FG2 v FG1 Die Rolling

[steev42]

Putting this in Tavern because I'm not sure if it's a bug or not.

My group has finally gotten tired of the die rolling algorithm in FG2. At least for the D20.

FG1 always seemed to work nicely; it was random, and the roll displayed always corresponded with the roll of the die.

This doesn't happen in FG2. We have an *overwhelming* preponderance of 1s and other low numbers; and about once out of every 5-6 rolls of a d20, the numbers don't match what's shown on the die. Note when I say overwhelming, I mean it. In our last session (6 players and 1 GM) we had 2 critical threats (one of which was on a 19) against at least 7 critical fumbles (which only happen on a 1). Not only that, but 6 1st level characters took 10 rounds to take out 4 goblins and a dog, due to an extreme number of low (2-5) rolls. Unfortunately, I wasn't the GM on this one, so I don't have the log to count up the numbers.


I'm mainly wondering if others have experienced the same shift since migrating to FG2? Did the die rolling algorithm change, or are we just experiencing a string of really bad luck for an overwhelming number of sessions? Any data recording I could do to confirm/deny problems with the roller?

[Dachannien]

You could try rolling a d20 about 200 or 300 times and plotting the results. Note that if there is something fishy with the PRNG, it's hard to tell what could affect it - rolling more than one d20 at a time, other players being connected, other players rolling dice in the interim, etc. If the PRNG is acting reasonably, though, none of this stuff should matter.

There are statistical tests you can do on the data to determine whether there is a statistically significant bias in the PRNG.

[kalmarjan]

Back in the day this was happening to me with FGI. I even posted a poll about it.

Things cleared up though, seems like a run of bad luck. ;)

Sandeman

[Griogre]

I have found the FG die rolling to be random over time - you may go through a few more lows and highs than actually rolling dice though.

I saw a study once that rolling dice by hand is actually a bit less random than a random number generator - primarily because of the way you can switch hands or switch dice (which actually had the same effect as shifting your grip), either which can have a small effect and interfere with true randomness.

My feeling is the dice in FG are far more random than some players would like. I feel pretty confident in saying if you went back and plotted the die rolls from your log you would not find a statistically significant divergence from random. Don't get me wrong, you might well find some runs of high or low numbers but not more than would be expected given the number of rolls.

On a less serious note – you can’t discount that some players/characters are just cursed when rolling dice. ;)

[Sorontar]

If I go through a bad run I change the colour of the dice and tell everyone I am changing my dice, I know it's sad but it makes me feel better :D

[Mellock]

If I go through a bad run I change the colour of the dice and tell everyone I am changing my dice, I know it's sad but it makes me feel better :D

Hehehehe... you're not the only one. :D

[Griogre]

LOL, I tell my players to change their die colors too when they are moaning about bad rolls. :D

[scytale2]

Yeah I'm having the same problem with the lottery....never the numbers I want...

[steev42]

You could try rolling a d20 about 200 or 300 times and plotting the results. Note that if there is something fishy with the PRNG, it's hard to tell what could affect it - rolling more than one d20 at a time, other players being connected, other players rolling dice in the interim, etc. If the PRNG is acting reasonably, though, none of this stuff should matter.

There are statistical tests you can do on the data to determine whether there is a statistically significant bias in the PRNG.


I may have to actually do that. And regarding manual v shortcuts -- It's a mix, at least on my part.