Heya,

I was wondering, I've never really looked into it, but I imagine there is a statistical significant difference between using a 1d100 and 2d10 for rolling a 0~100 number in this system. Both in real world and using computer based random generates for dice (rollers). I can't imagine they are really worlds apart in terms of the average they roll. But I do think they have to be different, because taking the average of a 1d10, twice, is quite different from a single 1d100 average.

I'm not looking for power play or using a rolling system to improve averages. Just looking into for the sake of interest.

Thoughts?

Very best,

Anydice.com might be able to do the math. Duno if it does percentile off hand tho.

Assuming you're using it as percentile, the odds of getting a particular number are (1d10)*(1d10). Since there's only one combination for each number, this is a 1:100 chance of a particular number turning up.

Assuming all dice are fair, Errant is right. I'm inclined to think that because no die is perfect there's probably more bias using 2d10, but that's not a calculation I feel like doing right now.

For example, how do you roll a 42 with a d100? You get that one exact side out of 100 possibilities, or a 1% chance. Rolling it with a 2d10 requires you to roll a "40" on one SPECIFIC 10, and then a 2 on the other 10. There's a 10% chance of doing the first roll, and then assuming you succeed you have another 10% chance of rolling the 2, and 10% of 10% is 1%, so we're okay.