Testing in sandbox with mirrored Hollidays and cl_showpos turned on and 4x time slow for better visibility.
Bullet lands, damage happens, sound effect plays, old debuff of crackshot is removed if present.
Then some weird delay for whatever unknown reason, approximately 0.2s.
And then the slow is applied.
This, with the ~0.1s it takes for the movespeed slow to smoothly apply, accounts for the amount of missing slow that is visible when testing with normal speeds.
Math:
velocity (Hollidays movespeed) = 322.85U
crackshot T1 slow = -75%
velocity after slow:
EXPECTED = 322.85-(322.85-78.74)x(1-0.75)= 139.77U
ACTUAL 170U
duration (default) = 2s
rate of fade = (322.85-139.77)/duration = 183.08/2 = 91,54/s = 9,15/0,1s
appoximate weird delay = 0.2s
approximate movespeed rate smoothing = 0.1s
total time = 0.3s
calculated missing amount of slow = 0.3 x 9.15/0,1 = 27.45U
actual missing amount of slow = 170-139.77 = 30.23U
So it feels like:
crackshot goes on cooldown;
the slow value is calculated;
the duration timer starts counting and slow starts reducing;
weird delay of 0.2s;
then the reduced slow value is applied and then it takes the 0.1s to ramp to it.
Also tested it with superior duration and the actual velocity after slow is 165U which coincides with my theory that since the duration is longer, the ramp is slower, ergo the delay eats up less of the slow.
QED
Valve... Pls fix.
Bullet lands, damage happens, sound effect plays, old debuff of crackshot is removed if present.
Then some weird delay for whatever unknown reason, approximately 0.2s.
And then the slow is applied.
This, with the ~0.1s it takes for the movespeed slow to smoothly apply, accounts for the amount of missing slow that is visible when testing with normal speeds.
Math:
velocity (Hollidays movespeed) = 322.85U
crackshot T1 slow = -75%
velocity after slow:
EXPECTED = 322.85-(322.85-78.74)x(1-0.75)= 139.77U
ACTUAL 170U
duration (default) = 2s
rate of fade = (322.85-139.77)/duration = 183.08/2 = 91,54/s = 9,15/0,1s
appoximate weird delay = 0.2s
approximate movespeed rate smoothing = 0.1s
total time = 0.3s
calculated missing amount of slow = 0.3 x 9.15/0,1 = 27.45U
actual missing amount of slow = 170-139.77 = 30.23U
So it feels like:
crackshot goes on cooldown;
the slow value is calculated;
the duration timer starts counting and slow starts reducing;
weird delay of 0.2s;
then the reduced slow value is applied and then it takes the 0.1s to ramp to it.
Also tested it with superior duration and the actual velocity after slow is 165U which coincides with my theory that since the duration is longer, the ramp is slower, ergo the delay eats up less of the slow.
QED
Valve... Pls fix.
