|
Dreiko Shadrack -> The issue of the core assumptions for pet accuracy and damage. (3/6/2023 20:25:35)
|
The core assumptions for a player's turn state that the pet is always present, but CHA is not. This poses a problem as having the accuracy tied to CHA and a Mainstat makes it impossible for the pet to have the assumed 85% expected accuracy and 20% melee damage.
This issue does not affect guests as they are not part of the core assumptions for a player's turn.
Reasoning:
CHA*2/25 + DEX*2/25 + LUK/50
Is the current formula, we'll just assume it's tied to mainstat instead in 50:50 Ratio so:
250*2/25 + 250*2/25 + 250/50 = 45 bth
With the primary providing 18.75 bth as normal. They get another 38 as standard at level 150 so:
45 + 38 = 83
Then monster blocking is
FLOOR(150/4 + 55) = 92
Plus
Monster Stat Blocking = 225 * 1/50 = 4.5
92 + 4.5 = 96.5
Then:
(100 + 83 – 96.5 ) / 100 = 86.5
All is well, except you need both CHA and Mainstat to make that work. Without CHA, you lose 18.75 bth as mentioned, which comes out to 67.75% accuracy.
Assuming you're at 0 CHA your pet deals half damage, 20% melee, so you're dealing 67.75/86.5 = 78%~ = 15.6% melee.
If you take CHA out of the equation as per the core assumptions, the pet deals 78% of what it should. For clarity the 50:50 ratio doesn't matter since any amount of CHA at all makes pets weaker than they should be for 0 CHA users.
The easiest solution, in our estimation, is to simply remove CHA from the accuracy formula for pets and instead scale it purely on mainstat (or alternatively not having it scale with any stat to begin with).
EDIT:
A revised solution that isn't as easy but still fixes the issue while not penalizing Beastmaster builds but still penalizing 0-mainstat builds:
quote:
If (Max(STR/DEX/INT) / Vstat) < 1 {
Max(STR/DEX/INT) * 2/25 + CHA * 2/25 + LUK/50
}
Else
{
If (CHA / Vstat) < 1 {
Min(Max(STR/DEX/INT), Vstat) * 4/25 + LUK/50
}
Else
{
Cha * 4/25 + LUK/50
}
}
Explanation (as a level 150): If your highest mainstat isn't 250 then you use the 50/50 split formula; if your highest mainstat is 250 or above but your CHA isn't then use only mainstat for accuracy; if both your highest mainstat and CHA are 250 or above use only CHA for accuracy.
Furthermore under this current model while I did say "penalizing 0-mainstat builds" it's more accurate to say that those builds will not benefit from these proposed changes, absent the changes and following the current plans for CHA nothing would change for them from how things already are.
spoiler:
I still say that making it not scale from stats at all and putting the burden of fixing inequality on item support is the most accurate solution. :B
|
|
|
|
