Favorite Dice Mechanics?

[jjordan]

What's your favorite dice mechanic? 1d20? 2d10? 3d6? 1d100? Dice pool?
Why is it your favorite?

I find I tend to prefer a dice mechanic which produces a bell-curve. When players are able to use bonuses to increase their probability of success this allows them to actually execute plans with some sense that they will actually be able to carry out the plan and not be 1 5% die roll away from failure at any given time. It does come with the need to constantly be doing addition, both of the dice and of the bonuses, and it does make exceptional successes harder to achieve. I think I'd prefer a 3d12 mechanic because it would give decent range, would make it fairly easy to add the dice, and would produce a decent bell-curve. But I also think most people would hate it in comparison to the other systems.

[Jason]

The big problem with bell-curve systems is that modifiers matter more in the middle than they do on the ends. If I'm rolling 2d6 and my target is 8 or more, a +2 modifier to my roll gives me about a 30% boost in my odds (from 41.67% to 72.22%). Pretty significant.
But if I'm rolling 2d6 and my target is 10, a +2 only gives me about a 25% boost in my odds (from 16.67% to 41.67%) .
And if my target is a 12 or more, a +2 only gives me about a 14% boost (from 2.78% to 16.67%). Not nearly as big a boost. You get diminishing returns from your modifiers as you reach the ends of the dice curve.

I still play Traveller, which uses this basic mechanic (2d6 + skill and attribute modifier, beat an average target of 8), but it is important to remember how the modifiers scale when you're running it or you'll end up thinking your odds are a lot better than they actually are. GURPS has the same problem but as it uses 3d6 instead of 2d6 the curve is wider and the effect is a little less prominent.

For D&D's flat d20 rolls, a +1 is always a 5% increase in the odds, until you get to the point where you'll only succeed on a 20 or fail on a 1 anyway.

Favorite dice mechanics?

I like the "roll and keep" system of the AEG versions of Legend of the Five Rings. It's an interesting mechanic. It lets you raise the target numbers in order to give yourself more spectacular results as well.
I liked the old Earthdawn rules where you rolled different dice based on your modified step number. That was fun, although you had to recalculate and look at a table for pretty much every roll. The exploding dice also meant that you had at least a small chance of succeeding at anything you could roll for.
I like The One Ring with it's quirky d12 + skill level times d6 mechanics. Basically you'll always have a 1 in 12 chance of success in that game, and the designers were very careful in choosing the target numbers for what happens when you don't roll the Gandalf rune on the d12 or worse, roll the Eye of Sauron.
I liked FASA's Doctor Who system, where you compared skill to resisting skill on a table and then rolled 3d6 against a target number. Everything in the universe fell into one of seven categories of skill or difficulty.

I find the old d% methods like in Call of Cthulhu or the Warhammer RPGs to be adequate, if unexciting. d20 works fine in the same way, though D&D3 got a little out of control with the difficulty numbers. D&D5 works much better that way.

The Shadowrun mechanic of "roll all of your d6s and then re-roll everything that rolled a 6, throw away the rest; rinse and repeat" was a little too time-intensive for my tastes.

I don't much care for systems where every die roll requires a table to decipher, like Rolemaster. FASA's Doctor Who does that too, but at least it's always the same table.

I don't generally like custom dice where no actual numbers involved, like FFG's Star Wars. They take entirely too much time to find the result and often come up with self-contradictory results.

[OldTrees1]

I really like:
Take 10. The ability to swap out a roll for a known quantity. Usually it is right below average, but even a Take 5 on a 1d20 would have merit.

While I generally use 1d20, I do like bell curves and inverse bell curves (nat 1s and nat 20s be the most common). Although bell curves can prevent mastery.

I also like exploding dice.

I wish I intuited the math behind dice pools better. Recently played a game with rolls like 14d{0,0,1,1,1,2}.

[Saint-Just]

I like bell curves very much too. The worst problem with GURPS 3d6 is that it's too low-resolution. But adding more dice makes the curve too curvy and they had "no unusual dice" in their mission statement. I wish GURPS was 3d12 or 3d20.

There are also systems which I like for no good reasons. Dice pools for example (oWoD or even weird poker of Cthulhutech). Though exploding dice are very bad IMO.

