Dice Tools in od&d

unified resolution systems are a big thing in modern rpg design. generally, the question i see a lot comes down to this: "why have multiple different [dice] tools for different things when you could have one system that handles everything." i used to feel this way, but since od&d has become my preferred system to a pretty strong degree, i've kind of changed my mind on it.

why use multiple tools when you could just have one system? because sometimes different tools are better suited to different jobs.

as i see it, od&d has three basic tools:

binary checks: roll 1d6 and compare to an x-in-6 chance to check if shit happens, otherwise shit doesn't happen. usually, the chance of shit happening is a 2-in-6. use this for things you check constantly, and only want to "work" sometimes (forcing doors, listening at doors, finding secret doors, wandering monsters). your stats don't really effect how good you are at putting your ear against a door.

spectrum checks: roll 2d6 and consult a table or matrix. use when you want to have a somewhat predictable result. used for morale, reaction rolls, and turn undead (morale+reaction rolls for corpses). 7 is the most common result (most people aren't interested in helping you right away), but there is room for modification (if they've heard about your reputation, they might like you more or less).

big stupid checks. roll a d20 and try to hit a target number. "high stakes dice." useful for when you want granularity without a fixed curve. allows you to to differentiate between chainmail (AC 14) and chain+shield (AC 15). Generally good for things where you want the success chance to be below 50% (hitting and passing saves), but with more variance then a binary check, and thus more ability to be modified as characters grow stronger.

to me, each of these tools serves a basic purpose which the others do not. you could use a d20 for all of these things, but it might suck. A 1-in-6 is approximately a 3-in-20, a 2-in-6 is approximately a 6 or 7-in-20.

spectrum-type checks are also interesting because of how they interact with modifiers and bonuses (such as those to morale and reactions). instead of adding a predictable +5% like on a d20, they essentially shift the distribution curve up or down, moving the goal posts rather than the range.

in my opinion, these tools are well assigned for what they are doing, and make a lot of sense in play.

of course, there are also times when a universal system would be preferable! immediately coming to mind are games where the skills of different player characters are emphasized (one of you is sneaky, one of you is charming, one of you is strong), and so it makes sense for all of these skills to work on the same math (such as the d20 system). also, games which emphasize story and character over specific exploration procedures typically benefit more from having a consistent way of resolving skill and chance.

basically: tailor dice to the game you want to make. different games want different dice.