Reaction looks slow from the spectator / game POV?

[babooz99]

When I watch my gameplay from someone else's perspective whether it be from another player or from the game's replay/demo, my reaction time looks off. It's like I'm delayed and my reaction time is slow as shit. Is this some kind of display lag? On my end it doesn't look like im slow, enemies are insta peeking me and im reacting as humanly fast as possible

[Chief Blur Buster]

When I watch my gameplay from someone else's perspective whether it be from another player or from the game's replay/demo, my reaction time looks off. It's like I'm delayed. Is this some kind of display lag? On my end it doesn't look like im slow, enemies are insta peeking me

In a replay, things can be temporally different.

During active realtime gaming, the enemy positions may be slightly out of sync between the local/remote system, due to internet lag fluctuations possibly, interacting with the game lag compensation systems / etc.

But when you playback a game from a replay system -- the game is displaying accurate positions for both you and the enemy. So your input-lagged and internet-lagged reactions actually becomes more human-visible during replay.

Now, this depends on what kind of game-replay engine they are using, but photontime-vs-photontime can be in sync in replay, unlike in live play. This is because when you're watching a replay, your local game engine is using already-reconciled timestamps during the replay, rendering locally yourself and the enemy at what the server thought was the actual timestamps of actual actions.

Replay isn't subject to internet lag, especially for replay engines that can replay your game while offline -- that requires knowing the timestamps of everything and all events, and it's all rendered locally -- with none of the realtime dissonances from server lags and internet lags (and lag fluctations).

This is normal for replay to look slightly different than live play, for Internet-based games where each separate player are a zoo of different latencies. So the server POV is necessary different from everbody's live POV, and the server has to lag-compensate everything the best it can.

Your hitreg weirdnesses (bullets hitting enemy that don't register, and bullets missing enemies that register) usually disappear for most replay engines in most games, when you playback a replay of a game -- because you're defacto playing back a recording of the server's POV.

The bigger the lag zoo effect, the worse the difference between live vs replay. (I say "lag zoo" to refer to the latency/jitter differences between all players).

For example, (1) LAN game, versus (2) a same-city everybody-is-FTTH game, versus (3) an international connections-zoo game where everyone is on LTE / 5G / cable / FTTH / DSL. A replay of item (1) will look far more accurate than a replay of item (3).

This stuff is up Battle(non)sense's alley. Watch his channel for some of these lag weirdnesses.

There are ways to try and mitigate some differences between live and replay (somewhat), but is outside the scope of this reply;

[blackmagic]

i have some different exprience but similar just in valorant...

when i spectate my teammates then i notice often and see that they all play with same delayed desynced movements and bad random broken hitreg similar to what i have...total off and tryharding...


dont know there hardware specs but pings are low same as mine 20ms.




i still believe that clients somehow has huge affect and effect on each other in an match...some are getting slowed down because of that even with stable pings and good fps...but everything feels wrong and broken....

then on other side some players are lucky and getting like an advantage because of that all bullshit and are like unaffected by other clients.



its still all an mystery for me.


cant understand how this possible and if my client has such huge affect on an match and makes it broken for everyone too in that match or if it has nothing to do with that and im just crazy...haha



i just need somehow to track down the players with who i played a match before and then follow and watch how there next matches will be going but just without me...and how they perform...

that would be rly interesting to see and to know for me.

[Chief Blur Buster]

i have some different exprience but similar just in valorant...

when i spectate my teammates then i notice often and see that they all play with same delayed desynced movements and bad random broken hitreg similar to what i have...total off and tryharding...

Just to be clear, there's two different things:
(A) Spectating a live game;
(B) Replay of a past game

Behaviors and mechanics will be different for (A) versus (B)

My previous post specifically refers to (B)

[blackmagic]

i have some different exprience but similar just in valorant...

when i spectate my teammates then i notice often and see that they all play with same delayed desynced movements and bad random broken hitreg similar to what i have...total off and tryharding...

Just to be clear, there's two different things:
(A) Spectating a live game;
(B) Replay of a past game

Behaviors and mechanics will be different for (A) versus (B)

My previous post specifically refers to (B)

can you explain more the differences ?


is live spectate more accurate than replays ?


i always believed that live spectate must be 99% accurate and shows to me exactly what other players experience in exact that moment and time.

or is it all too some delayed and i getting the spectate information with some lag time difference ? which is normal but still it must be accurate and show whats exactly going on and happened in that time and moment...


or is live spectate system too not to trust ?


this all topic so hard for me to understand...

[babooz99]

i have some different exprience but similar just in valorant...

when i spectate my teammates then i notice often and see that they all play with same delayed desynced movements and bad random broken hitreg similar to what i have...total off and tryharding...

