Homebrew general

>>54837259
You will likely have to make compromises at some point.

>>54838503
Well, I guess that's a given.

>>54837259
Use a 20 sided die
>linear probability
>each number represents a 5% increment
>modifiers remain small and easy to apply in your head
>no silly bell curve throwing off the scale (ie, making a +2 bonus better than a +1 in a nonlinear fashion)
>single die is cheaper than multiple (not everybody has an old yahtzee set lying around)
>gimicky dice mechanics won't distract from your well written rules and well imagined fluff

>>54841196
I gotta say, I dislike d20, it's got a broad set numbers, making the effect of small modifiers pretty unreliable.
Maybe a d10 though, hmm.

>>54837259
Lurk anydice and do rolls there.
2d6+mods will give you an average of around ~7+modifiers; and the modifiers in my humble (and probably retarded) opinion should never go above the half the possible max of a roll; so with 2d6 you should never have a modifier above 6 because it will trivialize most rolls.
http://anydice.com/program/c8a3

What's also important is what kinda math you using, is it roll under, is it regular D&D result vs. TN/AC, is it dungeonworld, dicepools in storyteller/shadowrun...
Read up on some system and steal their die system.

Also do tell what you aim for; what does easily eyeballed mean? Instantly see if a roll was good? d20's king here IMO.

>>54842090
Was going for some kind of roll against TN, or other rolls.
With eyeballing I meant, I want the players to be able to have a feel about their chances of success before rolling, without having to go into math.
Something dicepools for example lack a lot.

>>54841651
d20 offers a little more granularity than d10. d10 could be very good for a light, fast game.

And +1 is a pretty minor bonus, but not as minor as the nitty gritty percentage bonuses you see in some systems.

+2 is just as easy to work with, and can represent a more substantial bonus. Anybody can add +2 to a roll in their head almost automatically.

You can even do degrees of success--for example, by modifying results that are closer to the target number.

>>54842664
Single dice then my man, turning a dicepool into "hits, wins, successes" takes a bit of a while. Single dice are quick and simple, and are a great choice. And are also very friendly to people coming in from video games.

My homebrew operates on d6es, with 1 being -1win; 2,3 being 0wins, 4,5 being +1win, and a 6 being +2wins. This gives you an approximate "if you roll 8 d6es, you're gonna have about 4 wins", so rolling more dice is almost always gonna give you a better result and my players (playtesting guinea pigs) like it.
I'm using d6es because I'm in a third world country where a dice set is 10$ (and that's how much a day's pay is, let alone the fact that there are only 3 stores in the entire coutnry that sell dice sets), but 6d6 is about 50 cents.

I'm practically copying Shadowrun but having hits on results other than 5 or 6 (shadowrun can approximate hits to 1/4 of dice pool rolled).

Just steal a system from someone else (or at the very least do your own research as to why those games were designed to use XdY+Z).

>>54844472
I like any lightweight system that uses d6s.

File: unquestioned_authority.jpg (47KB, 400x195px) Image search: [Google]

47KB, 400x195px

2d6 will never be "easily eyeballed" because each side has a 16.67% chance of appearing and most people cannot easily multiply that number.

d10 is easy to eyeball because a lifetime of using a base 10 counting system teaches people to understand probabilities in increments of 10%.

2d6+mods is simple to use due to the ubiquity of PbtA games that use this mechanic, but it does not lend itself to making informed decisions based on mathematical probability; it merely produces an easy to interpret result.

That said, how "interesting" a core mechanic is has nothing to do with what shape of dice you roll - it is derived entirely from the rules surrounding how to interpret that die roll. For example, I would argue Freeform Universal has a more interesting core mechanic than D&D because regardless of the dice involved it has a broader variety of narrative outcomes.

And beyond that "interesting" itself is more a function of your target audience than an objective value.

>>54845087
I should add: multiple dice, regardless of shape, is too much of a confounding factor for most people to compute accurately.

File: 08-King-of-Hill-Hank_l[1].jpg (22KB, 400x300px) Image search: [Google]

22KB, 400x300px

>>54837259

I'd suggest using only one type of die and dealing with that.

>>54841196

Or use a d10 or d100 for better probability.

Hm, I think you have argued me to try just d10, instead of 2d6

Use a percentile die and a cube of LED lights.

The result of the percentile die can determine the base level of success.
Then you measure the angle of the shadow cast by the dice relative to the player that rolled, then multiply the result by that value divided by 360 and multiplied by 10.

>>54837259
>3d6/1d100 roll under.
The latter is just percentages. The former lets you use the dice themselves to scale difficulty (easy tasks are done with 2d6, difficult ones are done with 4d6, and so on).

>1-3d6, only the highest roll counts
So, say that all abilties are given a score from 1 to 3. That is the number of dice you get to roll when you use that ability, and you only check to see wether the highest of those dice is equal to or higher than the difficulty. Normal difficulty being, say, 4.

>color coded dice pools
Can be fun if you've got a huge collection of similar looking dice and ok with the idea of the GM (and possibly the players as well) improvising what the results mean. So, the player rolls a pool of, say, white dice to figure out if he succeeds or not, but may choose to add some red dice to the roll which also add to his success but every red dice added to a roll has some other negative impact on the character or story. And/or there may be a single black die which does nothing at all unless it lands on one particular face, in which case the action is an automatic failure or it has catastrophic consequences or the like.

>>54851608
What does the shadow measure?

>>54852191
Anal circumference.