From: Charles C. Allen (cca@newton.physics.purdue.edu)
Date: Sat 18 Aug 1990 - 09:56:33 EEST
Here's a short writeup of an alternate characteristic generation
method. Hope it gives some people ideas....
Charlie Allen ====================================== SIMPLE VERSION
In RQ2 and RQ3, the limit a characteristic may be trained to is determined by the starting value of the characteristic (or a related characteristic) or some racial maximum. Why not choose what a "maxed out" character will look like, just like you choose what it starts out like? The procedure is:
That's it! When you want to increase a characteristic, you are limited by the "ultimate" value. It works best if POW is increased like other characteristics, and not via the normal RQ "power gain" roll. Personally, I always thought that was a little too gross. Even the best roleplayers used to throw Disruption at trollkin so they get their chance at a POW gain.
This has the added benefit (in my opinion) of differentiating more between "fighters" and "magic-users". If a character has high ultimate STR and CON, he probably won't have high ultimate POW.
COMPLEX VERSION In the following discussion, only STR, CON, SIZ, INT, POW, and DEX (6 characteristics) are considered. APP is ignored (author's personal preference). The racial minima and maxima referred to can simply be the corresponding minimum and maximum values obtainable from the dice given in the monster description, or may be twiddled for playability.
Let L = racial minimum for a specific characteristic M = " maximum " " " " S = starting value for the characteristic C = current " " " " U = ultimate " " " "
Basically, when a character is created, the ultimate values determine the best that character will be (characteristic-wise). The current values start at the starting values (gosh, imagine that), and can increase up to the ultimate values through training. The current values are what get used in play for calculating bonuses, etc.
Define s = (S - L)/(M - L) c = (C - L)/(M - L) u = (U - L)/(M - L)
These numbers are all fractions between 0 and 1. The s's vary in [0,u], the c's in [s,u], and the u's in in [s,1]. The only exception is that the SIZ must always be the same. When determining starting characteristics, 6*0.4 = 2.4 points may be distributed among the s's. Similarly, 6*0.6 = 3.6 points may be distributed among the u's. Obviously, the 0.4 or 0.6 can be twiddled to what you think is reasonable.
The numbers used in play are the S's and C's and U's, which are obtained by inverting the above equations:
S = L + s*(M - L) C = L + c*(M - L) U = L + u*(M - L)
In "paper" games, S, C and U are integers for playability. The incremental cost for one point of a particular characteristic is
ds = du = dc = 1 / (M - L)
For humans, judicious choice of the limits, such as:
STR, CON, DEX [6,18] SIZ, INT, POW [8,20] M - L = 12 for all
results in the same cost for all characteristics, so they can just be swapped one for one. If the costs aren't all the same (ducks, for instance), then a spreadsheet can be used, or approximations for the costs calculated. Personally, the author has used the following for some of his RQ3 characters.
STR, CON [6,18] POW, DEX [6,21] SIZ, INT [8,18]
All characteristics except SIZ can be increased only by training. This includes INT and POW!
Training time should be based on the relative current value c and the incremental cost:
training time = c * dc * T (to go from C to C + 1)
where T is something like 10000 hours (that's a guess, this has not been playtested extensively).
Charlie Allen cca@newton.physics.purdue.edu
---
This archive was generated by hypermail 2.1.7 : Fri 10 Oct 2003 - 01:54:02 EEST