The Mushroom Press Design Blog

From the ground up

When I started tinkering with Legends of the Wulin the base character had a 7-die pool. You roll the pool, you get sets of matching dice, and you get an action per set.

Seven struck me as a lot of dice. Now that I’m working on a similar system, I realize the number must have come about after a lot of playtesting.

My own flailing attempts into the math behind the set-matching have proven… Beyond my mathematical understanding. 

But I have noted some undeniable trends:

-Sets are increasingly likely with higher die pools (duh)

-It is way more likely to get multiple sets of two than a single set of three or higher

-The likelihood of getting larger sets is really small, even with a big die pool (assuming 10 is the max)

So it makes sense to start with a big die pool like that. It puts a 3-die set on the table as a viable option. It makes rolling a set of 2, or even a couple of sets of 2 pretty good.

Even the NPCs never have less than a pool of 5. Probably for the same reason.

Coring down deeper produces weird results. What happens if you only have one die? How about two?

Clearly you can’t get sets from a single die. You also only get the “best” result of 10 one in ten times.

So, when you get two, your odds of getting a set of two are also 1/10. Twice your chances of rolling a natural 20 on a 2-sided die.

Three dice is where it starts to get complicated, and you start to see the real weirdness of the math. You go from 10 possible results (1-10 on a D10) to 100 results (100 unique combinations of 1-10 evolved from 2D10) to 1000 (You get the idea).

The strange thing though is that sets of two dice have this kind of lopsided lurch that doesn’t match with the pool of possible results as cleanly. If you spread out the 1000 results in front of you and count all the times that sets of 2 come up, then you get around 270 results (excluding the triple sets, which have ten).

So you go from 1/10 to about 1/3 (it’s a little less).

My ability to math out the results disintegrates above that. I’m not savvy enough with figures to set it up like a formula, so I have to write out the results in excel or by hand and just look at them.

But those results bear out in play. Eff it that’s trend enough.

That kind of system really brings out the best of Wuxia combat. Lots of actions, flurries of activity, etc.

So the paradigm is that it gives you

1) No answer (no sets)

2) One answer (one set)

3) Multiple answers (more than one set)

Which produces weird results when you ask it a binary question.

“Can I climb this wall?”

“How much can I lift?”

And stuff like that. It tells you inconsistent things from roll to roll. And what are you supposed to do with two answers? Take the higher set? What did the other set mean then?

"Can I climb this wall?"

-Not this time

-Yeah, with probably a 20-something result

-You can climb it, and you can do something else if you want

"How much can I lift?"

-Nothing this time

-About 20-something’s result worth, and rarely a little more

-You lift about 20-something’s result’s worth and do something else equally well at the same time

See its weird. People ask binary questions of an RPG system. Giving them answers like the above is awkward. This system does not do that well.

My proposed solution has been to allow people to just add the damn dice together. That way, more dice = higher result on average, with a little wiggle for the outliers. Way more consistent.

So there’s a dual resolution system. Adding the dice together (which in my estimation is the more intuitive) and sets (which are better, but less intuitive).

The dual resolution has the upswing in that it works with really low die pools. If you’re rolling one die, but 1-9 are on the difficulty scale, then you get an answer as to what you can do.

Adding the second die gives you more reliability, and it makes sets possible. So you’re significantly better than those 1-die chumps.

3 dice starts to really take advantage of all the options in the combat system. They have enough dice to attack and defend reliably.

You could conceivably start characters at low die pools like these and play them. Which opens up 6/10ths of the die pools that weren’t available in the older design.

I think that having the complication of a dual resolution is worth those results.