Monstrous Compendium Vol. 1: Spelljammer Creatures
Frozen Sick
D&D Beyond Basic Rules
Basic Rules (2014)
Sign in to view your library & saved favorites.
Sign in- Marketplace
Monstrous Compendium Vol. 1: Spelljammer CreaturesFrozen SickD&D Beyond Basic RulesBasic Rules (2014)Sign in to view your library & saved favorites.
Sign in
I generally have had the feeling that D&D Beyond digital dice typically roll below average.
This evening I played a session in which I rolled 12 d20s and these were the results I got:
1, 1, 5, 1, 3, 10, 1, 7, 4, 3, 1, 4
The average of that is 3.42 (expected 10.5), sum of 41
5 of those are natural ones, which is extraordinary enough. (about 1 in 5,850)
The chance of rolling a 41 in 12 dice is roughly 1 in 30,000-50,000
It's possible I had an extraordinarily unlucky night. I also am not entirely sure that this can be confirmed by staff members (I can see a roll history to an extent, so I know it's tracked at least in cache or maybe its just appended to the html)
It might be worth investigating the driving forces behind dice rolls, but of course, as mentioned, it could just be extraordinarily bad luck. Nevertheless, given the incredible statistic, I think it's worth at least a look.
Did you try using more than 1 dice set? Your data set is not extensive enough to support that result if you don't have at least a few more digital dice sets to compare & contrast.
Also, be VERY careful with digital dice theorizing.
Try not to fall into digital dice conspiracy theories some some do here.
DM, player & homebrewer(Current homebrew project is an unofficial conversion of SBURB/SGRUB from Homestuck into DND 5e)
Once made Maxwell's Silver Hammer come down upon Strahd's head to make sure he was dead.
Always study & sharpen philosophical razors. They save a lot of trouble.
While those seem like imposing odds, consider how many people play on DDB every day. (No idea, but it's a lot. Almost certainly in the several thousands at a minimum.) If it didn't happen to somebody once in a while, that would be a sign of a problem with the die roller.
A fundamental thing about randomness is that it's spiky -- any given stream of random numbers, especially a short one, is likely to have some kind of statistical anomaly.
And we are very good at spotting patterns in noise. Very useful in a wide variety of situations, such as spotting a tiger in the long grass of the savannah, but it does lead to spotting patterns in randomness. It's why we see shapes in clouds.
You were particularly unlucky, and it was a very noticeable string of results. Had it been 6, 6, 11, 6, 9, 15, 6, 13, 9, 8, 6, 10, you're just as unlikely to roll such a median set, but you probably wouldn't have noticed.
They've explained the mechanics of the die roller in the past: It uses the browser's RNG to choose the initial orientation, and then launches the die into the physics model with a somewhat randomized vector. They also, IIRC, claim to have tested it, but even if the physics model is deterministic in some way, the initial randomization should be sufficient -- the Javascript pseudo-random number generator in every major browser is known to be adequate. (I believe everybody uses the Mersenne Twister, which is considered to be quite good.)
It is, I admit, not impossible they've changed something that broke it, but your sample is far too small to indicate that, and such a breakage would be affecting far more people than just you.
I also vaguely recall that, in the past, browser extensions that interact with DDB would sometimes break the roller, but I very easily could be misremembering.
I figured, at the least, if it were a widespread issue this would be a voice in the chorus. And yes, it is very difficult to draw a conclusion based on any number one person could feasibly put together. Nevertheless I rolled another 50 dice today and it completely disagreed with my earlier sentiment:
14,11,2,8,2,11,8,18,14,19,11,13,3,16,3,3,16,6,3,6,15,5,15,13,20,14,2,15,4,17,13,1,3,15,1,11,19,11,20,1,1,5,18,11,16,16,15,8,13,15
50 d20: sum 520 (expected 525), average 10.4 (expected 10.5)
I just happened to get a one in ten-thousands unlucky streak yesterday. In a table with a lot of players that only goes a few hours, you really only get to roll a handful of d20s and so things don't get to eventually skew towards to expected mean.
A conspirator's note: These weren't actually sending checks to a Virtual Tabletop via Beyond Roll20 - when I sit down next session, it's gonna feed me 1's again
One person can feasibly put that together. Its just time consuming, and requires good record keeping. This is just being a victim of a statistical fallacy, and then jumped to a conclusion without testing the hypothesis first. Something that happens a LOT with players between bounded accuracy actively squashing variables and limiting bonus stacking options. We're not concerned when we're performing above the odds; but get very suspicious if we think we're performing below average.
That said, my table still does the practice of dice jail.
Well of course; that's just good practice.
Feature request for DDB: Virtual dice jail: lock them up and switch to a different set. Every time you roll dice, the offending die is sitting there in a cage in the corner. People would pay for this.
I would like a Virtual Dice Cuck Chair.
https://www.tiktok.com/@run_dmg/video/7475097626397003038