What math skills do game designers need to know

[OgreBattle]

My ability to do math has rapidly degraded since college, and it wasn't really great to begin with.

But rolling dice and stuff requires a grasp of math to figure out the odds of things, and having good math helps greatly in figuring out when you have a problem.

Being unaware of it largely allows bad design to carry through with Mr.Cavern fudging everything behind the screen.

I can figure out D&D stuff fine as it's all +/- multiplication and division, but calculating die pools like shadowrun or frank's warp cult/asymetric threat leaves me at a loss.

What are other systems out there, and what's the basic math skills needed to figure them out?

[Cynic]

Simple probability is important. One of the things I usually think about when I play is the rng. If I can look at what my character does and see how they interact with the dc is important. So that's mostly arithmetic.

[kzt]

Statistics and probability are key. If the game designer doesn't fully grasp how likely a given roll or event is and what modifying that number changes in terms of the chances of success or failure it's pretty likely the entire game will be a goat rope.

[RiotGearEpsilon]

Being able to program up a quick monte-carlo simulation can do a lot to compensate for difficulty with probability.

[codeGlaze]

so, math wise, why bother with dice pools over more accessible pecentages?

[Hicks]

Because dice pools always have a chance (however small) of giving the lowest result regardless of their size, while additive totals will eventually stop giving the lowest result.

In shadowrun, no matter how skilled you are, you could always fail; but your chance of failure goes down the bigger your dicepool becomes.

In DnD, A d20+11 roll will always make any roll of DC 12 or less, there is no chance of ever failing a DC 10 task. The exception being the klunky autofailing on a roll of 1 rule, but even then your chance of failure will always be at least 5% regardless of how skilled you become or how many positive situational modifiers you accrue.

[Username17]

Hicks got it in one. With Dicepools, your variance goes up as your power does. This keeps characters feeling "human", because they can't actually just stand around picking their ass while enemies shoot automatic weapons at them. Sure, they are going to win, but they still have to try because low level mooks are never actually off the RNG.

On the flip side, it makes dicepools a rather weird fit for stuff like tanks, because the nature of dicepools keeps an M-1 Abrams "feeling human" which is just super super weird.

-Username17

[fectin]

One option might be straight-up adding successes to Abrams. Then they never "fail," but could still be overwhelmed.

[codeGlaze]

Those kinds of dice pools run off of successes, though, correct?

Additive dice pools would completely negate that 'humanizing' aspect, I presume.

[tussock]

D&D combat stuff is mostly a comparison of, uh, negative binomial or multinomial distributions. You getting enough hits to kill them before they get enough to kill you, with the death spiral effect of falling odds when your enemy can focus fire, complicated further by variable damage and contextual odds changes and ....

Which is far harder to calculate than most people think, but that's an advantage anyway, because it lets players who are mathematically very safe feel like they're not.

[name_here]

Game designers probably can get away with calculating odds for success on individual actions and letting playtesting tell them how combat works in practice.

[Stubbazubba]

Those kinds of dice pools run off of successes, though, correct?

Additive dice pools would completely negate that 'humanizing' aspect, I presume.

If you directly add successes (i.e. giving an Abrams tank a dicepool of 10 + 5 successes), then yes, that gets rid of the humanizing aspect and establishes clear minimums for their success. They can't get less than 5 successes.