Hacker News Reader

Comment by fleshdaddy | Hacker News Reader

I see some folks in this thread underestimating casinos and exactly how far the odds are tilted against the players.

I grew up in Vegas and have a lot of family and friends who work in the casino industry. A particular family member is an executive at one of them and told me that he has control over the odds on machines (within the legal limit) down to a particular machine being used by somebody. Apparently this means if they need a particular high roller to stick around for awhile and spend some money they can go ahead and grease the odds on the machine they’re using to give them some winnings to spend. Then they can just make it back the next day or whatever by cranking it back and making the odds worse. Honestly at first I thought he might be exaggerating but some time later I had a friend who worked corporate finance at one of the big casinos independently tell me the exact some thing.

These casinos though are really no joke. It’s one thing to skew the odds, and most people understand that, but being able to skew them in real time feels a lot more rigged. I assume some machines don’t operate this way because for many years the smaller casinos had machines that you could break even or win a bit consistently in poker, and some casinos even advertise odds. Still though its pretty shocking if my understanding is correct.

View on HN

I call bullshit. There’s no way an industry as highly regulated as gambling which is effectively a license to print money, would be manipulating the odds on machines that are individually registered with the state gaming commission. The risk to the business for being completely shut down outweighs any weekend profit for getting a whale to stick around.

It probably depends how the regulations are written. Can the odds be _better_ than advertised without the casino getting in trouble? Then they'd quite plausibly have a range they could play with.

As far as I know, they have to be exactly equal to the advertised payouts. This prevents stupid shit like making one of your slot machines "hot" as an advertising trick.

Game odds are very rarely advertised. Sometimes it is trivial to calculate or simulate the game (blackjack, roulette, craps, video poker) because all of the game parameters are known (i.e. 52-card deck, 38 slots on an American roulette wheel, two 6-sided dice) as are all the payouts.

Slot machines, however, are opaque because the probability of hitting any specific combination is unknown to the public. The machine can promise that a 7-7-7 result will pay $5000 but you have no idea how often a 7-7-7 is supposed to hit.

Sometimes a casino will take some less popular machines and configure them with an extremely low payout percentage (3% for example) and because the players would not otherwise know this, the casino will put a sign up "97% payout!" to try to attract some value-oriented players, just like a retail store takes old merchandise and puts it in a bargain bin at 50% off.

What rule is that against? I don’t necessarily doubt you I just dont have any intuition about this. Because it seems like its already completely legal to play unfair games.

There is no need to tweak the odds. Every machine is scheduled exactly to give a certain amount back while raking a profit. They're not there for you to profit, at least in the long term.

An example of this is the simple claw machines. One was full of teddy bears and I knew the operator. He told me point blank the machine will drop 16/17 times, "someone's got to pay for the teddies, and he ain't me!"

In many cases the stuffed animals at carnival games cost less than the $1 it costs to play the game—you could win every play and they’d still make money off you.

Yes, but remember that an operator has further costs.

First, he has to pay the machine that wasn't cheap to buy outright.

Second, he has to share revenues with the venue hosting the machine. That goes between 20-60% AFAIK.

Then there is the maintenance cost, time spent lubing and adjusting, and restocking + travel time.

So yes, overall, the operator better make sure you don't just pay for the teddies!

It's highly-regulated in favor of the house. These gambling concerns are criminal-adjacent corporations with political connections and weak campaign finance laws.

To call them "criminal-adjacent" seems like a conservative western tradition. I'm not a casino apologist, but it's not like they have some hidden agenda. It's gambling. Did you misunderstand that? It's a game of odds and the house controls the odds. Did you misunderstand that?

Edit: Clearly you don't misunderstand that, so what's the complaint?

Gambling has a very strange effect on the human psyche. I never hear anybody come back from Disney World bragging about how much they won. But I've met tons of people who "always win in Vegas". I know a few people who seem addicted to Disneyland. I never understood it. Season passes every year. Maybe one visit a month. Even when they don't have children. It seems weird to me. But the effects of gambling on people's brains is even more weird.

It is only criminally-adjacent in places that perceive a giant risk to people and so have made it illegal. Otherwise, it is as criminally-adjacent as any source of wealth or power that has an incentive to build barriers to entry and to stifle competition.

