|
Also whoever mentioned XCom’s RNG BS, they deliberately added a “save scum” option that you need to turn off if you want a different outcome.
|
#
?
Sep 8, 2023 13:37
|
|
|
|
| # ? May 9, 2024 02:27 |
|
|
exmachina posted:Another game that deliberately uses poor RNG generation is Doom. There is a preset list of scrambled numbers from 0-255 and each call for random values just takes the next value off the list. This has been abused by TAS players. I mean, I don't know if I'd call it 'poor' except in a pure mathematical sense. Most humans won't ever recognize it. Also, with a static RNG table, that helped allowed the easy sharing of demos (since they are just user inputs that will get the same reaction) and that helped allow for some of the world's first speedrunning. For anyone who hasn't seen it, here's a video by some guy describing the computer code behind the Doom RNG. It's actually super fascinating. https://www.youtube.com/watch?v=pq3x1Jy8pYM
|
|
#
?
Sep 8, 2023 14:35
|
|
|
Schwarzwald posted:Back when I worked in a casino, the big new thing the suits were excited/worried about was Puzzle & Dragon. To hear them describe it, it was a gambling game where even when the player wins they don't get any money. if you have kids you need to make them watch this presentation, all the mobile game brain hacks straight from the horse's mouth and they're proud of it: https://m.youtube.com/watch?v=xNjI03CGkb4
|
|
#
?
Sep 8, 2023 15:16
|
|
|
On the topic of rng, are those electronic blackjack games you get in casinos and sometimes in high street bookies actually rigged to hell or does it just feel like it? I swear I've never seen the dealer get exactly what they need for 21 so often, even off starting cards like 5 and 6
|
|
#
?
Sep 8, 2023 15:22
|
|
|
ThomasPaine posted:On the topic of rng, are those electronic blackjack games you get in casinos and sometimes in high street bookies actually rigged to hell or does it just feel like it? I swear I've never seen the dealer get exactly what they need for 21 so often, even off starting cards like 5 and 6 Usually these are regulated to not do what you're saying. In principal it's easy to make a computer program that cheats though.
|
|
#
?
Sep 8, 2023 15:45
|
|
|
ThomasPaine posted:On the topic of rng, are those electronic blackjack games you get in casinos and sometimes in high street bookies actually rigged to hell or does it just feel like it? I swear I've never seen the dealer get exactly what they need for 21 so often, even off starting cards like 5 and 6 iirc Nevada state regulations require that any video representation of normal game objects (e.g. cards, dice, roulette wheel, etc) must follow the same rules and probabilities as their real-life counterparts, because otherwise you'd see a huge race to the bottom. Edit: That said, video blackjack tends to suck rear end, especially if it's one of those multi-game IGT machines or whatever. The paytables and rules are very heavily weighted against the player. I've seen ones that pay 1:1 on a blackjack, which, uhhhhhhh CapnAndy posted:Psuedo RNG in layman's terms: true randomness is very, very hard, but 99% of the time, all you need is to be random enough. I wouldn't call a weighted die or any other similar thing pseudo-RNG; they're still random. They're biased. There is still a lot of real chaos that goes into those rolls or throws and the results of any given roll are unpredictable. You can take a weighted die that lands on 1 50% of the time and predict it'll be a 1 and be right roughly half the time, but the outcome is still random. The randomness is just distributed unevenly. A pseudo RNG does not feature real randomness in the same way but functions on a series of potentially predictable inputs. If you know that the PRNG takes (x, y, z) and generates a new number out of it, even if the outputs of it are evenly distributed and fair, it's not generating purely random numbers, and if you know the inputs going into it and the function it uses you can predict further outcomes. It's a subtle, but important, difference.
|
|
#
?
Sep 8, 2023 22:06
|
|
|
Zamujasa posted:I wouldn't call a weighted die or any other similar thing pseudo-RNG; they're still random. They're biased. There is still a lot of real chaos that goes into those rolls or throws and the results of any given roll are unpredictable. You can take a weighted die that lands on 1 50% of the time and predict it'll be a 1 and be right roughly half the time, but the outcome is still random. The randomness is just distributed unevenly. CapnAndy fucked around with this message at 03:26 on Sep 9, 2023 |
|
#
?
