This is an idea for a deck to replace die-rolling.
The deck consist of 36 cards, or any intergral multiple thereof.
Cards are the size of playing cards and can be printed 9 to a sheet, so 4 pieces of index stock produces a basic deck. The orientation of a card is important--where it differs, the values of the two sides will be separated by a slash.
A card has two pieces of information on it: the die roll, and the number to add to your karma instead if using karma. Optionally, a "0.5 x avdamage / avdamage / 1.5 x avdamage" field could be added, and actually skew things such that lower rolls do more damage
A deck consist of N times of each of these:
* dx3, Karma+7
* 5, karma+2
* 4/6, karma+7
* 6, karma+6
* two copies of 7, karma+3
* 7/8, karma+7
* three copies of 8, karma+4
* four copies of 9, karma+5
* four copies of 10, karma+6
* 10/11, karma+7
* four copies of 11, karma+8
* four copies of 12, karma+9
* three copies of 13, karma+10
* 13/14, karma+7
* two copies of 14, karma+11
* 15, karma+8
* 15/17, karma+7
* 16, karma+12
dx3 means roll 1d6 and multiply by 3, for equal chance of 3, 6, 9, 12, 15, and 18.
If using luck, draw 3 cards and pick the best. If you're doing five actions this turn, take the top 5 cards off your deck, those are your rolls. Every turn you take strain, draw a card. If you're playing 30 NPCs, draw the top 30 cards and run through them. There's far less adding, and no dice to go all over the place...you just have to shuffle a bit between turns/when you run out.
Optionally, if everyone likes the deck idea, some futzing with character sheets can result in a very eligant system whereby you add your skill to the value on the card and any modifiers to the roll and get your success or failure margin. Combine that with damage modification codes on the cards, and everyone needs a d6 for when they draw dx3 and to roll their karma when they do a karma ritual (or, heck, that could be added too...36 turns out to be a great number of cards).
For better randomness, all cards should be double-sided with different values on each side.
(C) Jesse Cox, 2002. At some point, I'll probably even do something with the idea.