valorant desync/network ONLY thread

[mel]

For me personally as of now it is obvious that its all ISP related and it means there is nothing i can do!

I play the same if not worse than a week ago but for the past 2 days people just forgot how to shoot and react in time

rn im colleting some data related to jitter/packet loss to compare it later

some of you might be getting the wrong data about jitter/packet loss, you should consider exitlag's network tester thing, its a really simple tool

[ChristophSmaul1337]

It seems the netcode is applying interpolation at the beginning of the match.

This likely happens in every game that uses interpolation. Assuming you're talking about cl_interp, the time interval used to "buffer" data packets before rendering them out in order to achieve smooth motion, is more than likely a fixed value in the vast majority of games that use this kind of netcode. VALORANT does it, CS definitely does it, Battlefield does it, CoD too. It's not set on a "per-round" basis, it's most likely always a fixed value, as long as the server tickrate isn't changing. For example, in CS:GO it used to default the cl_interp value to 0.03125, for a simple reason: Official servers ran on 64 tick and the default amount of data packets that were kept in the buffer for interpolation to do its thing was 2, and as cl_interp is calculated by dividing cl_interp_ratio over tickrate, it results in 2/64=0.03125. You could change the cl_interp_ratio to 1, giving the game one less data packet to work with, which would reduce cl_interp to 0.015625. That's what a lot of people used to reduce latency a bit.

With modern games, including CS2, VALORANT and Battlefield, interp time (cl_interp) is set server side, so you no longer have control over it. Keep that fact in mind, then think one step further: What would be the reason cl_interp is set wrong for you, but not for everyone else? What are the odds that it happens to you in every single game, but not to the opponents or some people you may be in Teamspeak/Voice call with? If there was an actual bug in the way cl_interp is set, it would appear for everyone.

it's also strange how changing ping can fix interpolation

It can't. Ping has nothing to do with interpolation. Read: https://developer.valvesoftware.com/wik ... rpolation) All the ping changes is the amount of time the server has to roll back the game state to check if you landed a shot or not. Jitter is what's actually bad, not ping. However, your jitter has to be pretty bad in order to mess with the game's netcode, to the point where you'd see warning icons within the game's diagnostic overlay.


Alright, now that we have that out of the way, I want to contribute to this thread since it's about VALORANT specifically. There was a situation I witnessed about 2 years or so ago. You probably know about stat tracker websites for the game. What I've done to test something I was curious about was pretty simple: I had 2 accounts at the time. As you likely know, even an "unranked" gamemode like Deathmatch has a hidden MMR/ELO (which, in my opinion, is disingenuous and the mode should be renamed to "ranked deathmatch"), and you can estimate your hidden ELO pretty accurately. Play a round of DM, go to the stat tracker of your liking and look at the ranks of players in the DM game. Now, when I did my test, I had two accounts: One of them had a hidden ELO around the Silver to Gold mark, as every player in my DM games was always either Gold or Silver ranked in the actual competitive mode. On my other account, there was a hidden ELO of Immortal/Radiant, as every player in DM games on that account was either high Immortal or Radiant.

Funnily enough, in the Silver/Gold lobby, I couldn't react to anybody. The only times I got kills was when I was ratting in a corner and holding offangles. As soon as I was peeked on a somewhat common angle, I was dead, mostly before I could see someone on my screen. I ended that game somewhere around the bottom of the scoreboard with less than 20 kills (yes, I was actually trying!). Keep in mind also that I'm a former 4800 ELO Faceit player (Top 0.1% worldwide) in CS:GO and I have a 130ms average reaction time.

Immediately after that game finished, I logged out of that account, switched over to my other account and played another DM against Immortals and Radiants. I was winning that game by more than 10 kills difference, nobody was shooting back at me at all and I was called out for cheating by a guy who turned out to be Radiant. This happened quite a bunch of times every time I tried it.

