4 ms·
I've worked at 3 gaming manufacturers including implementing the payout table. Near misses were never coded into the design. If they did occur, they were conseq
by digler999 10y ago
I've worked at 3 gaming manufacturers including implementing the payout table. Near misses were never coded into the design. If they did occur, they were consequences of the way the reels were physically laid out. That choice was made from an excel sheet that outlines how many types of each award were expected. The primary statistics the mathematicians cared about were volatility (std deviation) and hold percentage. I can almost guarantee that no modern manufacturers have coded logic that says "If a loss was chosen, re-arrange the loss to appear close to a winning combination". There is no need to do that.
Also, virtual reel maps allow you to have wins over 10,000 units. That enables choice, some people might want to play a large-award machine. There is no deceit there. If the award says "8 million dollars", you know that winning symbol does not have a 1/22 chance of coming up. I have played megabucks, a progressive slot with a $10mm award, for a couple hours. I believe I got one winning megabucks symbol on the first reel, perhaps 4-6 times. That seems in line with its rough expected probability. No deceit. If the first symbol, and (!!) the second symbol came up with their true 1/22 probability, but conveniently the third symbol always missed, it might fool some people initially. But the people who play these machines often would figure it out and their gambler's superstition would kick in. They'd say "those machines are rigged, I've gotten 2/3 winning symbols hundreds of times but I never get the third. I wont play them anymore."
For anyone who is outright against all gambling, its important to remember that you will never shut it down. If you make it illegal, people will go online, or the mafia will have their "numbers" game, people will bet on sports through bookies, etc. The shadier the source, the dirtier it gets. I've heard of casinos in south america that had "inaugural wins", where they'd rig the machine to pay them the first time they sit down in order to hook them. That is sick.
The first reform I'd like to see is the $1200 tax limit increased. Thats an unfair tax on everyone, addict or not. It should be raised to account for the cost of living and inflation. I don't think it's been changed in at least the last 15 years.
Second, I think the optimal solution here is to legislate some sort of "gambler's bankruptcy", where someone surrenders their assets to a court, which then puts them on a forced budget and severely impedes their ability to gamble. That way it could perhaps prevent them from being homeless, sued, etc.
Third, I think that if you embezzle or otherwise use stolen funds to gamble that should absolutely be recoverable by the victim. If I buy stolen property, the court has no problem seizing that from me and returning it to its rightful owner. If the casinos sell entertainment and receive stolen funds, they absolutely should suffer the consequences just like an ordinary citizen. If they see a "whale" coming in and extend all these benefits to them, it should be their onus to verify that person's income.
- kirykl 10y agoExamples of a near miss that are indeed part of the design: - third wheel spins that change visuals/audio to enhance anticipation of landing that last needed symbol - symbols that look very close to eachother - symbols that appear to be stopping then move at the last moment - bonus wheels with false probabilities, or fake-out stops at high value slices - pre-determined picking on features
- digler999 10y agoDo you think there is a difference between a probabilistic near-miss - say, being dealt 4/5 cards to a royal - and a machine that already decided you lost and in attempt to tempt you, grabs 4/5 to a royal and shows them to you knowing full well you wont win ?
- kirykl 10y agoNot to the user. The internals are a black box, and what's presented is taken at face value. But In poker the user might expect the presentation to be based on probability (52 choose 5) and the 2nd example would result in more near misses than expected, exploiting the effect of a near miss. In slots there's not really such an expectation rooted in probability beyond what the machine presents on screen
- ianferrel 10y ago> Third, I think that if you embezzle or otherwise use stolen funds to gamble that should absolutely be recoverable by the victim. If I buy stolen property, the court has no problem seizing that from me and returning it to its rightful owner. This is a nice idea, but it's not a workable system for money. For one, because money is fungible, you can't actually track who has it. If someone steals $1 million from an employer, and also drains all his personal assets (total $1 million), gambling a total of $2m at five different casinos over the course of several months, who do you claw the money back from? Who got the stolen money? How could any casino reasonable have figured anything out here, since he could easily have proved he had $1m in assets prior to the gambling?