Drawn Card RNG Bias for Specific Hands

[thefallinglink]

Just wanted to put together some observations that make me think the game rigs drawn cards for certain hands.

I think most of us have noticed some bias in the cards we draw. For example, on ante 1 the game is very generous with drawing the cards necessary for a full house or a flush. I'll often get multiple full houses with few or no discards, which stops happening pretty quickly after that.

Or when playing for flushes on a fairly split deck, several times I will have 4 cards of the same suit and burn through 3 or more discards of 5 cards each just to get 1 card of the other suit. Without any bias and 20% of the remaining cards being the target suit, the expected value for the number of draws without replacement is <5, or one full discard. The probability of drawing the target suit increases as we discard more cards, so needing to go through 15 cards (~25% of the deck) for a single card of the target suit seems completely rigged.

Here's some example math that isn't 100% correct because I'm not looking at combinatorics or draws without replacement, but the actual probability should be even less because without replacement the likelihood of drawing the target suit increases with each discard. If 20% of the remaining cards are the target suit, the odds that 5 drawn cards are not the target suit is < (0.8)^5 = 32.768%. Fair enough. But the odds of drawing 10 cards not of the target suit (two full discards) is < (0.8)^10 = 10.7%. For 15 cards (three full discards), that's < (0.8)^15 = 3.5% This should not happen so frequently without something going on in the backend.

Sure, it could be just confirmation bias and pattern-seeking on my end, but I don't think so. It's happened pretty frequently. Reminds me a bit of XCOM missing 95% shots more frequently than 5% of the time.

Another example, a recent run on this seed (TJ3I17MS) with the checkered deck I was playing a mixed flush/two-pair build. Two pair is one of the most common hands, with straights being a bit less common. Usually two pair is very easy to come by. Yet somehow on this run, almost every hand I drew was a straight. I would often discard 5 cards and not even draw a pair. It really seemed like the seed was rigged for a straights build, which would have been great if that's what I was playing.

Not necessarily saying this is a bad thing - it can be heavily biased in your favor, and it can keep things interesting. But it can kill some runs in ways that feel unfair.

WDYT, does the game skew the probability of drawing the cards needed for certain hands? Is it tied to the seed, ante, or what other factors?

[Goblin]

Cards drawn are indeed tied to the seed, and from what I can tell not in a way that biases certain hands. If you replay a seed you get the same cards drawn as the other time you ran it, even if you discard different cards.

[Ronald Brain]

Seeded run has no RNG involved, everything's determined from the start. You can replay the same seed over and over and get the same card order, shop purchases and so on.

THAT SAID, it is somewhat possible to manipulate card draw on your next round if you shuffle your hand during your previous blind or during a Tarot / Spectral pack when the game present 8 cards to you. By moving the cards you're dealt in a certain order, they may or may not appear in the next hand's initial deal (or appear within the next few discards). In unseeded run, this is completely up to luck, but in seeded run, you can simply mess around and find out what works in your favor, then reset and go for that particular setup.