Sep 8, 2023 22:33
|
|
|
Also, famously, there are sharp-cornered moderately more expensive dice sold to RPG gamers since like the early 1980s by beloved crochety old man Lou Zocchi, founder of Gamescience, who would accost people at every convention to lecture them about the horrors of lovely uneven cheap dice and the wonders of his revolutionary products. https://www.youtube.com/watch?v=MRzg_M8pQms Zocchi was a national treasure and actually a really nice guy. Anyway yeah the point stands, just... good dice actually aren't much more expensive and you can totally spend an extra dollar to upgrade your lovely board game or whatever.
|
|
#
?
Sep 9, 2023 01:41
|
|
|
ThomasPaine posted:On the topic of rng, are those electronic blackjack games you get in casinos and sometimes in high street bookies actually rigged to hell or does it just feel like it? I swear I've never seen the dealer get exactly what they need for 21 so often, even off starting cards like 5 and 6 Legal gambling equipment is put under the most powerful microscopes regulatory bodies can develop. Getting caught rigging them has a high chance of causing a casino to cease to exist (via getting their operating licenses permanently revoked). They're not fair. The payouts are quite favorable to the house. But they are honest, because not being honest is outright suicidal for the casino.
|
|
#
?
Sep 9, 2023 02:28
|
|
|
flashbacks to the goon who posted a really cool ask/tell thread about working at Chessex and eventually had to delete his account and have mods memory hole the thread because some obsessive freak would not stop demanding proof they audit all their dice to ten decimal places
|
|
#
?
Sep 9, 2023 03:57
|
|
|
shame on an IGA posted:flashbacks to the goon who posted a really cool ask/tell thread about working at Chessex and eventually had to delete his account and have mods memory hole the thread because some obsessive freak would not stop demanding proof they audit all their dice to ten decimal places I've got two unread posts in that thread (it's still in my bookmarks!) and I assume those posts are the OP asking for the thread to be hidden from public view and a mod responding with an approval. It was kind of funny when it was just at the level of "they're just dice? for board games? Our business is in making dice that look cool and are fun to own.", but I'm sad the thread went away.
|
|
#
?
Sep 10, 2023 00:56
|
|
|
ThomasPaine posted:On the topic of rng, are those electronic blackjack games you get in casinos and sometimes in high street bookies actually rigged to hell or does it just feel like it? I swear I've never seen the dealer get exactly what they need for 21 so often, even off starting cards like 5 and 6 https://www.verywellmind.com/what-is-a-confirmation-bias-2795024 but to answer your question, yes gambling is rigged and you cant win at it
|
|
#
?
Sep 10, 2023 08:56
|
|
|
It's considerably worse in the UK and depending on the machine you are playing there isn't even an RNG to speak of. The slot machines in pubs are made to look random, but what you are going to win and how is decided far in advance. So much so that there was action taken years ago where a notice is now on all machines to state that you may be given something that looks like a choice (hi/low) but neither option will win. Then there are compensated machines. Which really plays into the usual gambler's fallacy of a machine being "due" a win after a long string of losses. Effectively the likelihood of you winning *is* dependent on what wins have come prior to you playing. Both of these are completely legal here and very common.
|
|
#
?
Sep 10, 2023 15:14
|
|
|
Tsietisin posted:...a notice is now on all machines to state that you may be given something that looks like a choice (hi/low) but neither option will win. I looked into a bootleg NES game that looked like it had been designed to be one of these, and reverse engineered a good amount of how it worked behind the scenes, and the hi/lo game is pretty much exactly as described. 1. Show a card to the player and ask for hi/lo. 2. Determine if the player should win, based on some factors. (Notably, if you would win "too much", i.e. more than you had put into the machine, you lose 100% of the time) 3. Pick a random card from the deck that hasn't been shown yet. 4. If the card would result in the "wrong" outcome, go to 3. 5. Reveal the card and pay out accordingly. The actual hi-lo game itself is a complete fabrication and the result is solely determined by what the game wants (and how much it decides luck plays a factor). quote:Then there are compensated machines. Which really plays into the usual gambler's fallacy of a machine being "due" a win after a long string of losses. Effectively the likelihood of you winning *is* dependent on what wins have come prior to you playing. This is also really common in redemption arcades. One example I have offhand is Bar-Ber-Cut Lite; see page 9. The game can be (and usually is) configured in a way where it simply will not allow a player to win unless x number of games have been played, where it will go into skill mode. They also advise pre-cutting one of the strings to make it look like someone else already won. Casino games in the states and Nevada specifically seem like a major outlier in that the games themselves are "fair", they're just designed in a way where "fair" is "pays out 60%~99% on average".
|
|
#
?