My friend who plays this game regularly recently observed something wild. He was in a 5 stack playing against what would later turn out to be another 5 stack. First match, they get rolled crazy, like 2-13 or something. What's important to note is that they weren't outplayed per se by crazy strats or util usage, but they simply couldn't shoot back because every death was the same as comonly described: Instantaneous deletion without the ability to react to anything. Next game they queue into the exact same 5 stack on the opposing team. Same players, same lobby, same server. Everything exactly the same, even the same map. You probably guessed it, opponents got annihilated 13-0. They even typed in chat how my friends were suddenly so good and gained godlike reflexes, while beforehand they would stand on angles for 500ms without realizing an opponent was there (from their perspective, of course).

The best part is when I was talking to someone who's not been playing PC games for too long. She's started tac-shooters with VALORANT, so she has no idea what an actual, real video game plays like. I was bursting out laughing when she asked me one day why VAL feels so inconsistent, that se can shit on people in one game, lose that game regardless, be on lower ELO in the next game but still be instantly deleted by players with lower skill levels than in the previous game.

I know someone who's regularly playing the game with a duo. He also was alerted to the fact that he would always instantaneously die to some random guy, spectate his duo mate and the same opponent that insta-destroyed him was missing all the shots on his duo. They decided one day to swap accounts, and would you know it... Suddenly roles got reversed and his duo was the one constantly being "pre-fire" insta deleted. His duo is also usually higher ranked than him, but in that "account-reversed" game, his duo got single digit kills while he was dropping a bombshell 30 kill game on them.

Ask around in your friends circle, I bet you'll hear similar things about the game.

I'm not saying this is conclusive evidence, but I strongly believe VALORANT has mechanisms in play that will manipulate your experience on the game depending on certain factors. I believe it's something similar to CoD's EOMM, and skill-based hitreg. Is everything above just random chance? Is it a coincidence that I just happened to be playing the game regularly at the time on the account that can't shoot back against Golds, and wasn't playing on the Immortal/Radiant account for a long time? Maybe.

I haven't played the game since that "DM incident" and I don't plan to do it, unless the devs actually address it, in whatever way they want. Either, they outright admit that they're using some form of EOMM, or they acknowledge that the netcode is fundamentally broken and detail strategies of what they're doing about it.

All I can say about VAL is this: Do not play that shitfest, do not take it seriously and do not interpret as much into it as you do. Try any other game(s) and see if you have similar experiences there, too. If the same happens for you in CS, BF, CoD, then you have grounds to stand on regarding possible problems with your ISP/Setup.

For me personally as of now it is obvious that its all ISP related and it means there is nothing i can do!

Which country/region are you from and what's your ISP?

[13n47]

It seems the netcode is applying interpolation at the beginning of the match.

This likely happens in every game that uses interpolation. Assuming you're talking about cl_interp, the time interval used to "buffer" data packets before rendering them out in order to achieve smooth motion, is more than likely a fixed value in the vast majority of games that use this kind of netcode. VALORANT does it, CS definitely does it, Battlefield does it, CoD too. It's not set on a "per-round" basis, it's most likely always a fixed value, as long as the server tickrate isn't changing. For example, in CS:GO it used to default the cl_interp value to 0.03125, for a simple reason: Official servers ran on 64 tick and the default amount of data packets that were kept in the buffer for interpolation to do its thing was 2, and as cl_interp is calculated by dividing cl_interp_ratio over tickrate, it results in 2/64=0.03125. You could change the cl_interp_ratio to 1, giving the game one less data packet to work with, which would reduce cl_interp to 0.015625. That's what a lot of people used to reduce latency a bit.