Just to be clear, there's two different things:
(A) Spectating a live game;
(B) Replay of a past game

Behaviors and mechanics will be different for (A) versus (B)

My previous post specifically refers to (B)

can you explain more the differences ?


is live spectate more accurate than replays ?


i always believed that live spectate must be 99% accurate and shows to me exactly what other players experience in exact that moment and time.

or is it all too some delayed and i getting the spectate information with some lag time difference ? which is normal but still it must be accurate and show whats exactly going on and happened in that time and moment...


or is live spectate system too not to trust ?


this all topic so hard for me to understand...

I believe replay of the server gives the most accurate depiction of what actually happens. There are some players that have incredibly non-delayed replays, they look like they have superb inhuman reaction times when you watch their demos. Some of them are for example pro players.

[Chief Blur Buster]

this all topic so hard for me to understand...

This is outside the Blur Busters ballgame at the moment, as Blur Busters currently more focuses on display's lag rather than network lag mechanics. I did have Battle(non)sense as a guest writer:

www.blurbusters.com/network-lag/

There are scientific differences of live-spectating (unreconciled/guessed/extrapolated stuff going on) versus replay (final/reconciled stuff).

Remember, the server is running realtime, and all players are behind realtime because of Internet latency. So the live players / live spectators are seeing estimated/guessed positions -- because that other person is often a few milliseconds of Internet lag behind you. The estimated positioning is a necessary system or you will feel awful lag (e.g. 100-200ms lag). They designed it to play with low lag + server guessing enemy positions, rather than play with high lag + server having perfect enemy positions.

Imagine, an enemy moving in a straight line, but you and your enemy is 100 milliseconds lag apart. (E.g. you're 50ms from the server, the enemy is 50ms from server, so you're 100ms away from your enemy). But the server is running quicker than 100ms lag, because your move-forward is instant, doesn't lag 100ms behind your pressing the WASD keys. The server lets you move instantly -- instant localfeel.

But it takes 50ms for the server to know you moved forward, and another 50ms (100ms total) for the enemy to know you moved forward. Imagine the enemy is trying to aim at you -- but your local position is not fully synchronized with what the enemy sees your position as.

So you've got all the weird position-estimating and position-correcting going on, with the attendant hitreg weirdnesses. What you see on your computer is not always perfectly in sync with what you see on your enemy's screen.

Whether you are a spectator live, or whether you're playing live. The server is actually playing "time machine" by guessing you and your enemy's future position, and trying to synchronize as perfectly as possible in realtime, despite all the network latency.

It's law of physics. The server is trying to act like a time machine, because everybody has an Internet lag away from the server, and is not preventing anyone from executing immediate local movements. That creates a time paradox, that the server has to guesstimate/estimate/compensate -- creating all the crazy lag compensation effects.

Sometimes the algorithms are bad, and sometimes the algorithms are better, it's very hard to optimize for the lag zoo -- the users on FTTH playing against users in another country over a long laggy connection, as an example.

But once you playback a past historical recording, the timestamps are already reconciled, so playing back a past historical recording is time-accurate, because it's all actuals rather than guesses (weird hitreg because of server mis-guesses / mis-estimates).

This is kind of (more or less) as far as I go -- I exit my professional territory of displays. Obviously this territory overlaps (lag in general), but latency is a ferociously long and complex chain, of internal and external latencies, and my castle's drawbridge right before the local end Ethernet cable.

Now; to get to the next advanced level of explanation -- you need to ask someone for help to explain this better; subscribe to Battle(non)sense's YouTube Channel:

CHANNEL: https://www.youtube.com/channel/UCP7QY6 ... -stL-pNFrw

Lag Compensation Algortihms by Battle(non)Sense

phpBB [video]

[13n47]

I've made a topic on Valorant's netcode in the past. There's something funny going on with how the lag compensation works in that game. There's a way to tweak it by using third party tools but YMMV.

I roughly have 37ms ping to London and the enemy movement speed feels 150% the normal speed nearly every single match. Increasing my ping by 20ms or introducing a 10ms ping spike every second to my inbound latency will correct this lag compensation algorithm. The enemy movement slows down and makes tracking and hitting my shots much easier. I'm really sad that the developers haven't been able to fix this for such a long time even though I have given them this information.

Playing on Stockholm where I get roughly 15ms ping the shots hit faster and enemy movement is smoother but I feel getting prefired when peeking way more often than on other servers. With lower ping it feels like killing other players is slightly easier but you're also getting killed way more often. The gunplay feels just so wrong in this game

[rvteen]

There's a way to tweak it by using third party tools but YMMV.

please can you pm me the tool?

btw i haven seen that if i play with constant loss i tend to play much better and game feels in sync

[mago]

is it csgo? demo are 32ticks no ?