Manageable Dice Pools?

The homebrew forum

Moderator: Moderators

Post Reply
User avatar
Hiram McDaniels
Knight
Posts: 393
Joined: Mon Sep 15, 2014 5:54 am

Manageable Dice Pools?

Post by Hiram McDaniels »

Lately I've been pondering success-counting dice pool systems like WoD and Shadowrun. Granted, I don't have a ton of experience with these systems at the table, because I find them cumbersome and slow, because they often involve what-the-fuckhuge handfuls of dice that need to be separated into hits, misses, double-misses and super-hits.

Particularly, I've been looking at the available material for Exalted 3, which I thought was going to be cleaned up and streamlined, but does not appear to have any better at all.

My understanding of how these systems work is thus:

*GM sets a target number, which is the minimum result each dice must show in order to be a "success" or "hit".

*Sometimes, a task requires more than one "hit" to succeed.

*Die pools are made up of dice from attributes, dice from skills, and dice from extraneous factors like environment, circumstances and SFX (powers, spells, disciplines, implants, etc.).

*It occurs to me that I could be making a mental goulash of Shadowrun and WoD.

Anyway, I had some untested ideas for streamlining a dice pool system, mostly in regards to limiting the sources that dice come from, and cutting the number of operations for things like combat resolution. From here on out, I'm going to be using Shadowrun terminology because I like it better.

Dice Pools

Dice pools are made up solely of attributes, which range from 1 - 5(ish). Skills actually do something else, which I'll get to in a minute.

I'm pondering whether to use traditional-type attributes (Strength, Dexterity, Buoyancy, etc.) or to use action-type attributes (Combat, Magic, Athletics, Subterfuge, etc.).

Threshold

Success or failure is determined by a Threshold that the GM assigns each task. Threshold is the minimum number of hits needed to achieve success. These range from 1 - 5, with 1 being an average task for an average person, and 5 being a task of epic proportions.

*No botches. A 1 is a miss, but does not subtract from successes.

*Dice showing 0/00 still count as two successes.

*Rolling over the threshold counts as a critical success and grants extra benefits, rolling just below the threshold is success with a cost, to reflect actual degrees of success.

Target Number

Generally, dice showing 6+ (d10) = Hit. GM never touches target numbers, these are player-facing statistics.

What skills do, instead of adding to the number of dice rolled, actually lower the target number for each die to be considered a hit. Skills are rated by ranks (novice, journeyman, expert, master), with each one decreasing the base target number by 1.

So Sneaky Pete the sneakthief has a master rank in the sneak skill, while Slow Charlie the Inadvertent Can Kicker has no rank. This means that when Charlie rolls a sneak check, any dice showing a 6+ is a hit, but when Pete rolls a sneak check, any dice showing 3+ is a hit.

Special Effects

SFX like spells and powers and such don't really add to dice pools. Instead, they either DO STUFF like discreet effects, or they grant automatic hits to certain rolls.

So Super Strength doesn't mean that you get a bunch more strength dice to roll, it just means that if you absolutely, positively have to punch down a wall, you already have a number of hits in the bank before even rolling.

I don't have a specific setting or game in mind for this yet, though I'm leaning toward Dresden Files-esque urban fantasy. Are there any games that already use some of these ideas, or is there a fatal flaw that I'm missing that makes this whole thing a complete nonstarter?
The most dangerous game is man. The most entertaining game is Broadway Puppy Ball. The most weird game is Esoteric Bear.
Blicero
Duke
Posts: 1131
Joined: Thu May 07, 2009 12:07 am

Post by Blicero »

If you have a dicepool system, and you have mechanics that change the number of hits required for success and mechanics that change the target number that determines what a hit is, you have probably created a system that is mathematically intractable for most people. It will be difficult for players and MCs alike to gain any intuition about what tasks count as easy, hard, etc. This is bad, and you should not do it.
Last edited by Blicero on Mon Feb 09, 2015 2:08 am, edited 2 times in total.
Out beyond the hull, mucoid strings of non-baryonic matter streamed past like Christ's blood in the firmament.
User avatar
Hiram McDaniels
Knight
Posts: 393
Joined: Mon Sep 15, 2014 5:54 am

Post by Hiram McDaniels »

You have a fair point.

Also, even with clear difficulty benchmarks laid out in the rules the probability curve for a 5 die pool is pretty unforgiving, to the point where if you're not maxed out, you can't compete.

Man, I just want small die pools. This is exactly why I prefer aggregate systems.
The most dangerous game is man. The most entertaining game is Broadway Puppy Ball. The most weird game is Esoteric Bear.
User avatar
tussock
Prince
Posts: 2937
Joined: Sat Nov 07, 2009 4:28 am
Location: Online
Contact:

Post by tussock »

The formula for standard deviation on dice is SQRT(n(x+1)(x-1)/12) for ndx. Game designers should know that, but they often don't.

If you want something like a d20, using dice with only two numbers (success and not), you need a pool of 133 dice with 50% success (133d2), averaging 66.5 successes (1d20+56). Same variance.

5d2 (as 0,1), like you suggest, may as well be 1d4, as you core resolution mechanic. Changing the rate of successes from 50% reduces that variation. To get the equivalent variation of moving up to 1d6, you need to use 12d2.

Aggregate dice don't even help that much. 3d6 matches 1d10+5 to within a few percent, and adding dice (like d6 Starwars) increases your average by more than the variation of your previous rolls very quickly. Going from 3d6 to 4d6 is like getting +7 in a d20 system, emboldened by the huge bonus.

Even using a d20, just +2 on heavily repeated rolls quickly overrides the per-die variance. For something like an original 4e skill challenge, you might face 15 checks, so the standard deviation per die is well under 2 points. Yet a hard challenge was +10 DC over an easy one. That just doesn't work for anyone but specialists. Not that the 4e design team understood any of that at all (see also, the 5e design team).

--

I'm of the opinion these days, the only reason to roll five dice is if you have at least six results you care to distinguish, including some extremely rare ones. Clearly categorised result steps are the best use of small dice pools (more fun than figuring out a stepped margin of success). But you probably only want 2-3 dice (for 3-4 results) and also use a big die type to allow some room for varying bonuses, and stacking your skill, attribute, gear, and situational numbers without exploding the game or being unable to distinguish things you care about.

So you can have a smaller die if your guns are pistol, assault rifle, and cannon, and your stats are normal, maximum, superhuman, than they are if you want to mechanically distinguish between an AR15 and an M60 by anything other than clip size. Noting that the dice effectively shrink as you check more of them to get a result, the bonuses on a d20 check made four times are worth twice what they are on a single roll.
PC, SJW, anti-fascist, not being a dick, or working on it, he/him.
Post Reply