Dice Question

>>50044404
>"50% chance of +1" do you need to be about equal to "100% chance of +1"?
I'm no mathematician, but two 50's are 100

>>50044404
If you want a small dice game, you shouldn't be speaking in terms of percentages at all.

Make success a 50/50 and either reroll the result on failures or add up to +2 to a roll. Or both. If they can do a 5+ on either roll A or roll B, it's close enough to a sure thing it's practically a gimme but will never be a gimme.

>>50044448
This isn't a small dice game, but a mechanic I've been contemplating adding to my full RPG. Right now there is only advantage, but maybe bonus dice will be more interesting?

The reason I ask is because certain abilities grant either advantage (flat +1) or bonus dice (50% chance of +1), and in order to cost these abilities I need to find an equilibrium between the two.

If you have them, bonus dice are rolled alongside your regular dice. Each bonus die that rolls 4+ adds +1 to your total (the game uses only d6).

Therefore, I'm thinking the risk of two bonus dice is about equal to a safe +1; in this scenario you have an 75% chance of a +1 and a 25% chance of +2. You could get a +2, but you're risking your +1.

>>50044404
Infinite 50%s could never give a +1, theoretically.

One 50% is 50% hit, two 50%ers probably have a 75% to give a +1 (if either rolls +1), three probably does 87.5, 4 goes to 93.75, 5 goes to 96.88 (diminishing returns here), tested with rolling highest 1 of Nd2 on anydice.com

In a nutshell, 5 50%s are a +1 (96.88% chance is close enough to 100%) if your dice are up to d20 (d20 works in 5% increments, a +2 to hit is a +10% to hit).

As a fellow shitty homebrewer I'd advise against such ideas because it's not a small dice game idea.
You could get your players to gamble on it (instead of a +1 they get a 50% of +2 or 0).
This is either overpowered, letting them limit break depending on how your math works.

And also, it's an extra step to rolling a die and could bog the game down; if you're the gambling types that like risk assessment, go for it. Wouldn't recommend this for newbies.

>>50044404
Two. Look up the statistical term "expected value".