[Democratus]

I like bell-curve systems with a "twist".

Force Dice in WEG Star Wars. Stunt dice in Dragon Age RPG.

[HumanFighter]

d20 FTW
There's always something exciting about rolling a natural 20

[Ajustusdaniel]

Having spent most of my tabletop RPG career immersed in d20 systems, I find that the icosahedron has a certain totemic appeal that's completely unassociated with it's actual mechanical advantages or lack thereof.

For what it does, I greatly admire PbtA's fairly straightforward system of Roll 2d6 with minimal modifiers, with set ranges of numbers corresponding to a failure, a mitigated success and an outright success.

[Xervous]

SR4e dicepools. It’s a left leaning curve that stretches rather than shifts. Extra dice are always relevant but not always the most efficient investment to pursue. Degrees of success are pretty easy to handle and the Take 4 option (guarantee 1 hit for each whole 4 dice) gives you a decent mechanical cutoff for hand waving away trivial tasks.

[Guizonde]

d100's, no questions asked. i like the way it's so easy to read your percentage chance at a glance rather than multiplying by 5 on a d20. i like it combined with degrees of success or failure to see how badly or how well you did. when i came across the ffg warhammer rpg mechanics, i was in love. i learned the entire system in about 20 minutes, contrary to dnd 3.5 which took me over 2 years to really understand properly.

exploding dice are very fun too. i like making dm's cry in frustration with my stupid luck and rolling 5-6 times in a row. when i'm a dm, i like seeing the joy on the faces of my players when they do just that.

[Anonymouswizard]

Percentile dice, roll under, matches are special, higher is more extreme. It's arguably easier to grasp than any other system, even 1d[whatever]+modifiers, and most people have a good knowledge of their chances no matter if you start at 00 or 01 (which gives the system a couple of quirks but is not bad). Matches are easier to identify as criticals or advantages or whatever you're using them to trigger, and degrees of success become much easier to calculate.

Plus you can use the fact that both your tens and units range from 0-9 to fold effects calculations into the die roll, which my favourite system (Unknown Armies) uses to make attacks and damage a single roll. Sure you can fold that stuff in with other systems, but percentile dice gives you a little bit more variation on how.

EDIT: another reason higher is extreme is the superior way to do percentiles is that it makes cheating less rewarding. 'lowest first' or your trick dice will still get you more successes against the odds, but they're more likely to be bad successes. And I've seen people roll continually critical successes in Deathwatch before, no wonder Carnifex went down so fast there.

[Laserlight]

I like exploding dice. They're risky (when the villains get them) but when your player says "He's going to kill me when he gets an action anyway, so I'll make a reckless attack. I hit and do 2d6+2 damage for...explodes...again...another one...okay...keep going...heh....okay, that's 41 points of damage" and cleaves him in twain.

[Yora]

I really like the "dice pool with no counting" mechanic from Blades in the Dark.

Your rank in an ability determines the number of d6s that you roll. The die with the highest number is your result. All other dice are discarded.

I like the idea of your ability rank determining the number of d6 that you roll, but always found dice pools annoying because you quickly have to add together half a dozen or more dive for every single roll you make. In this system, you just have to sweep over your dice to see if you spot a 6. If not you sweep to see if you spot a 5. And so on. This actually works better with dice with dots instead of numbers, as most people have practiced recognizing the patterns of dots instantly as children for years.

For further simplification, the numbers mean the same thing for every action roll. 1-3 on your best dice means failure and you suffer a negative consequence. 4-5 means a partial success and you accomplish what you tried to do and also suffer a negative consequence. 6 means full success and you accomplish what you tried to do with no negative consequence. (The simplicity is a bit broken by getting a greater success if you have two or more 6s, but it's still really simple.)

[Duff]

I have fond memories of; d% roll under, with crits for really low numbers, and botches for really high numbers.
roll 1 d 10 twice. Once for your 10s and then your 1s.
and you roll "Oh" [pause with baited breath while you pick up the dice] "dear" as the 2nd 0 appears plunging your hopes of a crit into the despair of a botch

[Quertus]

Hmmm… I'll second "take 10" (and most anything else that obviates the need to roll dice at all).

I like exploding dice - they give players memorable moments / stories to tell.

