Favored dice mechanic.

[aramis erak]

for single type, my preference is d6 roll low count successes against Skill - DifficultyMod. (Essentially, Arrowflight 1E).

For non-single type? dice pools by ability using customized dice and two-axis or 3-axis output.

L5R (Ring is functionally equivalent to attribute) Ring d6's & skill d12s, count your choice of 1 to ring dice.
d6: Blank, (opportunity+ Stress), (Opportunity), (success+ stress), (success), (explosive success + stress).
d12: 2× (Blank), 3× (Opportunity), 2× (Success+Stress), 2× (Success), 1× (Success+Opportunity), 1× (Explosive Success + Stress), 1× (Explosive Success).

I'm fond of T2K 4e... but wish they'd added d14 and d16 to give a six step instead of 4 step range. It's d(skill) + d(att), each read separately, with a d6/d8/d10/d12 range for attributes, and d0/d6/d8/d10/d12 for skills. Rolls of 1 being potential problem, 6-9 being 1 success, and 10+ being 2 success. Using the mode with two additional dice steps would need 14-16 as 3 successes. It doesn't require special marked dice, but those do add to the feel..

 

[zarionofarabel]

D100 or Percentile. Like in Mythras.

Thought Pendragons D20 version (5% increments) is also good.

Most other dice mechanics just seem either too complex or simply lacking in usefulness.

 

[Committed Hero]

I like the things that Delta Green does with the percentile system: the blackjack nature of the roll that allows for easy comparison, and the fact that doubles are a critical that can go either way.

 

[Snarf Zagyg]

I prefer 2d6.

A roll of 2, 3, or 12 is a fail.
A roll of 7 or 11 is a success.
If you roll anything else, you roll again. If you roll the same number before you roll a 7, it's a success with a complication. If you roll a 7, it's a failure with interesting effects.

 

[Celebrim]

Over the years I've become very much a proponent of linear systems with wide ranges, whether a D20 or a D100.

The fundamental problem with non-linear systems like dice pools, exploding dice, and so forth is not only is it difficult for a GM to manipulate and understand the math on the fly, but it also often becomes very clear after prolonged use that the designers themselves did not understand the math they created and instead relied on feelings of elegance in design that actually produced highly dysfunctional math in practice.

I feel like so many systems are inadvertently creating illusionism by hiding the math under complexity shielded from view by the fact that humans are bad at intuiting probabilities.

 

[Celebrim]

I prefer 2d6.

A roll of 2, 3, or 12 is a fail.
A roll of 7 or 11 is a success.
If you roll anything else, you roll again. If you roll the same number before you roll a 7, it's a success with a complication. If you roll a 7, it's a failure with interesting effects.

At least the probabilities in that system are well studied and readily available to anyone wanting to do the research. Though, you might find task resolution takes a surprisingly long time in many cases, and the drama of all that rolling may get tiresome if the stakes are low.

 

[Snarf Zagyg]

At least the probabilities in that system are well studied and readily available to anyone wanting to do the research. Though, you might find task resolution takes a surprisingly long time in many cases, and the drama of all that rolling may get tiresome if the stakes are low.

Well, I might have neglected to mention that I am big believer in playing TTRPGs for high stakes and cash money.

Laissez les bons temps rouler!

 

[Aldarc]

I'm not really a fan of the swingyness of the d20 roll, so I gravitate towards games that involve rolling two (e.g., PbtA, Ryuutama, Fabula Ultima, etc.) or more (e.g., Fate, Fantasy AGE, Cortex, FitD, Year Zero Engine, etc.) dice together. The added complexity of these systems tends to be overstated, especially given how often d20 systems involve d20 + Mod + Mod + Mod + Mod + Mod + etc. with additional bonuses that are often forgotten by the players.

 

[Committed Hero]

At least the probabilities in [craps] are well studied and readily available to anyone wanting to do the research. Though, you might find task resolution takes a surprisingly long time in many cases, and the drama of all that rolling may get tiresome if the stakes are low.

That would make an interesting timer mechanic.

 

[Sanguinarious]

I prefer 2d6.

A roll of 2, 3, or 12 is a fail.
A roll of 7 or 11 is a success.
If you roll anything else, you roll again. If you roll the same number before you roll a 7, it's a success with a complication. If you roll a 7, it's a failure with interesting effects.

I'd pass on that.