With modern games, including CS2, VALORANT and Battlefield, interp time (cl_interp) is set server side, so you no longer have control over it. Keep that fact in mind, then think one step further: What would be the reason cl_interp is set wrong for you, but not for everyone else? What are the odds that it happens to you in every single game, but not to the opponents or some people you may be in Teamspeak/Voice call with? If there was an actual bug in the way cl_interp is set, it would appear for everyone.

it's also strange how changing ping can fix interpolation

It can't. Ping has nothing to do with interpolation. Read: https://developer.valvesoftware.com/wik ... rpolation) All the ping changes is the amount of time the server has to roll back the game state to check if you landed a shot or not. Jitter is what's actually bad, not ping. However, your jitter has to be pretty bad in order to mess with the game's netcode, to the point where you'd see warning icons within the game's diagnostic overlay.


Alright, now that we have that out of the way, I want to contribute to this thread since it's about VALORANT specifically. There was a situation I witnessed about 2 years or so ago. You probably know about stat tracker websites for the game. What I've done to test something I was curious about was pretty simple: I had 2 accounts at the time. As you likely know, even an "unranked" gamemode like Deathmatch has a hidden MMR/ELO (which, in my opinion, is disingenuous and the mode should be renamed to "ranked deathmatch"), and you can estimate your hidden ELO pretty accurately. Play a round of DM, go to the stat tracker of your liking and look at the ranks of players in the DM game. Now, when I did my test, I had two accounts: One of them had a hidden ELO around the Silver to Gold mark, as every player in my DM games was always either Gold or Silver ranked in the actual competitive mode. On my other account, there was a hidden ELO of Immortal/Radiant, as every player in DM games on that account was either high Immortal or Radiant.

Funnily enough, in the Silver/Gold lobby, I couldn't react to anybody. The only times I got kills was when I was ratting in a corner and holding offangles. As soon as I was peeked on a somewhat common angle, I was dead, mostly before I could see someone on my screen. I ended that game somewhere around the bottom of the scoreboard with less than 20 kills (yes, I was actually trying!). Keep in mind also that I'm a former 4800 ELO Faceit player (Top 0.1% worldwide) in CS:GO and I have a 130ms average reaction time.

Immediately after that game finished, I logged out of that account, switched over to my other account and played another DM against Immortals and Radiants. I was winning that game by more than 10 kills difference, nobody was shooting back at me at all and I was called out for cheating by a guy who turned out to be Radiant. This happened quite a bunch of times every time I tried it.

My friend who plays this game regularly recently observed something wild. He was in a 5 stack playing against what would later turn out to be another 5 stack. First match, they get rolled crazy, like 2-13 or something. What's important to note is that they weren't outplayed per se by crazy strats or util usage, but they simply couldn't shoot back because every death was the same as comonly described: Instantaneous deletion without the ability to react to anything. Next game they queue into the exact same 5 stack on the opposing team. Same players, same lobby, same server. Everything exactly the same, even the same map. You probably guessed it, opponents got annihilated 13-0. They even typed in chat how my friends were suddenly so good and gained godlike reflexes, while beforehand they would stand on angles for 500ms without realizing an opponent was there (from their perspective, of course).

The best part is when I was talking to someone who's not been playing PC games for too long. She's started tac-shooters with VALORANT, so she has no idea what an actual, real video game plays like. I was bursting out laughing when she asked me one day why VAL feels so inconsistent, that se can shit on people in one game, lose that game regardless, be on lower ELO in the next game but still be instantly deleted by players with lower skill levels than in the previous game.

I know someone who's regularly playing the game with a duo. He also was alerted to the fact that he would always instantaneously die to some random guy, spectate his duo mate and the same opponent that insta-destroyed him was missing all the shots on his duo. They decided one day to swap accounts, and would you know it... Suddenly roles got reversed and his duo was the one constantly being "pre-fire" insta deleted. His duo is also usually higher ranked than him, but in that "account-reversed" game, his duo got single digit kills while he was dropping a bombshell 30 kill game on them.

Ask around in your friends circle, I bet you'll hear similar things about the game.