I pretty well hate the WoD dice pool mechanics (although its own variant on exploding dice can be OK). However, take the same basic concept (attribute + ability, roll that many dice) and instead *sum them* against a target number, and allow the player to determine the effect of degrees of success (and, presumably, degrees of failure), and I love the idea.

Speaking of hating WoD, I have a love-hate relationship with systems which make it easy to tell that your GM is incompetent, like WoD or 5e D&D.

I like the simplicity of the math of d20 (no temporary modifiers, please - Persistence or go home), but dislike 2e THAC0 (and hatred early D&D attack roll table lookups). I like rolling on the pretty table on the back of the Marvel facerip books, but I hate the table for H&H (it feels dumb, like it should have just had "d20" level of "simple math" instead).

I think I've hated every 2d6 and 3d6 game I've ever played, *except* Battletech. I think it's because the mantra of "short range is 4, I ran is 6, your defense makes 8, one woods makes 9, pulse lasers makes 7, targeting computer makes 6" is part of the fun. (Or do war games not count? If they don't, Battletech *technically* has an RPG…)

Speaking of the Battletech RPG, I enjoy anything that manipulates probability / dice rolls: Battletech specialties (roll 3d6, drop one), Marvel facerip Probability Manipulation, Marvel facerip karma, ShadowRun good karma, M&M hero points, 2e D&D Moment, Warhammer Fate Points, etc etc etc.

I don't much care what die I'm rolling - I care what kind of stories I'm going to walk away with, and whether it feels like wasted effort getting there.

[zarionofarabel]

D100 roll under. Blackjack or Price is Right for opposed rolls. Crit at super low rolls and fumble at super high.

In other words, the system used in Mythras. Just sings for me, best roll mechanic ever IMHO!

If one combines say, hit roll, location roll, and damage roll, all in one go, making an attack roll a die pool of sorts. It makes for some quick and easily resolved combats.

[Mastikator]

TBH I feel like this is asking "what is your favorite tool, hammer or screwdriver?". Kinda depends on if we're using nails or screws.
For skill checks I like 2d10, but for initiative I like 1d20, for weapon damage, it should vary based on the weapon.

[Saintheart]

I like WEG's d6 system for no more highbrow reason than that there's nothing quite like being able to rack up about 10d6 and throw them all at once to determine how awesome you are at doing something.

[Khedrac]

TSR's Top Secret SI had a nice twist on the d100 percentage-based system.

Two of the main problems with the standard d% roll low system are:
1) Calculating criticals (10% of skill is easy, but many people have problems with 5% or 20%).
2) Opposed rolls - how to tell who does better (and although I have no issues, I know a lot of people who cannot understand Chaosium's Resistance table).

So, Top Secret SI used:
1. Skill/success chance is a percentage - roll under to succeed.
2. The higher the roll the better the result (unless other roll is a crit - then higher crit roll wins).
3. Crits are doubles (11, 22, 33, 44 etc.) that as successes - so a 55 is a very good crit for a skill of 56 and a failure for a skill of 54.

All the simplicity and elegance of the basic percentage system, but with critical and opposed rules that were trivial to apply.

[Anonymouswizard]

TSR's Top Secret SI had a nice twist on the d100 percentage-based system.

Two of the main problems with the standard d% roll low system are:
1) Calculating criticals (10% of skill is easy, but many people have problems with 5% or 20%).
2) Opposed rolls - how to tell who does better (and although I have no issues, I know a lot of people who cannot understand Chaosium's Resistance table).

So, Top Secret SI used:
1. Skill/success chance is a percentage - roll under to succeed.
2. The higher the roll the better the result (unless other roll is a crit - then higher crit roll wins).
3. Crits are doubles (11, 22, 33, 44 etc.) that as successes - so a 55 is a very good crit for a skill of 56 and a failure for a skill of 54.

All the simplicity and elegance of the basic percentage system, but with critical and opposed rules that were trivial to apply.

Yep, as I said higher up the thread this is my favourite system. I first encountered it in Unknown Armies and loved it (although UA separates critical results and matched results) and also love it when it's used in Eclipse Phase. It's so much easier to use in practice.

Going further into mechanics, I also love flip flops. This is where in a d% system certain abilities allow you to swap the digits on your side after you roll. Turning 62s into 26s (or other beneficial flip flops) has saved my character's life on more than one occasion.

