Welcome to Gaia! ::

So we've been waffling back and forth on just what to do with this seemingly loopy dice server. Initially, everyone was getting 20s on seemingly every roll. That caused us to question the real randomness of our dice here on NPG and institute some changes to deal with the possible lack of a random dice roller.

I decided to do some investigating and thanks to everyone who posted in "Test Thread - GO!!!" for all the data we gathered. I compared every roll that's been made so far to some actual d20 rolling and came up with some interesting conclusions. Allow me to share my results.

For our purposes here, a roll of 10 or higher counted as a successful roll (because when you add in your modifiers, it will most likely put you above 15) and a roll below 10 counted as a failure. Based on a pool of 100 rolls:

When rolling 2d20, you will succeed a whopping 86% of the time. 9% of your successes will be 20s.
• 42% of the time, one dice failed and the other saved the roll.
• 44% of the time, both dice succeeded.
Conversely, you will only fail 14% of the time...but 8% of these failures will be 1s.

That information isn't too surprising when you figure that rolling 2d20 vastly increases your odds of success because you have 2 dice pulling for you instead of just one - if one fails, the other can still save you. But we have to wonder...do multiple dice affect one another's outcomes when rolling simultaneously? To answer this question, I rolled 20d20 five times (100 dice).

64% of these dice were successes. 6% were 20s.
36% were failures. Only 1% were 1s.

Though I can't jump to any conclusions, this leads me to personally believe that multiple dice DO in fact influence one another. At least when you roll a lot of them together. However, I don't think that rolling 2d20 creates any significant influence on one dice to another.

So what about rolling 1d20 at a time? Does this give us a more accurate portrait of the success/failure ratio? More importantly, does it give us any insight as to the dice servers bias towards success or failure? I'll let the numbers speak for themselves.

52% of these rolls were successes. 3% were 20s.
48% were failures. 4% were 1s.

This seems like a very balanced ratio considering that we split the d20 in half (10+ = success, 10- = failure). You would expect around a 50/50 split if the dice were truly random. That's exactly what we got with a slight bias towards success.

Actual physical d20s served as the control in our little experiment. We know these guys aren't tainted by any dice server - just good ol' fashioned luck. They were all rolled simultaneously in an enclosed cardboard surface, 10 at a time until 100 dice were rolled. Here are the results:

60% of the time, an actual d20 rolled a success. 4% of these rolls were 20s.
40% of the time, an actual d20 rolled a failure. 4% of these rolls were 1s.

True, it's not a 50/50 split, but it's pretty close...what more can our experiments uncover?

There's no question over whether real d20 influence each other when rolled simultaneously (how could they? It's not like they can intentionally work together). So the final piece in the puzzle is the control of the 2d20 success ratio. If our server is truly random and 2d20 aren't influenced by the multiple-dice success syndrome, then real dice will roll success/failure at about an 80/20 split.

82% of the time, the dice rolled a success. 7% of the time, the dice rolled 20. However...
• 57% of the time, only a single success was rolled on 2 dice.
• The other 25% of the time, both dice showed a success.
18% of the time, the dice showed abjected failure. 8% were 1s.

This is a far cry from the 42 single success/44 double success ratio given out by our infamous dice server. Where the dice server split the ratio down the middle, actual dice show (as one would expect) that a single dice saves the roll more often than both dice succeed.

All this information leads me to the following conclusion: Our dice server is random but only when rolling 1d20 at a time. I conclude this because:
• Rolls of 1d20 on the dice server very closely mirror rolls of actual single d20s.
• Rolls of multiple d20 on the dice server seem to influence one another towards success. This is proven by the fact that almost 50% of the time, both dice succeed when rolling 2d20...in actual dice rolling, this occurs about half as often.

So what does all this mean in terms of the rules? Well, it means that you guys were rolling bad numbers in NPG Academy because you're just plain unlucky. The dice server IS random when rolling single d20. With this in mind, we're going to go back to the ruling that all characters, by default, roll 1d20 for a feat.

Questions? Comments? Lamentations?

1d20 sounds good to me though. Are we still using the dare system?

We'll give it a go then, but what are you going to do about Stunts like Serpentine Strike?

So then we aren't going with the PCs/Bosses roll 2d20s and mobs and minions roll 1d20?

Okay then. Sounds good. *will never be able to use Dare system effectively*