I'm not saying this is conclusive evidence, but I strongly believe VALORANT has mechanisms in play that will manipulate your experience on the game depending on certain factors. I believe it's something similar to CoD's EOMM, and skill-based hitreg. Is everything above just random chance? Is it a coincidence that I just happened to be playing the game regularly at the time on the account that can't shoot back against Golds, and wasn't playing on the Immortal/Radiant account for a long time? Maybe.

I haven't played the game since that "DM incident" and I don't plan to do it, unless the devs actually address it, in whatever way they want. Either, they outright admit that they're using some form of EOMM, or they acknowledge that the netcode is fundamentally broken and detail strategies of what they're doing about it.

I get what you’re saying about interp being fixed server-side, but here’s the part that doesn’t add up for me: if it really was always fixed and not affected by anything else, I shouldn’t be able to change how smooth enemy movement looks just by manipulating ping or creating controlled packet spikes. Yet I can. For example, if I throttle a small % of packets for a short window (like 10–30ms), suddenly sonic fast enemy animations normalize and I can track them way more consistently. That shouldn’t happen if it’s purely a fixed server setting.

To me, that points to something else in the pipeline — either how lag compensation is applied in borderline cases, or how interpolation reacts to uneven packet pacing. Ping itself isn’t supposed to “fix” anything, but in practice, raising my ping, spiking my ping, changing network buffering setting or forcing small bursts of loss can smooth out gameplay. If it were just placebo, it wouldn’t be that repeatable.

So while I agree that interp is technically locked, I don’t think it tells the whole story. There’s clearly some sensitivity in Valorant’s netcode to packet pacing or timing irregularities, which explains why people sometimes get better hitreg on a higher ping route or by using tools to shape traffic.

I do not believe in the EOMM theory.

[drosku]

It seems the netcode is applying interpolation at the beginning of the match.

This likely happens in every game that uses interpolation. Assuming you're talking about cl_interp, the time interval used to "buffer" data packets before rendering them out in order to achieve smooth motion, is more than likely a fixed value in the vast majority of games that use this kind of netcode. VALORANT does it, CS definitely does it, Battlefield does it, CoD too. It's not set on a "per-round" basis, it's most likely always a fixed value, as long as the server tickrate isn't changing. For example, in CS:GO it used to default the cl_interp value to 0.03125, for a simple reason: Official servers ran on 64 tick and the default amount of data packets that were kept in the buffer for interpolation to do its thing was 2, and as cl_interp is calculated by dividing cl_interp_ratio over tickrate, it results in 2/64=0.03125. You could change the cl_interp_ratio to 1, giving the game one less data packet to work with, which would reduce cl_interp to 0.015625. That's what a lot of people used to reduce latency a bit.

With modern games, including CS2, VALORANT and Battlefield, interp time (cl_interp) is set server side, so you no longer have control over it. Keep that fact in mind, then think one step further: What would be the reason cl_interp is set wrong for you, but not for everyone else? What are the odds that it happens to you in every single game, but not to the opponents or some people you may be in Teamspeak/Voice call with? If there was an actual bug in the way cl_interp is set, it would appear for everyone.

it's also strange how changing ping can fix interpolation

It can't. Ping has nothing to do with interpolation. Read: https://developer.valvesoftware.com/wik ... rpolation) All the ping changes is the amount of time the server has to roll back the game state to check if you landed a shot or not. Jitter is what's actually bad, not ping. However, your jitter has to be pretty bad in order to mess with the game's netcode, to the point where you'd see warning icons within the game's diagnostic overlay.


Alright, now that we have that out of the way, I want to contribute to this thread since it's about VALORANT specifically. There was a situation I witnessed about 2 years or so ago. You probably know about stat tracker websites for the game. What I've done to test something I was curious about was pretty simple: I had 2 accounts at the time. As you likely know, even an "unranked" gamemode like Deathmatch has a hidden MMR/ELO (which, in my opinion, is disingenuous and the mode should be renamed to "ranked deathmatch"), and you can estimate your hidden ELO pretty accurately. Play a round of DM, go to the stat tracker of your liking and look at the ranks of players in the DM game. Now, when I did my test, I had two accounts: One of them had a hidden ELO around the Silver to Gold mark, as every player in my DM games was always either Gold or Silver ranked in the actual competitive mode. On my other account, there was a hidden ELO of Immortal/Radiant, as every player in DM games on that account was either high Immortal or Radiant.

