[HOWTO] Fix "interpolate bug" // peaking advantage on CS:GO
J3di from zBlock wrote a very interesting thread on CSGO PRO private forum.
ATM only possibility to fix it is to put the cvar directly into the server launch command line, so basically most of people won't be able to access it but I've already asked Valve to allow us to edit it.
Here is a copy of J3di thread :
Cvar "sv_clockcorrection_msecs" defaults to "60"
So this is something I discussed a while ago, as spotting this in the SDK struck me as somewhat odd. The result was in VALVE not changing this.
I understand why there is a need for jitter correction like this, but I've never liked this implementation.
I feel it should be down to the client to decide how much they need, this basically performs the same feature as "cl_interp" but isn't being configurable.
All zBlock match-servers currently have this set to "30" and Verygames DM have been running with "0" on their servers for months now.
It's also presently set at 0 in a couple of competitive league configs.
If you were to remove this cvar (or set it to 0 in CSGO) I'd be happy if the result was more relaxed cl_interp restrictions, so that "cl_interp_ratio" could be between 1 and 3, instead of the current forced competitive values min 1 and max 2.
Reducing this value should have the largest impact on the game for competitive players.
For example I have a average jitter of between 2-10ms on my home Internet connection and a latency of about 15ms to many game servers.
There is no need for me to have the 60ms jitter correction, it's increasing position latency by about 4x! So I could easily get away without the jitter correction and rely on a tiny amount of extrapolation.
If I wanted a slightly smoother more accurate game I could increase cl_interp_ratio to 2.
I'd really appreciate it if this in particular could be reviewed."
We used sv_clockcorrection_msecs "0" (instead of default value 60) during ESWC and no-one complained about it (as expected).
I've also been using it ONLINE on CS:S on Team VeryGames DM server with same result.
So basically, if you want to test it, you need to ask you GSP to put :
+sv_clockcorrection_msecs 0 into the server launch option, it will basically remove peaking advantage
so atm (with def. settings) ingame you cant have ping <60 .. but what if you have ping 80? will it add +60 to this??
and will valve fix it? and full remove this?
in css is it cl_smooth ; cl_smoothtime?
in go atm its like - if u are waiting for opponent to come out, u are for example with awp - zoomed "on the corner" - he peaks, i know that i hit him in chest or head, but he doesnt get the bullet, then i die :) it looks like u shot too early. and this commands fix this :)