internet latency & effects on hit registration

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
Locked
User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: internet latency

Post by Chief Blur Buster » 07 Jun 2019, 18:26

Many YouTube videos galore.
This one is not a Battle(non)sense video but it may simplify understanding for those who need a simpler "ELI5" video:

phpBB [video]


Shots that are made that looks like they missed, but hit.
Shots that are made that looks like it hit, but missed.
This video tells you why in a more "Explain Like I'm Five" manner.

For seasoned gamers, hitreg feel is totally different at 60Hz versus 240Hz, 125Hz mouse versus 1000Hz mouse, 64tick versus 128tick, 30fps versus 300fps, near zero-jitter fiber connection versus very jittery slow cable Internet connection, etc. It screws wonky with the predictive hitreg algorithms. Sometimes you have to compensate with different gaming tactics. The algorithms in a game tries to be a Great Equalizer for all the players in a game, and it sometimes fails to try to level the playing field. But oftentimes, one has to tell themselves to "suck it up cupcake" in tolerating the hitreg physics hardcoded in a game.

The problem of a high performance gamer with sub-10ms button-to-pixel CS:GO, running on fiber connections, means even tiny 1ms-to-2ms ping-volatility-pattern-changes (regular volatility, erratic volatility, volatility spikes, intermittent volatility, etc) can actually now create noticeable hitreg-feel changes to a trained/seasoned gamer to a human on an older game running at ultrahigh framerates. Ugh. That's the name of the game of the falling noisefloor of error margins. We upgrade everything to the hilt, and then feel hitreg became worse for certain situations somehow. Crazy.

It's come to the point where high-but-perfectly-consistent latency is easier to score hitregs on, than low-but-volatile latency. That's because we can lag-compensate -- e.g. shootaheads/shootbehinds. A perfectly exact 10ms change means something is predictably 10 extra pixels ahead or behind 1000 pixels/second (1ms = 1 pixel per 1000 pixels/sec). So we can train ourselves to shootahead/shootbehind as a human compensation. If your enemy is moving 4000 pixels/second across your screen, 10ms change may means 40 pixel change. It's like aiming ahead physically in real life -- like if your bullet is slower or faster, you have to aimahead further if your projectile is slower (e.g. shooting with arrows instead of a gun) towards a moving target. Now your latency varies? Now, if your ping volatility worsens, you can't compensate quite as easily, because the invisible valid hitreg zone is jittering all over the place away from the actual steadily-moving enemy. Ping compensation algorithms try to work around that but, your ping jitter is not the same as your competitor's ping jitter, and the ping jitter can interact by rounding-off to the previous/next tick cycle -- the server ain't waiting 1ms longer to make sure your packet is included in the previous tick. So, sorry, your hitreg zone can go wonky even with 1-2ms ping volatility changes.

Think of it mathematically...
...1ms equals 1 centimeter per 1 meter/sec motion
...1ms equals 1 inch per 1000 inch/sec motion
...1ms equals 1 meter per 1000 meter/sec motion
...1ms equals 1 pixel per 1000 pixels/sec motion
...1ms equals 1 degree per 1000 degree/sec motion
etc.

Now vary that millisecond on your side.
Now consider your enemy may be varying (performance, ping, etc)
Now consider the network.

Being fast action games where flick turns frequently far exceeds such numbers, as soon as the noisefloor of other latency issues (by upgrading everything we can to the hilt), we more starkly self-see the latency noise of other things like the other gamers, server, netcode, etc. The millisecond we used to not feel, now can get felt.

We try to measure, but those single ping numbers don't tell much. However, one thing does not change: Post measurement proof of your problems. In troubleshooting we need substantiated claims, not unsubstantiated claims.

Nontheless. It's very, very hard to control consistency in netcode -- even if playing on an ultralow-latency LAN on a real fast server. It's going to blast statistical noise out the wazoo. Much easily hidden like dust swept under a rug, by your garden variety 30 frames per second on a 60 Hz monitor. A 50ms amplitude of noise easily hides 1ms well below the noisefloor.

But as we push performance, frame rates, refresh rates, tickrate higher, mouse rate higher, etc -- the noise amplitude tightens. A 1ms change is potentially more noticeable in an 8ms-amplitude noise, unlike 50-to-100ms-amplitude noise (let's say, the latency noise of the whole game from you to the other gamers -- all the equipment lags and network lags combined, including netcode and server). Sometimes figurative textbooks need to be rewritten. So many onion layers of statistical noise are being peeled and peeled more in ultrahigh-framerate older games, to the point where lagfeel differences of tiny ping volatility differences are actually legitimately perfectly complainable topic matters for the sensitive gamer, even among the placebo-claim mix and other reddit/commenboard dismissals. (But rheoretical question: What do we do, within our ability?) But we need measurements. Yes, MEASUREMENTS -- something I totally agree on. Hard data, I say!

Now, there's so much latency noise -- you have no control over your enemy's ping jitter, nor the server's ping jitter. All that is going to contribute to hitreg weirdnesses you can never control. You can only unpeel your own noise as much as you can by upgrading everything to the hilt like a $5000 fine-tuned system on a fiber-optic business gigabit connection -- and STILL have hitreg problems. Then it's alas, often a "suck it up cupcake" situation.

Now, if you know hitreg feel is much better testing your system at your friends (physically moving your system to your friend) and you ended up liking the hitreg feel better on your exact same system. If so, then perhaps you want to play the router lottery and/or business ISP lottery, moving, or some other drastic route. Hitreg feel consistency/inconsistency is never going to be perfect, you can only do your best.

