Board Thread:Dark Souls/@comment-6012540-20130122190436

I've been trying to figure out how attribute values are used to calculate the scaling damage for weapons. I've got a fair data set for the unmodified club and the damage seems to scale as floor(91*tanh((S-8.7)/26)) where S is the value of the strength attribute. It's not a perfect fit, but the float to int conversion applied to the data complicates regression analysis, (more data in the 20-35 range and 75+ range would help). The function seems to lie between hyperbolic tangent function and a-b*exp(-c*S) (where a, b and c are model parameters) though it is much closer to the former function. I'm posting the data set to see if anyone can do better. I've eliminated: a*ln(b(S-c), a*(S-b)^c, aS/(b+S), a*erf(-b*(S-c)), and a few series.

Using a data from a character with a 50 strength and a large supply of weapons that only scale with strength, it appears that the ratio of the damage of a weapon when two-handed to one-handed is roughly same (1.07) regardless of scaling rate (i.e. a,b,c,d,e or s). This suggests that the attribute dependence is independent of the scaling rate. Furthermore the scaling damage appears to be proportional to the base weapon damage (B), suggesting that the formula for the club should be floor(Z*B*tanh((S-8.7)/26)), where Z=1.05 is a paramater that is dependent on the scaling rate. Furthermore it appears that a,b,c,d,e and s don't refer to a constant parameter value, but rather a range of values within which Z would fall (i.e. if Z =1.05 or Z =1.1 the scaling rate would still be called A) .

If anyone knows the exact formula, could supply more data for the club (a similar data set for another solely strength scaling weapon), or could test the ratio of damage two-handed vs. one-handed for solely strength scaling weapons (to see if it is indeed constant across scaling rates) with a different base strength. It would be much appreciated. 