|
Maybe someone will find this useful. Did some tests on TR/Latency that i have been wondering about for some time. Sample size is very small but thats ok, i got a good feel on what works and not.
My conclusion is TR14/Low is playable. Noone uses this so its rendundant anyway. Next in line is TR16/Low which is used by majority of games i can find. TR20/Xhigh is also ok and probably should be the newest standard because TR20/Low is much the same as TR24/Low.
My measurements can be way off of course but how it felt cannot. Not much difference IMO between TR20 and TR24.
Results: https://pastebin.com/raw/7aKxTEFx
No tests done for 3v3 but ill get on it.
|
Happy Birthday MeSaber
|
Thanks! I personally dont celebrate anymore but a cake ill take
|
How big was sample, one game? Can u make legend for this data? Im not sure what i am reading there.
|
The samples are the amount of numbers shown then averaged out and result in seconds.
Example: TR24/Low was 14 samples and result averaged to 0.21 seconds (210ms).
The method was continuous rightclicks with a worker (from game start) and measure was from rightclick to the worker reacting, the time between samples was 1-3 secs for the worker to stop.
I also did it in rapid succession to see if it mattered if i kind of spammed, it didnt matter that much for my test so i kept the pause inbetween of 1-3 secs.
The worst sample i got was .99 (1 second) with TR8/Xhigh.
|
Posting data without any explanation of what that data means.. also I think that there is an official post (or by someone from the community?) about the exact TR to latency mappings.
|
I will read that if i find it LML. Thx
Edit: I thought it was kind of self-explanatory but guess i was wrong.
|
|
Thx ill read it. Cheers!
Edit: A lot of talk about DTR which i dont feel is working properly and should just be skipped. DTR overvaluates your latency and puts a worse TR than actually needed. I would guess it thinks the game is lagging/stuttering as we see it though it doesnt in our eyes.
Example TR24/Low i had no stuttering or issues in my tests, even so i got result of roughly 200ms anyway. While in the table in the blizzard post it shows 200ms should be matched with TR14/Low which i surely dont agree to.
But as i said my measurements can be way off. Could need a more precise method but that would require me to record my game and measure it by frames.
|
|
|
|