Different dice in mechanics

Okay, here's the promised extension to the mechanics discussion. I don't have much time to write, so I'll keep it short. Let me just toss out a couple of the systems I know and we'll save evaluating them for later replies.

D20: Uses multiple sizes of dice to determine outcomes, most of which come into play only for hit point generation or damage rolls. Resolution rolls use a single D20 against a difficulty number with various + or - modifiers. Top and bottom 5% are critical successes or critical failures.

D6 (shadowrun) Uses a dice pool of six-sided dice consisting of (attribute+skill level) dice. Resolution rolls are fives or sixes on a roll of the entire dice pool. Substitutes success thresholds for difficulty numbers, and success thresholds don't modify. Modifiers are applied directly to the dice pool by adding or subtracting dice. Critical successes are confirmed with explosions on a six. Critical failures are confirmed by a lack of any successes whatsoever.

D6 (West End): Uses a dice pool of six-sided dice. Dice pools are equal to (attribute+skill modifier) in dice. Attributes and skill modifiers are bought up slowly rather than one whole die at a time (e.g., +1, then +2, then another die is added). A wild die determines critical successes (by exploding into another die) or critical failures (by rolling a 1 and then confirming the degree of failure with another roll).

I'd like to put White Wolf and GURPS up here as well, but I want someone with more experience GMing or playing those systems to break them down. Also, feel free to expand on what I have here. I know there are variants on systems, or aspects of mechanics that I've forgotten to list, so please help me out here.

Oh, I know what I forgot. The West End D6 system uses difficulty numbers that are directly modified based on difficulty (+ or -).

When I get a bit of time, I'll try to toss up AEG's 7th Sea/Five Rings unless you all beat me to it.

AEG's 7th Sea and Legend of the 5 Rings use a D10 based system that uses a total-based dice pool. In other words, your dice pool is calculated like a normal dice pool (attribute+skill modifier). However, of the dice you roll, you keep a quantity of dice equal to your attribute and add them up. The resolution check is against a normal difficulty number that directly fluctuates + or - based on modifiers. Critical successes are exploding dice in the dice pool, and critical failures are...well, I can't remember, but it's probably something typical (e.g., get straight ones on the whole dice pool).

Rolemaster/MERP: used d% rolls + modifiers against a lookup table. Most rolls in the system are open ended. If you roll really high (like natural 100), roll again and add it. It's been too many years, I don't remember what happens on a natural 01, but you probably had to subtract an additional roll. If you get another natural 100, repeat.

Not to stifle the discussion, but I'd love to have this sort of info on tag pages, I'd like them to be a nice database of info for folks to learn what all these wild terms mean.

No discussion of dice mechanics should be without a mention of the use of percentile dice in games such as RuneQuest, Call of Cthulhu, and James Bond 007.

The simplest version is the 'Basic Role Playing' one, in which a character has a precalculated chance of success (which may depend on character attributes and experience or design decisions). A character's percentage may be modified by adding a bonus or subtracting a penalty (as in most BRP games) or by multiplying or dividing (as in theRingworld RPG.

The most sophisticated and flexible version of percentile resolution was the one found in Victory Games' James Bond 007. In this 'Ease Factor–Quality Rating' system, each character's ability at a particular skill was rated as a 'primary chance of success', which depended on attributes and a skill level. And each task has an 'Ease Factor' assigned by the GM. You multiplied the PCS by the EF to determine a percentile chance of success. The you rolled D% against that chance. Chance of success could get well over 100% when a highly-skilful character attempted an easy task: this was not without advantages, because you got a better quality rating (faster or more complete success, more damage on a hit, etc.) if you rolled under one-half, one fifth, or one tenth of your success chance (tabulated on the character sheet).

Resolution systems based on percentile dice allow for fine adjustments of chances and rarer fumbles and critical than in systems based on a d20 roll such as Bushido or 3rd ed D&D. And being direct and linear they avoid the problems with central peaking that plague "fistful o' dice" systems such as d6 and the counter-intuitive funkiness of combinatorics such as makes White Wolf's dice-pool mechanic in my opinion an unusable mess.

-Brett

Shame on me for not discussing percentile dice. I was just playing a percentile dice based system, too. I'm trying to remember what it was. I think it was the Warhammer Fantasy Roleplaying System or some kind of Palladium game, but I could be smoking crack.

Rolling percentile dice is definitely the cadillac of dice systems for elegant simplicity and knowing the exact percentage chance of making the roll (without doing the math in your head). But for all its subtlety, I think it can become a bit obnoxious in terms of modifying difficulty numbers.

I can't agree that percentile dice systems make it inherently difficult to modify difficulty numbers. They have exactly the same strengths and weaknesses as any other system based on a linear probability distribution, such as a d20, except that they are capable of finer adjustments than most. And if we're talking Cadillac, the Cadillac of d% systems is the 'Ease Factor-Quality Rating' system from James Bond 007, which in my opinion had the best system for adjusting task difficulty and assessing the quality of success that I have ever seen in a commercial RPG.

Systems in which a variable number of dice are rolled and their results summed have in fact a much greater problem in adjusting task difficulty because the standard deviation of the result scales with the square root of the average result. Neither a fixed addition/subtraction to the target number nor a multiplication/division of the target number to adjust difficulties produces a suitable effect. It's only because the mathematics is too complicated to grasp intuitively that many gamers don't realise how ugly the results are. Outcomes under such systems typically excessively favour the characters with slight advantages.

As for systems in which a variable number of dice are rolled, and the results depend on how many come up the same, or how many match or exceed a target number, the maths of these is extremely obscure, and its results defy intuition. The fatal effect is that neither a GM setting a task difficulty nor a character-player whose character is attempting it is likely to have any idea whether the odds are 50:50 or 90:10. (And neither, I'm sure, do the game designers most of the time.) That's a recipe for nasty surprises.

I really prefer d% dice systems.

If you think about it, all other game systems tend to break down the 'odds' of rolling a 8+ on 2d6, etc.

Just skip the middle man and roll percentile.

I made an in-house rule system where: * low was always bad * high was always good * tasks were broken into 4 group (easy, average, difficult, challenging) with modifiers So, if the GM decides this particular task was easy (i.e. opening a combination lock (under stress)), then the chance of FAILURE is only 25%. The player must roll 25% or higher. * Cinematic option - failure is ALWAYS possible (roll a natural 0) and success is always possible (roll a natural 00).

I've got a collection of rules for different aspects of the game - combat, task resolution, etc. I am trying to put it all together, do some 3D artwork for the manual, and release it.

But I've been working on it for years. It's one of those tasks that may never get done.

Cheers