https://clips.twitch.tv/GlamorousGiftedSowTakeNRG
:/ it happens sometimes...
Dylan and I saw it though.

Just did a double check to see if that is the case and it is.

This really makes me wonder why valve hasn't deprecated the cl_interp command yet.

uhh that math on rate is wrong...

Rate is measured in bytes while Internet speed such as Dial up is measured in bits. So 56Kb dialup = ~7000 , the default rate is currently 80000 not 120000

interp_ratio 2.2 would be forced to 1 in RGL, UGC and ESEA, forced to 2 in Ozfortress, but work correctly in ETF2L.

Also interp_ratio 1.2 with interp 0.01818 would result in this lerp 21.8ms so about 6.6ms more delay that having the lowest possible.

interp_ratio OR interp 0.01818
Personally I don't see much of a reason for projectile based classes to have full proper interpolation since hit reg is server side and the few cases where more interpolation is required the added latency to both models and projectile firing would just make hitting a shot harder. It's nitpicking over some fringe scenarios and only by a few ms so largely it doesn't matter

Logitech wireless gaming mice all are super low latency as long as you have good signal strength and little wireless interference.

I haven't personally got to test any of the wireless mice from logitech (mostly because they're so damn expensive) but I have used a G403 which is the same shape and sensor as the G703 but wired. The only complaint I have with that mouse and a lot of ergo style gaming mice is their is not good place to put the ring finger or the pinky without one of the 2 being uncomfortable.

for reference this is the rough finger placement I have for the MX500 style of mice
https://i.imgur.com/6LaJ9xC.png
there's a ridge on the right side of the mouse that allows you to put your pinky a bit under where I have my ring finger at

It seems a lot of people don't use this kind of finger placement and thus most gaming mice don't have room for it. So this will likely be a non-issue for you. At $66 that's a pretty good deal. However if you have issues with the Wired G Pro assuming it wasn't defective or you didn't somehow receive a G203 (same shape worse sensor) you'll likely see issues with a G703 since it's the same sensor and also wireless.

I guess your issue could be related to the "surface tuning" feature in LGS which realistically all that does is change the LOD of the mouse. If you haven't tried that give it a go, if you already have go back to the factory default tuning level and see if the issue persists.

Double click/phantom click issue

This is a common issue amongst all mice that have the switches rated for 50 Million clicks. The reason for this is the lower actuation force for the switch while using more or less the same materials, the lower actuation force means the switch degrades a lot slower if using only that really low actuation force BUT this comes at a cost of a whole lot less tolerance of higher than necessary force. So if you click REALLY hard you're going to break a 50M switch a lot faster than a 20M switch. Technically the G403 uses 20M switches but they're omrons and have less tolerance the typical Huano switches found in Zowie mice, the FK1 itself has a design that requires yet more force than most other mice and eventually you get used to that actuation force and if you ever switch mice that don't have huanos and you still use that force you will break hardware a lot faster.

This can also be an issue with wireless mice, including the really good ones from Logitech or the Razer Mamba if subject to wireless interference.

If you're using a G Pro Wireless the issue is usually related to wireless signal strength and or interference. Place the receiver closer to your mouse/mousepad and make sure you move any routers that have 2.4GHz wireless at least 6ft away.

https://www.teamfortress.tv/50149/logitech-g900-g403-g305-wireless-mouse-jitter

If you're using the original wired G pro however (PMW3366), that's extremely strange and you likely got a defective model or a fake.

I'm just going to throw this out there for all the people recommending various net settings

Most league configs have values that limit what you can use

UGC server config

sv_client_max_interp_ratio "1"
sv_client_min_interp_ratio "1"
sv_maxrate "100000"
sv_maxupdaterate "66"
sv_maxcmdrate	"66"
sv_mincmdrate "50"
sv_minrate "1500"
sv_minupdaterate "50"

ETF2L server config

sv_client_max_interp_ratio 5
sv_client_min_interp_ratio 1
sv_maxcmdrate 66
sv_maxrate 0
sv_maxupdaterate 66
sv_mincmdrate 66
sv_minrate 20000
sv_minupdaterate 66

Ozfortress

sv_client_max_interp_ratio 2
sv_client_min_interp_ratio 1
sv_maxcmdrate 66
sv_maxrate 0
sv_maxupdaterate 66
sv_mincmdrate 66
sv_minrate 100000
sv_minupdaterate 66

Relatively certain esea forces cl_interp_ratio 1 but otherwise I'm not 100% aware of all the settings they enforce.

so from the knowledge we have the most consistent net settings you can use are

rate 100000
cl_interp_ratio 1
cl_cmdrate 66 
cl_updaterate 66

Everything else is more down to what class you're playing and how consistent your internet is to the server.

https://i.imgur.com/Jla6HTF.png

Currently feels like -30F, got a call at ~2:20AM for work where a cell tower lost internet and they wanted people to try to fix a fiber in complete darkness with a windchill of about -40F which coincidentally is also -40C. Fun fact wearing gloves meant for winter while trying to splice a fiber is basically impossible.

-edit-

Jeez I need to read your post entirely before I start replying.

A few questions

1. Which version of my config are you using?

2. Which version of mastercomfig did you install?

3. Did the mastercomfig you installed come as a .VPK file?

This maybe helpful just to get your game back to a state where you can play (without a config though) https://github.com/mastercoms/mastercomfig#clean-up

God it would be terrible to play against a DeepMind AI in any FPS. Especially with the way it learns. It could literally have the dumbest most predictable movement ever and it still wouldn't matter cause it would hit 90+% of its shots on hitscan (limiting factor being computational resources and Source hit registration). For DeepMind AI the most efficient strategies would likely be picking whatever classes do the best hitscan Deathmatch at medium to long range so Sniper, Spy and Scout. If multiple AI could learn to play as a team we might actually see a medic. Demo and Soldier? probably not tbh.

The AI would have to be kneecapped with smoothed out aim and delayed reactions for players to even begin to stand a chance.

Not using null-movement script is better but only marginally so.

1. You don't need a script to move like that.

2. You have the ability to stop basically instantly, which despite what some would say is a bonus in of itself.
Ex. If you're moving to the left (holding down A) and you want to stop really quickly just tap and hold D and quickly let go and press A back down. Otherwise you'll just decelerate and move a little bit in the direction you were already going.

3. You have the option of holding down opposite movement keys and letting go of one of them to move again. While not really that big of a difference for some people this may be more comfortable to do.

Realistically it's not that big of a difference. Depending on how you normally rocket jump using the null-movement script can really fuck you or slightly help you.

I can't wait to see some live casts of this...

Having the wrong input setting on the monitor wouldn't say anything about the resolution you use.

Most monitors have an option to select input via the OSD menu usually your options are a few or even all of these (VGA, DVI, HDMI, Display Port) as well as an "Auto" option that just cycles through all options until it finds an input.

I wish sandblast streamed this

https://clips.twitch.tv/TubularLightDunlinBCWarrior