Sep 10, 2023 19:31
|
|
|
Zamujasa posted:This is also really common in redemption arcades. One example I have offhand is Bar-Ber-Cut Lite; see page 9. The game can be (and usually is) configured in a way where it simply will not allow a player to win unless x number of games have been played, where it will go into skill mode. They also advise pre-cutting one of the strings to make it look like someone else already won.
|
|
#
?
Sep 11, 2023 04:31
|
|
|
Why play slot machines, which loving suck rear end, when there is the option to play poker, and thus have the satisfaction of crushing your opponent's psyche beyond all hope and repair?
|
|
#
?
Sep 11, 2023 04:33
|
|
|
That's a rhetorical question but to give the honest answer: playing poker at a casino for money for the first time is way, way, way more intimidating that playing a slot machine for the first time.
|
|
#
?
Sep 11, 2023 06:21
|
|
|
A slot machine is a dopamine factory with all the flashing lights and bells and whistles, a game of poker requires actually thinking and skill. Sometimes you just want to shut your brain off and flush money away.
|
|
#
?
Sep 11, 2023 06:36
|
|
|
EorayMel posted:Why play slot machines, which loving suck rear end, when there is the option to play poker, and thus have the satisfaction of crushing your opponent's psyche beyond all hope and repair? But then there is also the chance that it is your psyche that gets crushed beyond all hope and repair...
|
|
#
?
Sep 11, 2023 10:06
|
|
|
Play poker long enough and you start thinking you did good even if you lose because you ‘played it right.’
|
|
#
?
Sep 11, 2023 10:40
|
|
|
Ritz On Toppa Ritz posted:Play poker long enough and you start thinking you did good even if you lose because you ‘played it right.’ Look, they had one out. I'm losing money if I don't get it in there, even if I lost my money this time.
|
|
#
?
Sep 11, 2023 14:05
|
|
|
CapnAndy posted:Yeah, this is legal in America too. Any arcade game that looks like a skill challenge and is offering big prizes for a win is in fact a giant loving scam that will modify your inputs and force losses until it's gotten enough money to meet a preset amount of profit, and only then will it allow someone to actually win. The string cutters, the key pushers, the stackers, anything. All scams. This made me think of something else slightly relevent for us UK peeps. Outside of the various ticket games out there, if a slot machine mentions the word "Skill", it has to be exactly that, a test of skill. It doesn't need to be either fair or easy, but if you have the skill to win then you will be able to do exactly that. You'll find most machines have a "Skill Stop" or similar feature. They will go ridiculously fast for the higher win amounts, but if you are good enough, you can win the jackpot every time.
|
|
#
?
Sep 11, 2023 15:54
|
|
|
One of the ways Stacker and its clones get around that is that there's always a "skill window" where you can win even if the game hasn't hit its payout % yet, it's just so small as to be effectively blind luck to hit (and of course the visual representation of the game doesn't match because the internal timing is much tighter than what the lighting shows)
|
|
#
?
Sep 11, 2023 17:54
|
|
|
Zamujasa posted:One of the ways Stacker and its clones get around that is that there's always a "skill window" where you can win even if the game hasn't hit its payout % yet, it's just so small as to be effectively blind luck to hit (and of course the visual representation of the game doesn't match because the internal timing is much tighter than what the lighting shows) https://www.youtube.com/watch?v=vXBfwgwT1nQ
|
|
#
?
Sep 11, 2023 18:01
|
|
|
I don't know how much I trust YouTube dude, where he's testing against a phone's touchscreen. He doesn't seem to show that it's capable of hitting times other than 1.000 when testing it, like 0.999 or 1.001. If you assume a 60hz display on the phone, you're looking at 0.01667 seconds every frame, which is a lot less accurate than you actually need for this kind of test. And if the timer is only capable of giving 1/60 sec results, then anything within that bounds (especially something as tiny as ±0.003 sec) isn't going to register differently either.quote:MODE 34 He then goes onto highlight another part of the manual to claim that it's clearly rigged and unwinnable, ignoring the rest of the text: quote:MODE 35 This is exactly what I mean -- the game isn't unwinnable, the timing window is just so small as to make it more or less random chance, and when it's had enough it opens the window to more "human" ranges. The game isn't unwinnable otherwise, it's just extremely difficult, and impossible to reliably time. In the end, the difference doesn't really matter, because the game is so difficult that skill merely makes winning more likely rather than ensuring it. But it's different from purely unwinnable. Sort of like a lottery. 
|
|
#
?
