I really enjoy the way that partial successes in Apocalypse World games continually complicate the fiction. Actions scenes get richer and more interesting as problems mount; the tension builds, and players feel like there’s more thing to react to. However, I’m not a fan of the “moves” system, so I’ve been thinking of ways to port partial successes over to more trad roll-under d20 games, like Whitehack, Symbaroum, or The Black Hack.

A common way to do this is just to say that missing by say 1-3 is a partial failure, while hitting by only 1-3 is a partial success, or something like that. The trouble is that you can end up with some odd quirks, such as a the fact that if you have an ability of 10, it’s more likely to get a full success or failure than it is to get partial ones. The way I see it, characters with average stats should be getting partial results much more often than either of the extremes, similar to Apocalypse World.

Here’s my solution: when making a test in a roll under-system, roll two dice instead of one.

- If 0 dice pass, the result is a full failure.
- If 1 die passes, the result is a partial success.
- If 2 dice pass, the result is a full success.

Here is what the probabilities of this system look like. The sweet spot is from 5-15, as you would expect. What’s interesting is comparing this to the probabilities for Apocalypse World. A 15 looks a lot like a +3, a 13 looks like a +2, and a 11 looks like a +1.

If you wanted more degrees of success than apocalypse world, you just add more dice. Here’s a system with 4 degrees of success: full fail, partial fail, partial success, and full success. With 4 degrees of success, the odds of a full success or failure goes down significantly, so you will be dealing with a lot of partial results. However, the split between bad results and good results will be symmetrical, as opposed to the 3-degrees system, where two out of three results allow the PC some degree of success. I suspect 4 degrees would work better for more realistic games, while 3 degrees would be better for more heroic games.

If you were using Whitehack or The Black Hack, you could still keep advantage and disadvantage by rolling an extra die and dropping the highest or lowest, respectively. Or you could just give players mods to their stat, which might be easier.

I’m planning on using a system like this for the project I’m working on now, The Wormwood Throne(or The Wormwood Crown…I haven’t decided) which will be more robust and detailed than Maze Rats, but the principle is applicable to lots of games.