Does anyone know how perception affects ranged accuracy?
Offence rating is improved by increasing the specific stat, but how does this affect ranged weapons?
What is the difference between perception giving ranged accuracy and offence rating?
1 Like
perception increases accuracy in general(more helpful in ranged as lower chances of hitting)
offence rating, idk much about, i ll be honest. as far as i know it focuses more on accuracy for pvp. idk if it is even applicable to npcs.
2 Likes
I was looking into the formula for miss chance (offense/defense calculation) and I noticed it is only really working at higher values than what is currently available in the game. I think I balanced it around having 2000 defense vs 1500 offense for example. However at lower values it’s always hitting the bounds and resulting in a 95% hit chance basically always.
I’ll look into improving this formula and releasing it here so those that are math inclined can run some hypothetical calculations, but even if you don’t like math you should be able to intuit how it’s working just by playing too. You’ll also be able to see the role of perception at that point too.
3 Likes
if possible, can you actually give the formula(i am not going to find exploits with it but i am curious, also i might help you in improving it since i often need to work with people who deal in stats)
2 Likes
ofc analysis of stats is a bit different for us than game dev but i wanna try my hands on it anyways
2 Likes
\text{miss chance} = \max\left(0.05, \min\left(0.95, (\text{defense} - \text{offense}) \cdot 0.001\right)\right)
3 Likes
i have an easier and temporary solution with the 0.95 issue, you can simply add randomizer by replacing value of 0.05 with this: 0.01*time(in seconds/ms on the clients clock or a standard clock that differs from normal time zones).
3 Likes
although you will have to link value of time by a few lines of code
2 Likes
if you want to randomise in a standard clock, simply add 0.1*mod(sin(time on clock)) instead of the function i gave above.
3 Likes
Kaks was asking about the attribute formula too so here it is:
modifier = (attribute - 40) \times 0.01
3 Likes
you can use this for the input i think
1 Like
i read that paper again while i was having lunch today. first of all, the paper largely comes into effect after miss chance has been calculated. it will also be much more biased towards whatever final value the initial algorithm gives(in this case, the equation i have tagged). i havent analysed the equations fully but did conclude these 2 things after looking at the operations.
the matrix i was talking about earlier is generated from nothing but the bits of the numerical values provided by algorithms answer.the matrix undergoes a few operations and at last the result of those operations gives you a matrix used to generate a vector. the value of vector is what decides you missing or hitting. the closer the vector is to either of the vectors representing “hit” or “miss” decides it to be more precise.
3 Likes