I have been noticing that the dice roller has been having issues for a week or so now. I'm in a lot of Play By Post games, running several myself in addition, and the dice roller has been rolling either very very low across the board, or giving nat 20s. There are few rolls between 10 and 19 as far as I can tell. If it happened occasionally, I'd be fine, that is just luck of the roll, but it is very consistent. I've seen multiple forums where we get 3+ natural 1 rolls in a row. Many groups have had conversations about trying a 3rd-party dice roller, due to the performance of this one. I've started letting some of my games roll with advantage always, just due to this issue, and have also seen some DMs who let everyone reroll ones for the same reason. it has gotten to the point that we expect low rolls, and are surprised when we get one that actually lets us have a chance to hit or pass the DC check. Low single digit results are the norm right now, with double digits only available if you have a very large bonus to the roll or in the very infrequent high rolls.
Then explain how one group gets 5 nat ones in a row? How one character gets 5 in three rolls in a row? I mean these things can happen, but they have been happening in every single game I am in, and I am in a lot of games. I have several DMs who are complaining about the dice roller. I do not know how the seeds are set up for the random number generator here, but I feel there needs to be some sort of change there. Yes, when you just roll dice after dice, you get a variety of rolls, but that is not what is happening. You have a variety of players rolling a variety of dice, and getting results that are very similar, low for the most part, with a few random high rolls.
Have you actually tracked all of the rolls to rule out any perception bias or run an analysis to see if the rolls are random?
People are terrible for seeing patterns where there aren't any and thinking random isn't random. I'm not saying that's what you're doing - but the roller is very random from my experience.
Hey, I‘m fairly new here and definitely have observational bias since I‘ve only been tracking myself. I haven‘t noticed any weirdness with the dice that aren‘t d20s, since I rarely roll those myself. I‘m definitely in one strange campaign where those 5 nat 1‘s happened, and another where I don‘t think my character has managed to roll under a 12 even once despite making checks almost constantly.
I‘m not saying it‘s necessarily happening here, since I‘ve only been playing for a month or so, but I can‘t help but think of early MS Excel code, which used to have issue with its own randomizer. While it did randomize after the first initial use, the very first instance of using the randomizer was very definitely NOT random at all. Maybe something like that is going on in the code?
The code here says the player rolled a 1d20 and added his bonus to it (+3). He rolled an 11, the code added +3, which resulted in a 14. If you hover over the number displayed in the post, it will tell you that much.
However, the number displayed in the post is a 5, and on top of that, it's a red number, as if he'd rolled a natural 1 and then a +4 was added to it instead. There's nothing in the code to clue me in to why the completely wrong number is being displayed in that post. So I would say, something somewhere is definitely wonky in the d20 dice code.
That is a definite bug in the dice roller and it's happened a few times to me in the Super Yahtzee Thread as well. Editing the post in any way (and not changing the roll at all) will result in the warning that the dice roll was cheated - but it will fix the roll.
Either way - it's the tooltip that is correct - not the written number.
Just to note, there's technically no such thing as random. "Random" is just a pattern of results beyond our capability to predict or calculate but in math terms there is no true random: the code is an attempt to simulate randomness but it's not "actual" randomness. A lot of time and effort is gone into trying to improve how codes approach simulating randomness. Some involve use of laser arrays.
The point is, no matter what method you use, there will be moments where it seems less than random - this is not a bug, just a limit of the coding language being unable to create perfect randomness due to perfect randomness being technically impossible (at least for dice rolling/random number generation).
Technically, rolling dice by hand is even less random than using digital dice rollers.
So while what Enkida describes is a bug, the pattern of consistent low rolls or high rolls or such is, indeed, less random than expected but not actually a bug and there's nothing the devs can do about it without spending huge amounts of money for a very tiny, almost imperceptible, improvement to how the code simulates the illusion of randomness.
Rollback Post to RevisionRollBack
Click ✨ HERE ✨ For My Youtube Videos featuring Guides, Tips & Tricks for using D&D Beyond. Need help with Homebrew? Check out ✨ thisFAQ/Guide thread ✨ by IamSposta.
