calculate openings

[stickman]

i cannot figure out how to calculate openings.. the wiki appears to be wrong

wiki formula (and what was said by jorb):

(attack weight/block weight)^(1/3)
X
base opening %
X
(1-current opening)

However when testing i found the following:
Person 1 has 100 unarmed and is using Chin up, they have 1 point in flex
person 2 has 400 melee and is using to arms, they have 1 point in full circle

if Person 1 uses flex against person 2 it should be (100/400)^(1/3) X 15 X (1-0) = 9.44
HOWEVER when i test this it opens for 13%

if person 2 uses full circle against person 1 it should be (400/100)^(1/3) X 15 X (1-0) = 23.8
HOWEVER when i test this it opens for 18%


Does anyone know how to currently calculate? also how the heck is manuvers factored into this with making combat stats the same... example:

if i have 1000 unarmed and my opponent has 500 unarmed and we both use chin up... we are the same... however if the 1000 unarmed guy uses combat med and flexs... normally his 1000 would be reduced to an effective 250... but under the new rules does it get changed to 250 vs 500 which is considered the same?

if that is the case the stronger opponent gets FREE maneuvers with no disadvantages... normally i would test this myself but i dont have the time and im getting frustrated so i was hoping someone else has tested this stuff already.

[boshaw]

This is going off various posts from J/L in the past and it should be fairly accurate still as it was in w11 when it was sqrt(sqrt(atk/block))..

Code: Select all

def -> def_{old} + (((1 - def_{old}) * opening * (attackweight/blockweight)^(1/3)))


where attackweight expands to

Code: Select all

ua/mc * attackweight * maneuverWeight * Mu(cards)


and the remaining `attackweight` is what's stated on the Card, usually 0.8 - 1.0 depending on the Card. maneuverWeight depends on if you're using like bloodlust which can do up to 4.0 or like oak which is 0.5, and so forth.

blockweight expands to

Code: Select all

ua/mc * weight * Mu(cards)


Where `weight` is based off what the card says. (ie: shield up -> 2.5, etc)

In all cases Mu/Mu(Cards) is

Code: Select all

(1 + ((cards - 1) / 4.0) * 0.5


As for w12, from what i've seen in testing iirc the equalizer is only applied to your ua/mc vs their ua/mc depending what you and them are using. Any additional attackweight or maneuverweight is applied afterwards. Hence you're equalized on ua/mc and not weight. Also if you don't fall within the 2x range your stat will still be 2x, just not equal. For example if they are 1500 ua/mc and you're 500 ua/mc you'll be effectively 1000 ua/mc vs their 1500 ua/mc.

In both of your examples it should be as followed (bold is the changes):

if person 2 uses full circle against person 1 it should be (400/200)^(1/3) X 0.15 X (1-0) = 0.188988157484231
HOWEVER when i test this it opens for 18% (loftar only sends integer last i checked, probably rounded down to client)


This one is interesting since something is missing slightly as it seems when the weaker opponent is attacking the stronger their initial hit weight is counting for more than it should. I've tried other stat examples where this is the case and the weight always comes out slightly higher than it should be. This doesn't seem to be the case when the stronger opponent attacks the weaker.
if Person 1 uses flex against person 2 it should be (200/400)^(1/3) X 0.15 X (1-0) = 0.119055078897615
HOWEVER when i test this it opens for 13%

Extra example of two people within the 2x range (say 200 ua and 400 ua with chin up using flex) => 0.15 expected and 15% is observed.

also two more showing off the oddity with weaker opponents getting some boost...
P1: 924 ua, P2: 10ua, left hook (0.15) from 0 def, chin up for both

P1 left hook P2: 0.538234977946919 expected, 53% observed
P2 left hook P1: 0.041803303244664 expected, 9% observed

So something is definitely missing from the formula that only helps weaker opponents.

[stickman]

( (10*X)/924 )^(1/3) * 15 = 9

( (10*X)/924 )^(1/3) = 0.6

(10*X)/924 = 0.216

10*X = 200

X = 20

seems like the weak 10 uarmed guy in your example is punching as if he has 200 unarmed.. a 20x increase

[boshaw]

( (10*X)/924 )^(1/3) * 15 = 9

( (10*X)/924 )^(1/3) = 0.6

(10*X)/924 = 0.216

10*X = 200

X = 20

seems like the weak 10 uarmed guy in your example is punching as if he has 200 unarmed.. a 20x increase

Right, but it's not constant. It's something that scales based off your UA/MC and their UA/MC.

[Glorthan]

Btw, full circle only has a 90% attack weight.