[MoiMagnus]

I really like the "dice pool with no counting" mechanic from Blades in the Dark.

Your rank in an ability determines the number of d6s that you roll. The die with the highest number is your result. All other dice are discarded.

Systems that allow for multiple kind of dice and "roll all of them and take the highest" works quite well too.
(I only know obscure homebrews doing so, no official games)

[Ajustusdaniel]

Perhaps the most complicated dice system I've come across is Dogs in the Vinyard, in which the two sides of a conflict (usually, but not always, a player and the GM) roll a bunch of different dice (I think d4s through d10s) and then use the pools generated in a sort of push and pull poker game, possibly rolling more dice along the way.

[Jason]

TSR's Top Secret SI had a nice twist on the d100 percentage-based system.

Two of the main problems with the standard d% roll low system are:
1) Calculating criticals (10% of skill is easy, but many people have problems with 5% or 20%).
2) Opposed rolls - how to tell who does better (and although I have no issues, I know a lot of people who cannot understand Chaosium's Resistance table).

So, Top Secret SI used:
1. Skill/success chance is a percentage - roll under to succeed.
2. The higher the roll the better the result (unless other roll is a crit - then higher crit roll wins).
3. Crits are doubles (11, 22, 33, 44 etc.) that as successes - so a 55 is a very good crit for a skill of 56 and a failure for a skill of 54.

All the simplicity and elegance of the basic percentage system, but with critical and opposed rules that were trivial to apply.

Hand-to-hand combat was really very slick in Top Secret/S.I. You would make one percentage roll, trying to go under your skill rating. If you succeeded, the 10s digit told you how much damage you did, and the 1s digit told you the hit location. You could "bump" the hit location to another location by a number equal to your skill level (normally 0-5), so a hit to the right arm (4) could be moved to the chest (1) with a Brawling skill of 3, but not to the head (0). If you succeeded and rolled doubles then you had a crit and the hit location of your target was entirely filled in. Knock our their head or chest areas and they needed to make a Con check or fall unconscious.

One die roll was all you needed. Skilled hand-to-hand fighters could both do more damage and select a more damaging hit location, and they had more chance of scoring a crit - all in one roll.

[Khedrac]

Yep, as I said higher up the thread this is my favourite system. I first encountered it in Unknown Armies and loved it (although UA separates critical results and matched results) and also love it when it's used in Eclipse Phase. It's so much easier to use in practice.

Going further into mechanics, I also love flip flops. This is where in a d% system certain abilities allow you to swap the digits on your side after you roll. Turning 62s into 26s (or other beneficial flip flops) has saved my character's life on more than one occasion.

Apologies for missing your post. I like the flip-flops ideas - coudl be veryneat.

Hand-to-hand combat was really very slick in Top Secret/S.I. You would make one percentage roll, trying to go under your skill rating. If you succeeded, the 10s digit told you how much damage you did, and the 1s digit told you the hit location. You could "bump" the hit location to another location by a number equal to your skill level (normally 0-5), so a hit to the right arm (4) could be moved to the chest (1) with a Brawling skill of 3, but not to the head (0). If you succeeded and rolled doubles then you had a crit and the hit location of your target was entirely filled in. Knock our their head or chest areas and they needed to make a Con check or fall unconscious.

One die roll was all you needed. Skilled hand-to-hand fighters could both do more damage and select a more damaging hit location, and they had more chance of scoring a crit - all in one roll.

It has been so many years sicne I saw my copy of the rules (probably 30?) I had forgotten that until you mentioned it - yes - a really neat idea to speed gameplay.

[OldTrees1]

Going further into mechanics, I also love flip flops. This is where in a d% system certain abilities allow you to swap the digits on your side after you roll. Turning 62s into 26s (or other beneficial flip flops) has saved my character's life on more than one occasion.

