I've been tinkering with a new dice mechanic, but since I'm terrible at figuring probabilities, its hard for me to judge what kind of results I'll be generating.

Characters have 5 Attributes, rated from 1 to 6: Strength, Reflexes, Mind, Technology, and Presence. They also have skills, such as Shooting, Cooking, etc. Skills range from 1 to 6.

When attempting a task, roll a number of d6 equal to the appropriate attribute. A die that comes up a 5 or a 6 is a potential success.

Keep a number of potential success dice equal to a character's skill. A 5 counts as 1 success, and a 6 counts as 2 successes. Compare the total successes to a difficulty number (1 success needed for easy tasks, 3 for hard, 6 for heroic).

Example: Trog has a Strength of 5 and a "Smash Things with Rock" skill of 3. He hurls a rock at an innoffensive critter. He rolls 5 dice and gets 1,4,5,5,6,6. His skill of 3 allows him to keep three dice, so he retains a 5 and two 6s. The 5 counts as one success, and the two 6s as two successes each, for a total of 5 successes.

Any thoughts? Has this been done before? Is it too clumsy? I wanted to make competent characters good at what they do, but still allows weak characters a chance at achieving impressive results (hence double successes for each 6 rolled). I also wanted to keep the math low, and the number of rolls to a minimum. I used to play WEG's Star Wars game, and at higher levels, the rolling of buckets of dice and totalling them all up really halted the flow of the game. Anyway, hope I can get some feedback. Thanks for taking a look.

I think the mechanic looks good, but I am not a statician either so it could be off as far as odds go.

Okay, here's my assessment of the probability matrix. I may have gotten it wrong, but if so I'm sure someone will correct me.

[code] At least this many
Successes 1 2 3 4 5 6
Ability ---------------------------------
1 | 33% 0% 0% 0% 0% 0%
2 | 55% 11% 0% 0% 0% 0%
3 | 70% 26% 3.7% 0% 0% 0%
4 | 80% 41% 11% 1% 0% 0%
5 | 87% 54% 21% 4.5% 0.4% 0%
6 | 91% 65% 32% 10% 1.8% 0.1%[/code]
What I'm seeing is drastically limited utility for skills of four or higher. Even with an attribute of six, your odds of benefitting are not at all good.

How consistently would the same skills be married to the same attributes? Would "Smash with Rock" always be a Strength task, or would the caveman expect to use Smash with Technology to nap flint tools?

Welcome to the Forge LizardLips!

A couple of things. First, this type of thread is best posted in the Theory Forum. The Indie Design forum is reserved for actual projects in development. Is this for an actual project your working on? If not, don't worry, its not a big deal. Just remember it for next time.

Second, its kind of hard to evaluate a mechanic in isolation. Could you expand a little on your design goals?

That all said, the mechanic technically works. But I don't see the purpose in the "take the X number of highest rolls" bit. At least, not in conjunction with saying that only 5 & 6's are considered successes. All the "take highest" bit does is cut down the number of successes that people can roll. I mean, if the player rolls 553311 and can only take 3, what's the point? He only gets two good rolls. The only time the "take highest" bit matters is when a player rolls 555511, at which time they must discard one of the good rolls. I would understand it if you had something "take X number of highest dice and add the results together".

Also, if the goal is to "make competent characters good at what they do, but still allows weak characters a chance at achieving impressive results", I would suggest you think about an exploding die mechanic. If you're unfamiliar with the term, that means that when you roll a certain number - usually the highest or lowest number on the die - you get to reroll the die and add that extra result in. In this case, that would mean that whenever a 6 is rolled, the player gets to reroll the die/dice and add any extra successes they may get. That type of mechanic allows for freaky-lucky rolls where unskilled PC's get unbelievable results.

This system does, at least at first glance, present you with no chance of success when doing something for which you have no skill.
Is this intentional?

Roll stat, keep skill.
Hm... sounds familiar... ;)

The question becomes "What do I keep if I don't have the skill?"

The mechanic really de-emphesises skills (which I don't think is what you intended). There is never any reason to learn a skill to higher than the attribute it is normally used with. And statistally, there is little reason to extended your skill past 4 even if your attribute was 6.

For example, if your troll had six strength, then the difference between a skill of 5 and a skill of six will come in to play (1/3)^6 or 1 in every 729 times.

Another extreme, a character with a reflexes of 1 but a rifles skill of 6 would be routinely beaten in a sharp-shooting contest by a guy with reflexes of 2 and a rifle skill of 2.

Hmmm...

What about abilities that determine number of dice, but skills determine target number.

Perception of 4 would be 4 dice to roll.
Rifle skill of 3 would be he'd need a 3 or less to hit.

Max abilities out at 5.
Max skills out at 5.

Unskilled is a 1.
Amateur is a 2.
Skilled is a 3.
Professional is a 4.
Master is a 5.

Keep penalties simple. (- 1 to -3 range)
Haven't run the math on this.

Effect of success depends on what type of game you're developing. Storytelling, combat-oriented, player-driven, etc.