The point is, no matter what method you use, there will be moments where it seems less than random - this is not a bug, just a limit of the coding language being unable to create perfect randomness due to perfect randomness being technically impossible (at least for dice rolling/random number generation).
I would say that a person is going to see patterns by sheer coincidence and because the human mind wants to see patterns in everything - long before they see a pattern because of how bad computer algorithms are at (not) making random numbers.
Modern RNG's are plenty good enough for rolling dice. It's for things like cryptography where it really matters.
The point is, no matter what method you use, there will be moments where it seems less than random - this is not a bug, just a limit of the coding language being unable to create perfect randomness due to perfect randomness being technically impossible (at least for dice rolling/random number generation).
I would say that a person is going to see patterns by sheer coincidence and because the human mind wants to see patterns in everything - long before they see a pattern because of how bad computer algorithms are at (not) making random numbers.
Modern RNG's are plenty good enough for rolling dice. It's for things like cryptography where it really matters.
I very much agree.
Rollback Post to RevisionRollBack
Click ✨ HERE ✨ For My Youtube Videos featuring Guides, Tips & Tricks for using D&D Beyond. Need help with Homebrew? Check out ✨ thisFAQ/Guide thread ✨ by IamSposta.
This is true but there are ways to get close to random. There are many ways you can generate random numbers using a computer. The best way to use random is to give a computer input that comes from a natural phenomena that humans can't predict nor control. One such generator is RANDOM.org, which uses atmospheric noise to draw numbers. It is probably one of the more popular algorithms for generating random integers.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Hello,
I have been noticing that the dice roller has been having issues for a week or so now. I'm in a lot of Play By Post games, running several myself in addition, and the dice roller has been rolling either very very low across the board, or giving nat 20s. There are few rolls between 10 and 19 as far as I can tell. If it happened occasionally, I'd be fine, that is just luck of the roll, but it is very consistent. I've seen multiple forums where we get 3+ natural 1 rolls in a row. Many groups have had conversations about trying a 3rd-party dice roller, due to the performance of this one. I've started letting some of my games roll with advantage always, just due to this issue, and have also seen some DMs who let everyone reroll ones for the same reason. it has gotten to the point that we expect low rolls, and are surprised when we get one that actually lets us have a chance to hit or pass the DC check. Low single digit results are the norm right now, with double digits only available if you have a very large bonus to the roll or in the very infrequent high rolls.
Hmm. Let me try.
4 18 5 16 19 6 16 16 12 11 18 18
12 3 19 19 19 1 11 1 11 11 8 10
20 18 2 19 17 10 1 4 17 14 9 1
3 2 15 1 11 5 7 1 17 1 14 8
Looks random to me.
Then explain how one group gets 5 nat ones in a row? How one character gets 5 in three rolls in a row? I mean these things can happen, but they have been happening in every single game I am in, and I am in a lot of games. I have several DMs who are complaining about the dice roller. I do not know how the seeds are set up for the random number generator here, but I feel there needs to be some sort of change there. Yes, when you just roll dice after dice, you get a variety of rolls, but that is not what is happening. You have a variety of players rolling a variety of dice, and getting results that are very similar, low for the most part, with a few random high rolls.
Have you actually tracked all of the rolls to rule out any perception bias or run an analysis to see if the rolls are random?
People are terrible for seeing patterns where there aren't any and thinking random isn't random. I'm not saying that's what you're doing - but the roller is very random from my experience.
Mega Yahtzee Thread:
Highest 41: brocker2001 (#11,285).
Yahtzee of 2's: Emmber (#36,161).
Lowest 9: JoeltheWalrus (#312), Emmber (#12,505) and Dertinus (#20,953).
Hey, I‘m fairly new here and definitely have observational bias since I‘ve only been tracking myself. I haven‘t noticed any weirdness with the dice that aren‘t d20s, since I rarely roll those myself. I‘m definitely in one strange campaign where those 5 nat 1‘s happened, and another where I don‘t think my character has managed to roll under a 12 even once despite making checks almost constantly.
