Implausibly bad luck

[wrldtrvllr]

If you hold two dice and reroll the other four, then you should expect about a 23% chance of not getting any dice that match the first two. And yet, literally this happens to me about nine times out of ten. Does anyone else experience this?

[Meeplelowda]

this happens to me about nine times out of ten

Provide the game IDs to back up this claim.

[qwertasdfg]

If you hold two dice and reroll the other four, then you should expect about a 23% chance of not getting any dice that match the first two. And yet, literally this happens to me about nine times out of ten. Does anyone else experience this?

48

[wrldtrvllr]

If you hold two dice and reroll the other four, then you should expect about a 23% chance of not getting any dice that match the first two. And yet, literally this happens to me about nine times out of ten. Does anyone else experience this?

48

My mistake, I didn't specify the situation fully - meant to say, when you hold two, and re-roll the other four dice twice, so 8 die rolls in total.

[bionic20]

If you hold two dice and reroll the other four, then you should expect about a 23% chance of not getting any dice that match the first two. And yet, literally this happens to me about nine times out of ten. Does anyone else experience this?

I understand your frustration, but im not sure this is a KoT issue; BGA's RNG algorithm is evidently flawed. Any experienced Can't Stop player can share with you the craziest dice sequences happening over and over

[Jellby]

If you lock two dice with two different faces, the probability that a single die roll does not match any of those is 4/6. The probability that none of 8 rolls matches is (4/6)^8 = 4%, so it should happen one out of 25 times. But that's just a probability. In practice, it can happen from 0 to 25 times out of 25.

[Meeplelowda]

If you lock two dice with two different faces, the probability that a single die roll does not match any of those is 4/6. The probability that none of 8 rolls matches is (4/6)^8 = 4%, so it should happen one out of 25 times. But that's just a probability. In practice, it can happen from 0 to 25 times out of 25.

The OP's probability is derived from holding two of the *same* faces, e.g., holding two 3s, then rolling the other 4 twice and getting no more 3s. So the OP's calculation is correct, but I suspect the 9 out of 10 claim is just an artifact of the cognitive bias that you remember improbable things more easily than common things.

This biased RNG claim keeps arising in so many games: Can't Stop and backgammon among others. The claims are almost universally based on anecdotal recounts of something "fishy" happening (see two posts above), not a systematic collection of data.

I once had a "feeling" in another online game (Wolvesville, f.k.a. Werewolf Online, if you must know) that certain squares were more likely to be attacked the first night than others. Over the course of a couple hundred games I collected and recorded (you cannot rely on your memory for these things) results and threw the data into a spreadsheet and used the chi square calculation function to see if my feeling was correct. It wasn't. The results were well within what you'd expect to see. In other words, I was unable to reject the null hypothesis that it was random and there was no bias.

The human brain wants to see patterns in everything (this is the basis of many optical illusions) and is horrible at accepting randomness.

[Jellby]

Right, (5/6)^8 = 23.25%. And I completely agree with the human inability to evaluate and accept randomness.

Besides, the fact that "something very unlikely happened to me" is not proof of anything (except that something very unlikely can happen), no matter how unlikely it is. Even if it's a one-in-a-million chance, if one million people try, and one of them gets it, he/she will come here reporting "it happened to me, something's fishy".

[voriki]

And I completely agree with the human inability to evaluate and accept randomness.

Yup.

For example we'd take something simple. Flip a coin, 50% odds heads, 50% odds tails.
What are the odds of flipping 2 heads after each other? Correct 25%.

I flip once, heads. So some people might now think they have 25% chance to throw heads again, and 75% to flip something else, thus tails?
Ofcourse that's not true, the odds 50-50 reset themselves after each coinflip, probability has no memory, it does not recall what previous results were. Every time, you start with a clean slate. And once in a while you do end up with a combination of flips or rolls that seem highly unlikely, but yet, they still happen and it is not abnormal.

[Meeplelowda]

And once in a while you do end up with a combination of flips or rolls that seem highly unlikely, but yet, they still happen and it is not abnormal.

And just as important, if unlikely things never happened, then that would be abnormal. Low probability is not the same thing as zero probability, yet people behave and interpret results as if they are the same thing.