Vlad_Konung, so what about the 5d20 dice? I was toying with R this evening and draw some diagrams to prove mine and
Galdred's point of view.
Using 5d20 instead of 1d100 would just confuse people. The CtH in percents that they will see in game wouldn't mean anything.
The actual chance of scoring 'p' points by 'n' throws of the 's'-sided die is:
This is a 1d100 diagram, X is the score, Y is the probability. Everything is equal, just as expected.
And here is the 'Chance-To-Hit' diagram. (Just a sum of probabilities of scoring a number lower or equal to X) X is the score, Y is the CtH.
It's linear, and you can see, that the actual CtH (Y axis) equals the 'CtH' displayed to the player (X axis).
Here is the 5d20 diagram:
Looks binomial-ish. Again, X is the score, Y is the probability.
And here is the 'Chance-to-hit' for the 5d20: (sum of the probabilities of scoring the number lower or equal to X)
As you can see, the ACTUAL CtH (Y axis) is not equal to the 'CtH' visible to player (X axis).
Players with 'CtH' of 30% in reality will have the CtH of 4,4% ! And players with 'CtH' of 75% will actually have a 99,36% CtH !
So, people with 'CtH' lower than 25% will almost never hit, and people with 'CtH' bigger than 75% will never miss. It's still random, even more random than 1d100, but now it's also confusing as hell. The only solution I can think of is to display ACTUAL CtH, calculated with the above formula, but it would still be random, just with different chances to hit.