A suggestion about the blackjack shuffler

[elon]

Look, I don't have any reason to believe the blackjack shuffler is broken. I played blackjack a little when I first started ~1.5 years ago. It seemed fine. Haven't played since.

However, some of us were talking about all the complaints about it -- maybe they mean as little as the complaints that the poker is rigged (If it is, it's rigged in my favor, and I'm fine with that.)

So I started thinking, what's different between blackjack and poker? The most obvious answer is the 6-deck shoe in blackjack. It's easy to have index-, modulo-, etc- bugs (or some combination of these and other bugs) in a given Fisher Yates implementation.

I'm sure you ran some LONG test samples of the 6-deck output and statistically analyzed its distributions, but in case no one has looked at it in a while and there have been any code changes, I thought I'd mention it.

[Salvo GOP]

We understand your doubts  @elon, but there's nothing to worry about. We would like to reassure you that our Random Generator Number (RNG) got certified. Scaling tests and shuffling were conducted on enough samples to prove that cards/numbers are unpredictable, non-repeatable and uniformly distributed.

 

 

[elon]

As a VIP+ who doesn't buy chips (okay, I bought an inexpensive hat package when I first started, but it was October, leading up to Halloween), I don't doubt their claims of a fair RNG. I'll admit that I saw some SERIOUSLY improbable-seeming patterns my first month or two (e.g. consistent patterns of specific PP vs PP pairings), so I know where ppl are coming from, but what I've seen in the past year and a half in poker corresponds well with what I'd expect, when you adjust  for the high percentage of ppl who play long shots. (isn't that what hurts most? Ppl don't complain as much when beaten by a better hand that had them outranked from before the flop -- well SOME do, but they're SO CERTAIN that their (semi)bluffs are SO CLEVER that there must be rampant cheating. I LOVE playing against ppl like that)

I was making a specific technical suggestion for something that can happen if you apply a 52-card algorithm to a 6-deck boot. If you don't SPECIFICALLY KNOW what the modulo-, index-, and other bugs can arise in a Fisher-Yates when you change the number of cards, then my original post would be meaningless to you. If they say they've checked that, I believe them. It would be relatively easy to check/fix,

There's always one winner per hand, barring a tie. The rest lose -- i.e. 4-8 ppl lose every hand. There are MANY players facing you. If each is only "lucky" one hand in ten, you need to play sharp to stand a chance. The hands where you have ALMOST the "nuts" (the best hand) are the ones where you bet most and lost most for any tiny misjudgement. The odds are already so consistently in GOP's favor after rake that cheating would be 10x the effort for little reward. Those odds/rake are why casinos using physical cards instead of RNGs make money hand over fist. Are you such a great player that you know you'd win against eight ppl who will foolishly play every hand that looks good, even knowing that most of them will lose in the long run -- but not before stealing amny hands you expected to win. If so, play for real money instead of pretend chips.

I'm NOT great (though I played profitably for real money in the early days of online poker, when it was legit for USAns to play, I've studied enough to see gaping holes in my play). I'm only fairly proficient. I made a few million (sometimes small losses, some time bigger wins) week after week, slow but steady, and took over a year to reach 1 Billion chips, playing the mid-levels, and capitalizing on team bonuses. I would have been pounded flat if the numbers didn't work the way I expected them to -- the way statistics say they should. The same goes for most of my teammates.

 

 

[bcfdf]

You are lying again, you got certified for single deck shuffling, the post is about blackjack - 6 decks.

To say that is random when dealer gets 80-90% 17-21 cards, way to many 1 point advantage and the player busts almost every time dealt 12 or 13 it's laughable.

On 7/8/2019 at 12:00 PM, Salvo GOP said:

We understand your doubts  @elon, but there's nothing to worry about. We would like to reassure you that our Random Generator Number (RNG) got certified. Scaling tests and shuffling were conducted on enough samples to prove that cards/numbers are unpredictable, non-repeatable and uniformly distributed.