Sep 11, 2023 23:07
|
|
|
Tsietisin posted:It's considerably worse in the UK and depending on the machine you are playing there isn't even an RNG to speak of. I believe by law, if a prize is shown it has to be winnable
|
|
#
?
Sep 12, 2023 11:37
|
|
|
https://twitter.com/NeckarValue/status/1701706301829730668/photo/1
|
|
#
?
Sep 12, 2023 23:00
|
|
|
Vavrek posted:I've got two unread posts in that thread (it's still in my bookmarks!) and I assume those posts are the OP asking for the thread to be hidden from public view and a mod responding with an approval. Huh? Why wouldn’t the guy who kept demanding dice audits get warned then probated for ruining the thread? Or was he getting banned then kept reregistering?
|
|
#
?
Feb 4, 2024 06:50
|
|
|
Hyrax Attack! posted:Huh? Why wouldn’t the guy who kept demanding dice audits get warned then probated for ruining the thread? Or was he getting banned then kept reregistering? It was a different moderation regime at that time.
|
|
#
?
Feb 4, 2024 08:13
|
|
|
That thread wasn't memory holed because of the weirdo - I think they even got probated at some point? The OP requested it because they didn't want to out where they worked, but ended up giving too many details anyway that goons connected the dots and then got paranoid someone would see it, so they requested mods to hide it.
|
|
#
?
Feb 4, 2024 08:19
|
|
|
Exactly that. As a mod I can see that thread and that's basically what happened, OP asked for the thread to be obliterated because their work was concerned about potentially proprietary info being leaked to competitors and the poster was worried the thread could be brought up and get them written up or fired or something, and an admin immediately took care of it. It wasn't because of some weirdo or doxxing or anything, just a pretty reasonable worry that posting your company's technology on the internet for everyone to see isn't a great idea.
|
|
#
?
Feb 4, 2024 08:43
|
|
|
Great thread, what’s the usual casino policy on outside food & drink? Like I know probably couldn’t have Doritos at a small stakes table but if it were high enough stakes would they ignore it if other players weren’t complaining? Same with if someone had a big gulp or something at a machine they’ve been playing for hours? Also wanted to mention there was another good ask tell casino thread a while back from a goon who was security at a Tribal casino, had fascinating stories: https://forums.somethingawful.com/showthread.php?threadid=3859559&userid=0&perpage=40
|
|
#
?
Feb 8, 2024 01:53
|
|
|
I don't think I've ever been at a casino where they gave a poo poo about outside food or drink.... They aren't making their money on the cocktails or burgers.
|
|
#
?
Feb 8, 2024 02:49
|
|
|
At mine they didn't care about snacks or soft drinks and such, but anything alcoholic was a no-go. My state was very aggressive about overserving (and outright banned free/cheap drinks), so losing the ability to track how much somebody was drinking and cut them off was something that could get us in a lot of trouble. This may not be universal.
|
|
#
?
Feb 8, 2024 03:26
|
|
|
Yeah you can get away with a lot of poo poo but they don't gently caress around where liquor licenses are involved
|
|
#
?
Feb 9, 2024 12:52
|
|
|
Baddog posted:I don't think I've ever been at a casino where they gave a poo poo about outside food or drink.... They aren't making their money on the cocktails or burgers. yea regs are constantly bringing snacks cause they dont win enough to eat casino food
|
|
#
?
