keithcurtis said . It's not deciding in any manner. In 5e, you double the number of dice rolled, not the result of the dice. In PF2 it looks like this is an optional rule. FWIW, there is a statistical difference.* The reason more dice are rolled than are needed in most systems is that the Roll20 dice engine doesn't know the results before dice are rolled. It must roll all possible dice to cover all needs. A Roll Template has intelligence and will only display the dice results that are appropriate. This is what Vince was obliquely referencing. you posted a screen shot of what is causing an issue for you. Doubling the dice is less "swingy" than doubling the result. In 1d6*2, every result has an 18% likelihood of occurring. The curve is flat. In 2d6 you are far more likely to roll a 7 than a 12. In fact, 7 isn't even possible with result doubling. But I agree that doubling the result is simpler. It rolls two separate sets of dice, then doubles one. In the above example the first die rolled was a 1, the second die was a 2 which it doubled to give the four. The attack checks for a crit and will tell you if it is a crit, but it rolls both sets of dice every time. One set is just a normal roll and the other set is doubled for some reason. Again it's not an issue, just odd. It did cause a small discussion at the table as we tried to figure out why there were two separate sets of damage, but it was not an issue just an oddity that I wasn't sure if was a bug or odd coding choice. Seems like they should just use the same die roll for both calculations instead of rolling different sets of numbers for the same attack. EDIT: hmm not saving my pasted screen shot for some reason. EDIT2: got it.