I‘m not saying it‘s necessarily happening here, since I‘ve only been playing for a month or so, but I can‘t help but think of early MS Excel code, which used to have issue with its own randomizer. While it did randomize after the first initial use, the very first instance of using the randomizer was very definitely NOT random at all. Maybe something like that is going on in the code?
Here's some dicey weirdness I noticed in one campaign:
relevant post:
https://www.dndbeyond.com/forums/d-d-beyond-general/play-by-post/42585-above-below?comment=524
The code here says the player rolled a 1d20 and added his bonus to it (+3). He rolled an 11, the code added +3, which resulted in a 14. If you hover over the number displayed in the post, it will tell you that much.
However, the number displayed in the post is a 5, and on top of that, it's a red number, as if he'd rolled a natural 1 and then a +4 was added to it instead. There's nothing in the code to clue me in to why the completely wrong number is being displayed in that post. So I would say, something somewhere is definitely wonky in the d20 dice code.
That is a definite bug in the dice roller and it's happened a few times to me in the Super Yahtzee Thread as well. Editing the post in any way (and not changing the roll at all) will result in the warning that the dice roll was cheated - but it will fix the roll.
Either way - it's the tooltip that is correct - not the written number.
Mega Yahtzee Thread:
Highest 41: brocker2001 (#11,285).
Yahtzee of 2's: Emmber (#36,161).
Lowest 9: JoeltheWalrus (#312), Emmber (#12,505) and Dertinus (#20,953).
Just to note, there's technically no such thing as random. "Random" is just a pattern of results beyond our capability to predict or calculate but in math terms there is no true random: the code is an attempt to simulate randomness but it's not "actual" randomness. A lot of time and effort is gone into trying to improve how codes approach simulating randomness. Some involve use of laser arrays.
The point is, no matter what method you use, there will be moments where it seems less than random - this is not a bug, just a limit of the coding language being unable to create perfect randomness due to perfect randomness being technically impossible (at least for dice rolling/random number generation).
Technically, rolling dice by hand is even less random than using digital dice rollers.
So while what Enkida describes is a bug, the pattern of consistent low rolls or high rolls or such is, indeed, less random than expected but not actually a bug and there's nothing the devs can do about it without spending huge amounts of money for a very tiny, almost imperceptible, improvement to how the code simulates the illusion of randomness.
Click ✨ HERE ✨ For My Youtube Videos featuring Guides, Tips & Tricks for using D&D Beyond.
Need help with Homebrew? Check out ✨ this FAQ/Guide thread ✨ by IamSposta.
I would say that a person is going to see patterns by sheer coincidence and because the human mind wants to see patterns in everything - long before they see a pattern because of how bad computer algorithms are at (not) making random numbers.
Modern RNG's are plenty good enough for rolling dice. It's for things like cryptography where it really matters.
Mega Yahtzee Thread:
Highest 41: brocker2001 (#11,285).
Yahtzee of 2's: Emmber (#36,161).
Lowest 9: JoeltheWalrus (#312), Emmber (#12,505) and Dertinus (#20,953).
I very much agree.
Click ✨ HERE ✨ For My Youtube Videos featuring Guides, Tips & Tricks for using D&D Beyond.
Need help with Homebrew? Check out ✨ this FAQ/Guide thread ✨ by IamSposta.
Curious if a smaller die would show a pattern using the same concept as above:
111233436531
651665152256
145426351321
326625625262
That's a lot of red compared to green. Not saying it means anything, just curious. Another sample:
612162445132
465614131561
111662144416
343541415165
Still lots more red! For kicks:
521116441646
362422433216
454232622614
632434116432
Third time's the charm. Seems random enough to me.
I recently filmed my DnD beyond dice rolling the same numbers.
Any die 10 were 4's . As in i rolled 10 d 10 and all dice were 4's
Any 8's were 2's
20's were 4's
12's were 7s
I filmed for about 5 minutes going through different variations of dice and they all continued to roll exactly the same.
post us a link?
This is true but there are ways to get close to random. There are many ways you can generate random numbers using a computer. The best way to use random is to give a computer input that comes from a natural phenomena that humans can't predict nor control. One such generator is RANDOM.org, which uses atmospheric noise to draw numbers. It is probably one of the more popular algorithms for generating random integers.