Yes, people might have biases. In this case though, Vegas really was developed by the mafia.

This is an interesting question. I don't think the first casino in Vegas was established by the mafia. The mafia's involvement later on is well documented. The mafia's current involvement is not so well documented. If you believe that the mafia was eventually exercised from Las Vegas and look at the development before and after that point... We could have a reasonable discussion about whether the development post-mafia dwarfs the development under the mafia and whether the development under the mafia dwarfs the development pre-mafia.

But the claim seemed to me to be about the present day. Today, I think the industry falls about the same place as big-oil or big-pharma or big-auto or big-politics. That is, plenty of people would call them all criminal-adjacent. But it still makes me wince a little. I'm sympathetic to them all being heckled, though.

Gambling existed before the mafia. The mafia isn't that involved with Las Vegas any more.

Casinos have been the way for the mafia to operate legally.

What interest would a criminal organization have in a high cash revenue business?

Could it be a way to launder money? Jimmy "No Nose" comes in with $40k cash every weekend and loses it gambling, and "nobody knows" where a guy like Jimmy gets that cash so regularly.

Gambling has been well-known for some time as a way to launder money. For example see this[1] official government report in the UK or this[2] which is US-specific but much less detailed.

[1] https://www.gamblingcommission.gov.uk/print/The-money-launde...

[2] https://moneysmartathlete.com/betting-industry/money-launder...

The old skim was even better than that. Make $500k, report $300k and send $200k cash back to Chicago.

Casinos are spiritually in the same place as drug dealers and pimps. Profiting off weakness and vice.

If you like spending all your dime on drugs, dice and so forth then fine, but let's not pretend the guys selling it are fine upstanding businessmen.

This is not the right analogy.

'Criminally adjacent' means literally, adjacent to organised crime.

This is not the same thing as 'above bar businesses are 'adjacent' to anything that is not' , or 'aggressive competitive measures'.

You are wrong. And it is arguably worse than that.

I'm a little skeptical about "high rollers" playing on any machine at all, but the machines are, in fact highly regulated and highly controlled. I'm always amused to see ads about the "loosest slots in town." It's not really false advertising, because the slots are regulated and everybody sets them at exactly the legal limit.

But then there is what is allowed to happen when you don't win. If the machine has already calculated that you will not win on a particular "pull" it can manipulate the loss in advantage ways. For instance, it can land on 7-7-spinning..., spinning, one cherry away from 7. You just missed by 1! So close!

The world thinks they have uncovered a big new secret when they realize that social media sites study human physiology to optimize dopamine release in their victims^H^H subscribers^H^H account holders. Casinos have been at it for lifetimes.

I don't know why you would think there is a risk to the business if a casino lets a guest win an extra game or two.

Not every machine is at the legal limit. Just like not every black jack table is at the best odds for casinos.

> the slots are regulated and everybody sets them at exactly the legal limit.

It's unlikely there is any slot machine in Nevada "at the legal limit." Every gaming device must be set to theoretically pay out a minimum of 75% of wagers -- i.e. the maximum it may keep (or "hold") is 25%. This is, of course, a theoretical number, on any given day a particular machine may hold more, or it may have negative hold -- just like if you flipped a coin enough times, occasionally you'd get long streaks of heads or of tails, even though theoretically it should be 50/50.

In reality slot machines are typically configured to hold between 6 and 15%. A slot machine holding 25% would be an awful experience and simply not get repeat play. Slot manufacturers prepare each game with a range of options that the casino can choose from, but none of them would deliberately ruin their own game by offering a 25% hold option on a slot machine.

Nevada Gaming Control Act Regulation 14: 14.040 Minimum standards for gaming devices. 1. All gaming devices must: (a) Theoretically pay out a mathematically demonstrable percentage of all amounts wagered, which must not be less than 75 percent for each wager available for play on the device.

https://gaming.nv.gov/modules/showdocument.aspx?documentid=1...

> If the machine has already calculated that you will not win on a particular "pull" it can manipulate the loss in advantage ways. For instance, it can land on 7-7-spinning..., spinning, one cherry away from 7. You just missed by 1! So close!

Showing a “near win” for a loss is completely different than manipulating the odds in real time in response to the current user.