Huh, flip flops. What if we assume bigger is always better. There are 55 results with this distribution (multiple of 11 are half as frequent because it can't flipflop):

Unique results
10,11,20,21,22,30,31,32,33,40,41,
42,43,44,50,51,52,53,54,55,60,61,
62,63,64,65,66,70,71,72,73,74,75,
76,77,80,81,82,83,84,85,86,87,88,
90,91,92,93,94,95,96,97,98,99,00

All results
10,10,11,20,20,21,21,22,30,30,
31,31,32,32,33,40,40,41,41,42,
42,43,43,44,50,50,51,51,52,52,
53,53,54,54,55,60,60,61,61,62,
62,63,63,64,64,65,65,66,70,70,
71,71,72,72,73,73,74,74,75,75,
76,76,77,80,80,81,81,82,82,83,
83,84,84,85,85,86,86,87,87,88,
90,90,91,91,92,92,93,93,94,94,
95,95,96,96,97,97,98,98,99,00

Comparing to d%,
Overall there is an increase. If it were not an increase, you would not swap.
The increase ranges from +0 (highest result is 00) to 25 (25th lowest result is 50).
As expected these increases are strongest in the low range, but they are not immediately strongest. The lowest result (10) only has an increase of +9.

[Minty]

Personally, I tend to prefer extremely simple linear dice mechanics with easily calculable odds. The current system I'm using is a home brew D10 + skill vs target number system, which is pretty much optimal in my opinion. No critical success or failure, no margin of success, just pass or fail. I also quite like D10 roll-under systems. D20 and D100 roll-under systems are OK as well, as is D20 + skill vs target, but I find D10 simpler and nicer. I find target number systems preferable to roll-under-skill type systems, as they handle contested rolls and difficulty modifiers better and scale well.

I've almost always played rules lite home-brew systems, but my favourite off-the-shelf systems are Warhammer 2e, Dark Conspiracy/Twilight 2000, and Cyberpunk 2020. In my experience, systems with dice adding (GURPS), dice counting (WoD), or excessive modifiers (D&D) just drag too much because at least half the players in my group are mathematically challenged.

I detest gimmicky systems.

[Pex]

My favorite is for a boardgame not an RPG. For RPGs I guess Take 10/Take 20 for when it's appropraite to tell the dice to stop interfering, but the boardgame Kingsburg uses my favorite dice mechanic. You roll the dice to give you worker placement options which in turn provide resources you use to do stuff. I find it an appealing method.

[Telok]

I prefer 3d6 vs. target numbers and dice pools mostly because those systems seem to have (usually) put more consideration into how often you should get particular results.

I detest single die / flat probability systems with two exceptions, the Call of Cthulhu d% with it's doubling/halving skill to represent hard/easy tasks, and one of the Paranoia editions that runs on a d20 using a blackjack & margin of success method that reduces anything you need to one roll.

I enjoy the Lt5R / DtD40k7e roll & keep but I had to write my own dice roller app to be willing to use it.

[lightningcat]

One addition to the Take 10/Take 20 idea I ran across for a high level game was the Take 1 rule. The action took less time than normal, but was 1 higher than your skill bonus. For character with +20 to +50 bonuses, this just made them look cool as they did small things with no effort.

[Tanarii]

Unless rolls are extremely infrequent, as simple as possible to execute. Generally speaking that means one die over multiple dice, and addition if a single number over multiple, and no subtraction, multiplication or division.

A complicated system is fine if you only make a check once an hour for a crucial action. It sucks if you're doing it them every 30 seconds for combat results, or worse multiple of them for a single action in combat.

[PhoenixPhyre]

Honestly, the only thing I care about with my dice mechanics is that they're simple and fast. NO TABLE LOOKUPS (at least for things that aren't once in a blue moon rolls). I want actually rolling the dice and interpreting the results to be basically brainless. Everything else pales in comparison. I don't look to my dice to provide entertainment--they're just there to help the "action" move along and resolve uncertainty. Flat vs bell? Meh. Statistics really don't matter in practice, not with the number of rolls you're making. What you roll (or don't roll) for is more important than what you roll IMO.

Although anything that lets you roll a big fistfull of dice occasionally is its own kind of fun. But only occasionally.

Unless rolls are extremely infrequent, as simple as possible to execute. Generally speaking that means one die over multiple dice, and addition if a single number over multiple, and no subtraction, multiplication or division.

A complicated system is fine if you only make a check once an hour for a crucial action. It sucks if you're doing it them every 30 seconds for combat results, or worse multiple of them for a single action in combat.

I see we had the same thought at almost the same time. I'm just more prolix than you are, so you won.