000 includes a call of $600 and an additional raise of $1,400. Thus, your bet was just a $1,000 increase. Therefore, you needed to add another $400 or fold your hand. ( source )
At BetJamaica I played a total of 30 hands and ended up losing 21.5 units. What are the odds of that happening?
In blackjack, the standard deviation per hand is around 1.15 according to Vegas Strip rules ( source ). This value might vary based on specific rules, but since you didn't provide that information, we will work with 1.15. Consequently, for 30 hands, the standard deviation will be sqrt(30) × 1.15 = 6.30. While I'm not sure about the specific blackjack rules in play, let's assume there's a house edge of 0.4%. Therefore, out of 30 bets, you should expect to lose 30 × 0.004 = 0.12 units. This means your losses were significantly higher than projected by 21.5 - 0.12 = 21.38 units. Dividing that by the standard deviation gives us 21.38 / 6.3 = 3.39 standard deviations below expectations. The likelihood of this outcome is 0.000349, or 1 in 2862. I'm afraid this might not warrant any allegations. If you still feel something is off, it might be beneficial to gather more data.
I have a friend who once worked in a casino managing roulette games, and he mentioned that when players start to have success, the casino changes the dealer. I've also witnessed a staff member instruct a dealer to alter the speed of the roulette spins. Doesn't this indicate that casinos believe the dealer can influence outcomes in a non-random manner? Could this imply that players might find a 'lucky' table where a dealer consistently produces favorable spins?
Unfortunately, a lack of understanding can extend quite far in this field. I won't disagree that a keen observer might be able to time the wheel on a notably slow spin. However, aside from that point, swapping dealers does not alter the odds of the game. There’s no such thing as a fortunate or unfortunate dealer. It’s tough for some people to shake off their superstitions. I’ve repeatedly emphasized that the more outlandish a belief, the more stubbornly it is typically held.
I genuinely appreciate the praise for this website; it's filled with valuable insights! To provide some context: when deploying a Random Number Generator (RNG) for payouts in a finite system, like a million lottery scratch-off tickets, the RNG can be designed to either exclude non-winning options or include winning ones to ensure a fairer distribution of winners among the tickets produced. The aim here is to establish a balanced distribution within the cards while adhering to the required payout percentage. Is it feasible for such programming to take place in Nevada? Statistically, one might argue it's unnecessary, but isn’t it theoretically possible for a slot machine designed with a 97% payout to yield only 95% one year and 99% the year after, unless some oversight on the RNG is implemented?
Thank you for your kind feedback. Indeed, scratch-off tickets and pull tabs can be printed in batches. Each batch will have a designated number of win outcomes, and the overall return from that batch will align precisely with the manufacturer’s expectations. In some regions, where only pull tabs are lawful, outcomes can even be shown on a video monitor, similar to slot or video poker machines. However, the process is different in Nevada, as each play in the slots operates independently of previous outcomes. A machine expected to give a 97% return can easily payout less than 95% or more than 99% over the span of a year, especially if it's not played frequently.