The displayed result is a loss and what matters is the chance of winning or losing must match the registered odds for the machine as an independent event.

> I don't know why you would think there is a risk to the business if a casino lets a guest win an extra game or two.

Because if they knowingly manipulate the odds of a game they’d be in violation of their gambling license and be shut down by the State.

It's entirely true, modern slot machines get a ticket from a central server detailing the results it will show.

They do analytics on all of this using Splunk btw (they can actually afford it, lol), there are some really cool talks out there if you look around.

This usually depends on the jurisdiction. Class III machines are independent and each one contains its own RNG. Class II obtain their randomness from a central server, and are usually configured to actually be a bingo machine or some other pooled wagering mechanism, and the "slot machine" that you see is a mask for the pre-determined outcome.

In the US, almost all commercial casinos (Nevada, New Jersey, Mississippi, etc) use Class III games while native American casinos use a mix of Class III and Class II. Typically, Class III are preferred by casinos and by gamblers.

Here's another source: https://www.tahlequahdailypress.com/news/state_news/casino-1...

The actual regs matter. For example it is well known that with penny slots, the house takes a larger cut than larger denomination lots. Significantly, something like 9% vs. 5%.

Then supposedly Casinos put higher payout slots in high visibility areas such as the edge of a high traffic walkthrough, or with good sight lines. This would make a lot of sense from a marketing POV, make the winners more visible.

I imagine this works out because by assuming a temporary “odds” debt that is repaid over the next day or so by other customers.

You're assuming regulations == good.

In Vegas, slot machines for example have a legally mandated payout rate. This means it is illegal for them to be random and therefore fair. They are required by law to be rigged.

A process with a particular expected value can still be random.

I think in this case it's not a legally mandated expected value, but a legally mandated payout rate. As in, it actually must pay out a percentage of what's taken in, over a certain time period. I don't think a fully random process can guarantee that.

Of course it can. Unless you start playing word games to manipulate what “fully random” means.

You can set the odds of a fully random process, just take an rng returning a uniform distribution between 1 and 100 and set the bar at whichever threshold you want, or do the same with dice, the odds of any given face is 1/faces, so get however many faced dice you need then set the faces (or a threshold again for numbered faces) e.g. on a 20-faced dice you set the threshold at 3+ and you’ve got a 90% payout rate.

This is obviously incorrect: Let’s say they have to pay on each 500-play window, on true random, operator can end up with all 500 plays as non-paying and break the law.

This is why they are programmed to increase payout odds if there was no payout for long time, and this is why some players wait for such machines. (In modern casinos machines are pooled, so it’s actually more complicated)

An operator ending up with all 500 plays as non paying is not breaking the law, they are just incredibly unlucky. The spins are independent, and no amount of waiting for a machine to be "due" is going to do anything.

That's not how randomness works.

In your example, there is no guarantee that 90% of rolls will be a 3 or above. Statistically, over the long term, that will be the case, but I wouldn't expect the payout laws to be written in such a way that "eventually" you'll "probabilistically" hit your payout targets.

More likely it's something like "you must pay out X over time period Y". You can't 100% guarantee that with a random process. It's possible (and not even all that unlikely) that, over the span of a day (week, month, whatever), you'd end up with only 89.9% of rolls being a 3 or above, and that would violate the payout laws.

I think there's a few missing pieces of information (in this discussion, probably could be looked up by some intrepid soul)

Firstly what is the payout rate against, I could imagine it being:

- Plays - Time - Stake

Secondly, how is compliance defined:

- Convergence in expected value - Degree of breach of threshold

On the second one, it's kind of interesting if the degree of breach is 'never breach threshold', because when a machine is 'first turned on' the first play must always be a win (crudely speaking), if the second play has a sufficiently large stake (such that a loss would break the payout rate threshold) then that probably has to be a win too, etc until people can't keep exponentially raising the stages... This is a boot-strapping problem really but still.

If compliance is in expected value then I think any of the rates can be made to work

For a rate in terms of plays or stake then you can just have a binomial variable with some probability. In this case each player/play would be exposed to the same probability; each play would be independent.

For a rate in terms of time the plays would not be independent because the duration of previous plays and payout would affect what payout with subsequent plays would be, so you would need a dynamic adjustment of odds over the time domain, but this can still converge.

