Note to Self
Posted: Sun Apr 16, 2006 10:59 pm
XP_cum_min = 1/2*(2*(A+C)+B*L)*(L-1)
XP_to_next = A + C + B*L
Yeesh. They must have simplified the experience formula in RM2k3. The factors of A and C are the same thing, except C has a minimum of zero, not one like A does.
Max A, B, C factors are 2000/2000/1040 (A and C are exchangable)
Approximate formula for AP. (Small sample size - XP per encounter probably won't be linear later on) AP gained is linearly correlated with the sum of the XP of the encounter.
R^2 = .986, ANOVA p-value = 1E-9.
It mgiht be possible to use this formula to simplify the AP learning system a bit, by which I mean the AP required for each skill. Simply pick a level and calculate culm. XP for that level to ge the culm. AP for a skill of that estimated level. However, there might be a problem since the linear regression shows that the sum of XP is correlated to the sum of AP, and not individually. Not sure if it is a significant problem, though. Using this formula for every single encounter would defeat the point of the AP system, which was to allow skills grow independently from experience growth.
XP_to_next = A + C + B*L
Yeesh. They must have simplified the experience formula in RM2k3. The factors of A and C are the same thing, except C has a minimum of zero, not one like A does.
Max A, B, C factors are 2000/2000/1040 (A and C are exchangable)
Approximate formula for AP. (Small sample size - XP per encounter probably won't be linear later on) AP gained is linearly correlated with the sum of the XP of the encounter.
R^2 = .986, ANOVA p-value = 1E-9.
Code: Select all
f(x) #46;0062*x+0.863
if (f(x) > int(f(x)+.7)
A(x) = int(f(x))+1
else if (f(x) > int(f(x)+.4)
A(x) = int(f(x))+.5
else
A(x) = int(f(x))