Feb 10, 2024 09:55
|
|
|
BiggerBoat posted:Any instances of crooked dealers or employees trying to couple up with players for a cut? Seems like a dealer working with their friends in a way that's not incredibly obvious would be a relatively easy thing to pull off if no one got greedy about it. But, of course, they always get greedy and I know the cameras are really tight as well and catch crooked players. quote:a business where a single corrupt dealer can cost the casino US$250,000 in 15 minutes (if a dealer lets a player use a “pre-ordered” deck in a table game, for example) I know this is a bit of a necro, but some of the questions in this thread ran smack into an area I used to work in and some of the research I used to do way back. At that time, the company I worked for was one of the first doing electronic prescriptions (along with lab results and insurance claims), so the thought was to tie them together for long-term health data. I was very much concerned with privacy as we were planning on keeping this data for decades (and HIPAA was brand new). One of our big problems was keeping track of people when there's so much crappy writing and careless data entry. So a name like Robert Smith, Bob Smith, Rob Smith & R J Smith can look like different names. So we had to spend time trying to determine if they're the same person or not. This problem is called "Entity Resolution" or "record matching". And as part of my research into the subject, I came across a guy named Jeff Jonas. He was one of the lead guys in this area and he started by working for casinos in Vegas. His approach was called Non-Obvious Relationship Awareness. quote:An employee who works in surveillance has just put in for an address change in your payroll system. This same address is consistent with that of an individual arrested early last year for a $375,000 baccarat scam and now serving time in jail. You had always suspected help from the inside but had no evidence. This latest piece of data in the payroll system would be an important lead, but how would you ever discover this important new fact? He mentioned that crooked employees are a bigger threat to casinos than crooked players. Some countermeasures the casino might use would include the bit mentioned earlier, where the dealer has to show their empty hands to the cameras after scratching their nose. Jonas' & IBM's product checked data, like does the dealer live at the same address as someone who got trespassed? Or are they married to someone with a loyalty card? And these checks are done frequently, not just a single background check, done once and never again. He calls this Data finds data. IBM bought his company and merged the product into some of their database products. DoD & In-Q-Tel invested in his company, took some of his ideas and rolled them into tracking terrorists. If the R Smith problem was hard, keep in mind that there are about 20 different ways to render "Mohammed" in Roman letters. Most of Jonas' articles from 2002-2004 were of the sort where he was demonstrating how his product could have identified all the 911 hijackers *before* 911. As a result, almost all of the research dropped off the internet, and then most of the fluff articles disappeared. Much of the string matching research had been diverted into the Human Genome Project (DNA is a long sentence of 4 distinct letters, so it was of interest to companies with deep pockets) and after 911 it all disappeared into "defense research" (who have the deepest pockets of them all). Vegas 911, his history. "Threat and Fraud Intelligence, Las Vegas Style", some examples of what they're looking for and some pseudo-code for finding it. If entity resolution is a subject that you might find interesting, here's a decent entry level book and the thesis where I started my research on the subject. The last employer where this sort of research/idea would have been useful was my state's department of motor vehicles (officially called Transportation Cabinet, it also included "the highway department"). There were a number of people who left the vehicle registration in a dead relative's name, but being dead, they would not get their driving license suspended for lacking insurance. One of my after-hours projects was going through the dead peoples and updating their record in the vehicle registration system (totally separate from the driving license system), this would cause them to have to update things next time the car got registered. This state used to be one where people laundered salvage titles, so there are lots of scams involving cars still going on.
|
|
#
?
Mar 26, 2024 16:48
|
|
|
I'm interested in data and names and all this poo poo, that's a really informative post thank you Huttan
|
|
#
?
Mar 26, 2024 17:25
|
|
|
BiggerBoat posted:Speaking of that, what are the background checks like to work in a casino? Yeah, they crawl right up your rear end and take the grand tour. It wasn't even the casino doing the background check, technically; it was the government. To work in a casino you need a gaming license, and to get a gaming license, well... I no longer worry about privacy, because that's just not something that applies to me. I had to give them a full set of full and partial fingerprints, have my picture taken for facial ID, give them references, tell them about every time I ever was convicted of a crime -- and that form does not include the standard "excluding minor traffic stuff" language -- and give them permission to do a credit check, read my tax returns, and check all my bank accounts. Best part is that when your gaming license is up for renewal, you get to do all of that again, except for the references. And they care very much about all of it. Last time I got my license renewed, I got an email from the gaming control board all "hey, you missed some student loan payments, buddy" and luckily in my case I was able to go "yes but I've already got that taken care of and I have the paperwork to prove it", because I genuinely don't know if that would have been a disqualifier otherwise. And just to be clear, this was not a form email, this was the person who was processing my renewal reaching out personally because they'd seen something and wanted to give me a chance to explain. And yes, the casino did drug test me when I first got the job.
|
|
#
?
Mar 26, 2024 17:36
|
|
|
|
| # ? May 9, 2024 02:27 |
|
|
I took a class around that time period in college (I'm old but I just don't want to admit it. But it was after 911 and before Obama got elected) on Biological Programming as an elective. I forgot the exact name but it was work related to biology, like generic programming, and stuff people in the field of bio would use code for. One of the things was determining DNA sequence similarity using Dynamic Programming and Levenshtein Edit distance as a edit string metric.. So if just applying that to Arabic names somehow got you a ton of money from the DOJ and/or IBM. God drat. I missed out!
|
|
#
?
Apr 3, 2024 08:25
|
|
