I've noticed recently, that the digital dice rolls aren't exactly... random. Like two weeks ago, I was attempting to roll stats for a character, and I went 30 minutes without rolling a single stat above a 10, and ended up having to just roll them by hand. Last night, in 11 rounds of combat with 3 attack rolls per turn, I rolled below a 10 on a d20 a total of 27 times. I realize that this is all up to chance, but the probability of rolling so low, so often, is not realistic.
I've noticed this issue as well. It's prevalent on the D&DBeyond character sheet and Avrae on discord. I imagine that they both use the same algorithm since they're both from Beyond and why not just reuse code you already have?
My last session I rolled 3 nat 1's and my highest roll was a 14 before bonuses. Absolutely unplayable.
Rollback Post to RevisionRollBack
Forever DM here. Seeking to make a decent campaign out of nothing.
I've noticed this issue as well. It's prevalent on the D&DBeyond character sheet and Avrae on discord. I imagine that they both use the same algorithm since they're both from Beyond and why not just reuse code you already have?
My last session I rolled 3 nat 1's and my highest roll was a 14 before bonuses. Absolutely unplayable.
They do not use the same algorithm, not even similar ones. Avrae is an online-only bot developed before it was acquired by D&D Beyond and uses a pRNG API that generates high confidence randomness. The D&D Beyond dice are offline based, using a physics engine based pRNG seeded using state-unique non-repeating variables (combining things like system time, CPU load, number of active processes etc).
If you're seeing similarities in behaviour, that would be indicative of confirmation/observer bias more than anything else.
"state-unique non-repeating variables" then says "(combining things like system time, CPU load, number of active processes etc)"all 3 of which are repeatable mechanisms and only one of them requires tampering to repeat, the other 2 do not these also the bases of most PC pseudo-random generation mechanism so your saying they are using a standard Operating systems Random.
"state-unique non-repeating variables" then says "(combining things like system time, CPU load, number of active processes etc)"all 3 of which are repeatable mechanisms and only one of them requires tampering to repeat, the other 2 do not these also the bases of most PC pseudo-random generation mechanism so your saying they are using a standard Operating systems Random.
All three are non-repeating variables; CPU load and active processes do not follow a set pattern nor do they repeat on any interval. And system time is an ever incrementing value measured from Jan 1st 1970. Yes, you can modify your system time but that will cause some issues.
Also, the list is non-exhaustive; those were but three examples and there are lots of other variables that are combined in a non-linear fashion to produce a unique initial state. And the point isn't to be tamper-proof, the point is to be reliably random.
I've noticed recently, that the digital dice rolls aren't exactly... random. Like two weeks ago, I was attempting to roll stats for a character, and I went 30 minutes without rolling a single stat above a 10, and ended up having to just roll them by hand. Last night, in 11 rounds of combat with 3 attack rolls per turn, I rolled below a 10 on a d20 a total of 27 times. I realize that this is all up to chance, but the probability of rolling so low, so often, is not realistic.
Im actually testing the randomness of various dnd dice rollers i'll let you know the results!
:)
I've noticed this issue as well. It's prevalent on the D&DBeyond character sheet and Avrae on discord. I imagine that they both use the same algorithm since they're both from Beyond and why not just reuse code you already have?
My last session I rolled 3 nat 1's and my highest roll was a 14 before bonuses. Absolutely unplayable.
Forever DM here. Seeking to make a decent campaign out of nothing.
They do not use the same algorithm, not even similar ones. Avrae is an online-only bot developed before it was acquired by D&D Beyond and uses a pRNG API that generates high confidence randomness. The D&D Beyond dice are offline based, using a physics engine based pRNG seeded using state-unique non-repeating variables (combining things like system time, CPU load, number of active processes etc).
If you're seeing similarities in behaviour, that would be indicative of confirmation/observer bias more than anything else.
Find my D&D Beyond articles here
"state-unique non-repeating variables" then says "(combining things like system time, CPU load, number of active processes etc)"all 3 of which are repeatable mechanisms and only one of them requires tampering to repeat, the other 2 do not these also the bases of most PC pseudo-random generation mechanism so your saying they are using a standard Operating systems Random.
All three are non-repeating variables; CPU load and active processes do not follow a set pattern nor do they repeat on any interval. And system time is an ever incrementing value measured from Jan 1st 1970. Yes, you can modify your system time but that will cause some issues.
Also, the list is non-exhaustive; those were but three examples and there are lots of other variables that are combined in a non-linear fashion to produce a unique initial state. And the point isn't to be tamper-proof, the point is to be reliably random.
Find my D&D Beyond articles here