For a threshold like you describe then at such a point you would breach the threshold you may need to payout in a situation where you would have not paid out when converging in expectation.

I think what you are saying is that this shunting of probabilities would make the process not random.

However, I think the probability that a 'shunt' occurs could itself be modelled as a random variable dependent on the play history. Assuming some random properties to the sequence of plays (which is a reasonable modelling assumption), this would still be random.

It wouldn't be 'fair' in the sense of giving each player the same probabilities or for plays to be independent, but it would still be random just a distribution weirder than a bernoulli one.

Also FWIW:

> I wouldn't expect the payout laws to be written in such a way that "eventually"

That is exactly how I'd expect them to be written, seems the most natural way to do things, with the same ultimate benefit.

> payout laws to be written in such a way that "eventually" you'll "probabilistically" hit your payout targets.

This is exactly how they are written. Except, there is no specific payout target (this game must payout 92% of wagers), there is a mandated minimum, any game which exceeds the minimum is lawful. Note the word "theoretical" in each of these:

New Jersey Admin Code 13:69E-1.28A: (a) Except as otherwise provided in this section, each slot machine game which requires a wager shall have a theoretical return to player (RTP) equal to or greater than 83 percent.

Mississippi Title 13, Chapter 12, Rule 12.5: Rule 12.5 Minimum Standards For Gaming Devices. All gaming devices submitted for approval:v(a) Must be electronic in design and operation and must be controlled by a microprocessor or the equivalent. Microcontrollers are allowed.v(b) Must theoretically pay out a mathematically demonstrable percentage of all amounts wagered, which must not be less than eighty percent (80%) or greater than one hundred percentv(100%) for each wager available for play on the device

Pennsylvania 58 Pa. Code § 461a.7: (a) A slot machine may not be set to pay out less than the theoretical payout percentage, which may not be less than 85%, calculated using the lowest possible wager that could be played for any single play, or equal or exceed 100%, calculated using the highest eligible wager available.

They do not have a legally mandated rate, they have a legally mandated expected value and there are strong statistical tests done of the RNGs.

Source: have worked with a situation which required me analyzing the process of getting betting software through regulatory approval.

Edit to add: Here's the relevant regulations in Nevada. https://gaming.nv.gov/modules/showdocument.aspx?documentid=4...

The thing to search for in the regulations is the "Theoretical hold percentage", which is the expectation for the house expressed as a percent. The fact that they call it "theoretical" is probably enough to demonstrate that it isn't a mandated rate, but the theoretical average rate.

I don't think you understand what a legally mandated payout rate means. You can have a slot machine that is completely random, yet gives a consistent payout over a large amount of trials. Every spin is independent.

Fixed game odds vs payout odds can be done and be truly random (still not fair, but truly random). The problem with slot machines is they have to pay out a percentage of gross revenue. This means the game has to track how much money has been put into it and adjust the game odds to compensate, not simply count the number of rolls in the game. This cannot be truly random.

Slot machines do not have to pay out a specific percentage of gross revenue, and odds are not adjusted. The result of any individual outcome is random. Slot machines are programmed to theoretically pay out a certain percentage -- but in reality, some pay out a little more or a little less, just like the quarter in your pocket might land on heads 52 times out of 100 flips, or maybe 47.

You are positing a situation like, hey, that dealer just dealt that player two blackjacks in a row. Therefore for the next three hands we're going to take all the aces out of the deck to make sure there isn't another blackjack. Obviously this is not how casinos work, neither is a slot machine adjusting its odds to hit a specific payout target. The blackjack game, and the slot machine, don't have to, over the course of millions of wagers they are mathematically certain to get very close to that theoretical target.

No one manipulates stock offerings either. The SEC is just too good at it's job.

This is a myth, and a quick google search will support that. Wizardofodds mentions it here: https://wizardofodds.com/ask-the-wizard/slots/progressives/

The link you provided states that the technical capability is there, but Nevada has a rule against changing the odds without notifying the player.

> "The conventional gaming device or client station must be in the idle mode with no errors or tilts, no play and no credits on the machine for at least 4 minutes. After this time, the conventional gaming device or client station must be disabled and rendered unplayable for at least 4 minutes. During the time the machine is disabled a message must be displayed on a video screen or other appropriate display device notifying the patron that the game configuration has been changed." — Technical Standards for Gaming Devices and On-line Slot Systems 1.140

https://wizardofodds.com/ask-the-wizard/slots/progressives/

In other states, there may or may not be such a rule.