Funnily enough, in the Silver/Gold lobby, I couldn't react to anybody. The only times I got kills was when I was ratting in a corner and holding offangles. As soon as I was peeked on a somewhat common angle, I was dead, mostly before I could see someone on my screen. I ended that game somewhere around the bottom of the scoreboard with less than 20 kills (yes, I was actually trying!). Keep in mind also that I'm a former 4800 ELO Faceit player (Top 0.1% worldwide) in CS:GO and I have a 130ms average reaction time.

Immediately after that game finished, I logged out of that account, switched over to my other account and played another DM against Immortals and Radiants. I was winning that game by more than 10 kills difference, nobody was shooting back at me at all and I was called out for cheating by a guy who turned out to be Radiant. This happened quite a bunch of times every time I tried it.

My friend who plays this game regularly recently observed something wild. He was in a 5 stack playing against what would later turn out to be another 5 stack. First match, they get rolled crazy, like 2-13 or something. What's important to note is that they weren't outplayed per se by crazy strats or util usage, but they simply couldn't shoot back because every death was the same as comonly described: Instantaneous deletion without the ability to react to anything. Next game they queue into the exact same 5 stack on the opposing team. Same players, same lobby, same server. Everything exactly the same, even the same map. You probably guessed it, opponents got annihilated 13-0. They even typed in chat how my friends were suddenly so good and gained godlike reflexes, while beforehand they would stand on angles for 500ms without realizing an opponent was there (from their perspective, of course).

The best part is when I was talking to someone who's not been playing PC games for too long. She's started tac-shooters with VALORANT, so she has no idea what an actual, real video game plays like. I was bursting out laughing when she asked me one day why VAL feels so inconsistent, that se can shit on people in one game, lose that game regardless, be on lower ELO in the next game but still be instantly deleted by players with lower skill levels than in the previous game.

I know someone who's regularly playing the game with a duo. He also was alerted to the fact that he would always instantaneously die to some random guy, spectate his duo mate and the same opponent that insta-destroyed him was missing all the shots on his duo. They decided one day to swap accounts, and would you know it... Suddenly roles got reversed and his duo was the one constantly being "pre-fire" insta deleted. His duo is also usually higher ranked than him, but in that "account-reversed" game, his duo got single digit kills while he was dropping a bombshell 30 kill game on them.

Ask around in your friends circle, I bet you'll hear similar things about the game.

I'm not saying this is conclusive evidence, but I strongly believe VALORANT has mechanisms in play that will manipulate your experience on the game depending on certain factors. I believe it's something similar to CoD's EOMM, and skill-based hitreg. Is everything above just random chance? Is it a coincidence that I just happened to be playing the game regularly at the time on the account that can't shoot back against Golds, and wasn't playing on the Immortal/Radiant account for a long time? Maybe.

I haven't played the game since that "DM incident" and I don't plan to do it, unless the devs actually address it, in whatever way they want. Either, they outright admit that they're using some form of EOMM, or they acknowledge that the netcode is fundamentally broken and detail strategies of what they're doing about it.

I get what you’re saying about interp being fixed server-side, but here’s the part that doesn’t add up for me: if it really was always fixed and not affected by anything else, I shouldn’t be able to change how smooth enemy movement looks just by manipulating ping or creating controlled packet spikes. Yet I can. For example, if I throttle a small % of packets for a short window (like 10–30ms), suddenly sonic fast enemy animations normalize and I can track them way more consistently. That shouldn’t happen if it’s purely a fixed server setting.