Why would you quote anybody but Nevada gaming regulations in an appeal to authority on this question?

If you know anything about gambling, you will realise wizardofodds is the authority, and likely more trustworthy than the Nevada gaming regs…

He's the authority.

Wizardofodds mentions a specific methodology, which is not the issue.

Are odds changed after an idle time? They can be, practically. They can be legally, in NV.

https://gaming.nv.gov/modules/showdocument.aspx?documentid=1...

This is not a myth.

Casinos are required to display the current payout rate on any and all slot machines (EDIT: incorrect, see comment below). The way some machines become profitable is when there's a progressive jackpot that hasn't been hit. Imagine there's a high roller $100 slot machine with a 99% payout. That means each spin you have an expected value of losing $1. Now imagine there's a $100 jackpot that you have a 1% chance of hitting (to keep the math clean). That jackpot is now adding $1 of expected value to each spin, so now your expected value is $-1 + $1 = $0. As jackpots grow, all machines (which offer them) will eventually become +EV, or profitable, to play.

Casinos also regularly offer comps, bonuses, and other non-cash incentives to players who spend a lot of money, so it's pretty easy to end up with a +EV game. It's one of those scenarios where the casino gets to have their cake and eat it too. For instance, the way the jackpot may work is the casino takes $0.2 of each $1 of profit that high roller machine is yielding, and put it towards the jackpot. But when that jackpot gets high enough, the exponentially increased number of spins results in an overall profit gain for the casino, relative to just taking the whole $1 for themselves and having no jackpot.

---

Finally, as a slightly different segment of gaming there are some machines which are "technically" capable of being beaten with skill. For instance many casinos now offer Limit Texas Hold'em poker machines, where you are playing "real" poker against a machine. But limit hold'em is an incredibly technical and complex game which no human is ever going to be able to come out on top of against a strong machine now a days. But it's a lot more fun than playing e.g. chess against a computer because the optimal strategy for heads up limit hold'em makes you look pretty much like a maniac, to somebody who doesn't know much about the game.

> Casinos are required to display the current payout rate on any and all slot machines.

No, they aren’t, at least in Nevada. Some states make the whole casino disclose its average “return to player” metric though.

> Casinos are required to display the current payout rate on any and all slot machines.

This is not accurate in at least Nevada, Michigan, and Maryland. Casinos are required to disclose the overall average, but not the rate of individual slot machines.

Wow, you're absolutely correct. I was indeed referring to Vegas in particular. I suppose I conflated the rules for video poker (where payout tables are and must be listed each machine, which indirectly gives an exact EV measurement with optimal play) with slots. My opinion of slots just dropped lower - something I did not know was possible. That is just so sketchy.

Just to be clear, payout tables absolutely must be listed on slot machines. The _rate_ at which the payout is made is not required to be listed on a given machine.

That is, if you land on three cherries, you should be able to verify you got the right amount of money. But you don't know the percentage chance of landing on three cherries.

Absolutely. The difference I'm referring to is that video poker games must be dealt as if the cards were coming from a fair deck, so the odds of each event happening can be determined relatively easily given a fixed strategy, and consequently so can you exact expected profit/loss per hand, and when a machine can be profitable to play.

That the same is not true in slots is the thing that I missed. That a triple cherry pays 100 to 1 or whatever is meaningless if you have no way of knowing the odds of a triple cherry.

Truth.

Mistake no. 1 is expecting to beat the house in the long term when the expectation of all of these games, save PvP like poker, is less than zero a priori.

The MIT blackjack team might have found temporary counting and signaling systems to exploit, but the casinos were never going to let any group bleed money from them in the long term.

Every now and then some groups may try, but they're not going to beat a system that doesn't tolerate excessive winnings for long.

If you want to go to Vegas for a vacation and make $100 last with your own system of minimum risk, sure, go for it.

You fell for a very old urban legend. Please google it before spreading "friend who worked corporate finance at one of the big casinos" stories. Your friend was just having fun at your expense, as was the family member.

This is actually true that they can and do adjust the odds on these machines for high rollers and the like. It's all done by computer now, no lever. However, there are strict procedures involved. If there is a player on the machine, you can't change the odds or payouts without first kicking the player off the machine, and you have to make the payouts very clear (usually with a placard on the machine), and you have to change that too.

Looks like there’s actually some truth to it. Wouldn’t be surprised by some exaggeration but please google it before leaving a comment dismissing it.

I was curious so I looked it up. This site seems to have some technical details:

> Some slot machines operate on downloadable software, which comes from a central server. In these cases, the casino doesn’t have to open a game and change a chip.

> Instead, they can simply download the software to alter RTP. But just as with replacing an EPROM chip, casinos need to follow their jurisdiction’s guidelines when changing server-based payout percentages.

From https://www.gamblingsites.net/blog/inside-the-myth-that-casi...

There's a little bit of truth here, but it's mostly incorrect. Yes casinos set the odds for any individual slot machine. However the game manufacturers prepare the games with about 5 configurations, even the loosest config will still be well into the casinos favor (it's sort of like how Call of Duty has an Easy, Normal, Hard or Legend mode), i.e. the game configs might be to hold 7.3%, 8.6%, 10.1%, 11.5% and 12.7%. Of course the game's theoretical hold is not revealed to players.

The casinos are allowed to update a game configuration while it is on the floor and idle (not while it is being played). I doubt it has been tested but I believe Gaming Control Board would frown on configuration changes being made to accommodate an individual player, but even if a casino did this, they would not be "giving" a player any winnings, they would just be making it so the player was likely to lose at a slower rate, but still be losing, i.e. dropping from the 11.5% config to the "loosest" 7.3%.

This is mostly a moot point anyway, because it is perfectly legal for a casino to offer a player hundreds or thousands of dollars in free slot play (or free bets), that's how they give a specific player "winnings to spend."

I wonder if it's even worse.. by law they are required to have a certain statistical payout right?

What if the casino can enable "win mode" for their favorite customers staying in high priced rooms, and then turn it to basically zero for foot traffic. At the end of the month the machine paid out the correct ratio... Except you as a general player had zero chance to win because it paid out to their loyal customer earlier in the day.

Don't think its worth the risk. You'd lower the average percentage player payout from like 98% (might be defaulted lower in Vegas, not sure) to low 90s and bump a few to over 100? it doesn't sound worth the risks. regular high rollers will still go there and still be profitable with regular slot odds.

Casinos cannot alter odds on any machine (or rather they could, but it would be highly illegal) and have absolutely no reason to. Variance does that job for them quite fine.

I mean, the software and RNGs are audited. The legal odds have to be true in any given spin. If your suggestion it is only in monthly aggregate was true, they wouldn't waste it on favorite customers. They would accept a bid of millions of dollars to accept free slot machines from a vendor, have their customers lose money all month and then have a vendor rep come in once a month to win a suspiciously large payout that made the math work out. (Or some other third party, since I assume a casino employee could not win on their own machine)

The gaming machines are all networked Windows PCs with payout odds that are controlled by group, which can be defined into sections as small as twelve machines. (The exact number may have changed in the past decade.)

NGCB (Nev. Gaming Control Board) does regular on-site audits of the machines and the central configurations to ensure that they are compliant.

I don't believe this.

But it seems it is true.

The catch is the high roller has a favorite machine, so you change the odds before they arrive.

https://gaming.nv.gov/modules/showdocument.aspx?documentid=2...

"Once a game is initiated by a player on a gaming device, the rules of play for that game, including the probability and award of a game outcome, cannot be changed. In the event the game or rules of play for the game, including probability and award of a game outcome, change between games during a gaming session, notice of the change must be prominently displayed to the player."

Other source, not sure where this reg comes from - "Slot machines in Nevada must sit idle for 4 minutes before an adjustment to payback percentage or game rules can be made. When a change is made, the machine must be taken out of play with a message displayed on the screen noting an update is taking place. The machine then must sit idle for an additional 4 minutes after the change before play is allowed."

'feels rigged'.

My friend, the entire thing is rigged from top to bottom, from the time you see the ad, until you are back at home.

The only way to win is to accept the grift on those terms.

Reply on news.ycombinator.com