To me, that points to something else in the pipeline — either how lag compensation is applied in borderline cases, or how interpolation reacts to uneven packet pacing. Ping itself isn’t supposed to “fix” anything, but in practice, raising my ping, spiking my ping, changing network buffering setting or forcing small bursts of loss can smooth out gameplay. If it were just placebo, it wouldn’t be that repeatable.

So while I agree that interp is technically locked, I don’t think it tells the whole story. There’s clearly some sensitivity in Valorant’s netcode to packet pacing or timing irregularities, which explains why people sometimes get better hitreg on a higher ping route or by using tools to shape traffic.

I do not believe in the EOMM theory.

like cmon:)) marvel rivals admited to having it, don’t you think every other has it at a lower “level” just to make you addicted to it? until someone has the full code you cant be certain, more you play more likely to spend everything its a business!

[13n47]

like cmon:)) marvel rivals admited to having it, don’t you think every other has it at a lower “level” just to make you addicted to it? until someone has the full code you cant be certain, more you play more likely to spend everything its a business!

If thats the case it has the total opposite effect, cuz the game feels like shit and i can only count a handful of times i have played it since 2022. if the netcode was consistent then i'd play it more often and be more inclined to buy skins even. But the risk for Riot doing such a system outweighs its implementation, the negative press and possible lawsuits as they promo it as a e-sports game.

[mel]

@ChristophSmaul1337
i have two different isp cables, both of them are small local companies and they rent lines from bigger ones - rostelecom and transtelecom

about eomm - riot are far too incompetent for that, plus i never see streamers, even smaller ones, experience the level of ferraris that i sometimes have

as previous user mentioned, i once managed to manipulate it by toggling vpn and relogging, also going afk sometimes helped

[themagic]

@ChristophSmaul1337
i have two different isp cables, both of them are small local companies and they rent lines from bigger ones - rostelecom and transtelecom

about eomm - riot are far too incompetent for that, plus i never see streamers, even smaller ones, experience the level of ferraris that i sometimes have

as previous user mentioned, i once managed to manipulate it by toggling vpn and relogging, also going afk sometimes helped

Aahaha the AFK thing is the most funny mindfuck where even I can confirm that this a real thing...

Your teammates start too play like better and improve ? and even manage to win some rounds without you ? in the time when you are afk or you just don't move for some 2-3 rounds ?

But as soon you start moving and play again and try...everything like slowly degrades again just after some 1-2 happy round with you ? but your teammates like start to suck in that time and play worse again and fail and eveything sucks and degrades more and more and not only for you and until end of the match ?


Yes this real true Valorant experience haha...

Horror and Crazy...call help pls haha cause this very hard to explain and describe but i am 100% not Shizo...

[Jam3]

An idea for someone who has too much time on their hands or is smarter than me:

Pro City games between streamers usually have both perspectives of the shooter and victim. Time sync and compare both perspectives to see the desync. This is probably the easiest way to get enough quantitative data on shooter and victim POVs to convince anyone that desync is real.

Replay system will most likely be smoothed over since its all server side.

[mel]

Recently i have played val from a different city/region for a week, no big difference in ping (60 compared to usual 75-100)

BUT!

While playing on a terrible hardware and peripherals from an internet cafe, the game was 10 times easier to play, I only had 3 negative games in ~20 matches. Now im back at home, playing on a 480hz monitor w/ 9800x3d, way more comfy and it feels like playing against F1 drivers again.

The issue is real and its my ISP fault. I wish i knew what exactly to test my ISP for.

[donger]

Recently i have played val from a different city/region for a week, no big difference in ping (60 compared to usual 75-100)

BUT!

While playing on a terrible hardware and peripherals from an internet cafe, the game was 10 times easier to play, I only had 3 negative games in ~20 matches. Now im back at home, playing on a 480hz monitor w/ 9800x3d, way more comfy and it feels like playing against F1 drivers again.

The issue is real and its my ISP fault. I wish i knew what exactly to test my ISP for.

Why do you even wanna test it for lol, it's busted and the only thing you can do is changing it.