Sometimes it's just easier to change your gaming tactic to other game tactics that are more immune to ping-volatility problems. (Example1: Shooting someone running away or towards you, is more hitreg-reliable, than shooting someone strafing sideways across your vision. Example2: Double or triple shot to give a bullet spray to cover a wider latency error range to capture a bigger likelihood of hitreg, if the scoreable hitreg is lagging in an empty space slightly to the side of the strafing onscreen enemy. You see many Twitch gamers doing intentional double-shots and triple-shots to sideways-strafing enemies as a hitreg-compensation tactic)
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

mello
Posts: 251
Joined: 31 Jan 2014, 04:24

Re: internet latency

Post by mello » 07 Jun 2019, 18:39

open wrote: Also you say there is no point to measuring now. The point of measuring is to KNOW WHAT THE PROBLEM IS.
We know what the problem is. Problem is caused by network congestion and network performance fluctuations that are happening in the area where person lives, it happens because of network overload / too many people using internet at the same time. It happens most likely because ISP's infrastrusture is old or is just not good enough to handle the load, or even worse, it is no properly configured for all available services at high load scenarios. It creates problems with proper UDP packets delivery. This is obvious when you bother to read my posts. You won't be able to fix it anyway, reporting it to the ISP will also not work, they simply don't care about gaming/ping times/UDP packets.
open wrote: But I guess your advice is just to assume something. That I have never witnessed over 30+ isps in 6 states. And change isp without confirming that even though you could likely measure it in 30 mins or less depending.
And you never heard in the last 20 years about people complaining about hit registration problems in FPS games ? You must have been not paying attention. Or maybe you just never "experienced" it in the same way as some people are seem to not be bothered by motion blur, stutter, flicker etc. I have seen that happening, but more often when it comes to casual gamers, they often miss a lot of things that more experienced and dedicated gamers do know about and notice. For example, there a lot of streamers that are just your average skilled players, and then they rage about getting outplayed and they say things like: "how is that even possible", and you on the other hand can clearly see that they have made a mistake in the game or just clearly missed the shoot while not being even aware of that. This issued with hit registration was so bad at some point, that people started creating networking / system related tips & tricks trying to optimize both the PC and network side of things. This included bios settings, system setting, system services, registry scripts, disabling unused devices/services, enabling advancted features in the network card (Intel NIC's), altering network setting in the game config etc. It never fixed anything, outside of maybe system having more resources avaialble and maybe being slightly more responsive. Other people tried going with hardware route, by buying new routers, PC, fast monitors, gaming gear etc. all in hopes of mitigating the problem in some way. And, again no difference.
open wrote: So yeah not very good advice and you wanting to repeatedly post a wall of text over a distinction that was made but you missed a page ago, really doesn't help anyone. Please give more useful advice.
Wall of text with full explanation of the problem. Yeah not useful, not at all. Did you even bother to read that this problem is not fixable on clients side ? Its funny if you think that you can fix or help someone with that, when no one ever was able to do that in the last 20 years, except of using the methods i described. And some people are not even able to do that, and they are stuck with what they got with no other options possible in terms of choosing the ISP.
open wrote:Figure out and confirm the problem then act.
It has been done over 15 years ago. Please read my posts again, because it looks like you are still not getting it. The only thing that has not been figured out is what is exactly happening to UDP packets on a technical level. The most likely scenario is something like load balancing solution that ISP uses to improve network stability for all customers and it probably unintentionally ignores (lowers priority or something) time sensitive packets like UDP in favor of networkreliability and uninterrupted connectivity.
open wrote: I'm sorry I've given my 2 cents and I'm taking a break from this thread for now.
All comment are useful, the problem unfortunately was your attitude IMO and oversimplification of the problem without fully understanding it. I welcome any advice about testing side of things, but then again this it not something that can be fixed on the clients side unfortunately. And not everyone is capable of or even willing to do more advanced testing and looking into this issue.
Chief Blur Buster wrote:Discuss and debate away! Keep it crispy, not supernova-hot...
All is good, as long as are there is no name-calling and insults :)
Chief Blur Buster wrote: There are lots of networking problems that definitely exists that don't have easy measuring tools available for them. One example is Cisco has a lot of high end expensive ISP network analysis stuff that actually measures some really interesting data -- since the backbone people need these expensive tools to properly measure their big iron. A lot of that data doesn't ever get seen by you nor me, unfortunately. No user-downloadable EXE can tell you. Unfortunately I do not know if this is the case here or not, but I definitely keep an open mind (Blur Busters policy). The fact is, sometimes it's not easy to measure without a big spend...
A lot of ISP's are also using preprogrammed and preconfigured software/settings with minor adjustments to their particular networks. They only collect and analyze the data they are required to obtain or when they need to troubleshoot something, then they may make an use of more advanced analytical tools. I don't think that there are too many network technican geeks working for ISP's who look that deep into this kind of stuff. As long as everything works and there are no complains from clients regarding disconnections/loss of signal, this is all what matters. Unfortunately, when it comes go gaming, it is still considered a niche that no ISP is paying attention to or is taking it seriously. I remeber that maybe 10 years ago, one of the ISP's somewhere was offering "Gaming Package" as one of their internet network services. I do not remember exactly the details, but it included packet prioritization. It was also short lived in the end.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: internet latency

Post by Chief Blur Buster » 07 Jun 2019, 19:18

mello wrote:It has been done over 15 years ago.
Now.... I have something to say about that. Unfortunately what happened 15 years ago sometimes needs to be rewritten.

Back in the day, there was a huge amount of statistical noise of the dialup days where you had literally 500ms-amplitude worth of noise between you and your enemy. Let's consider 120Hz didn't exist yet in as widespread as today (unless you kept an old CRT around and intentionally ran at low-rez like 800x600). And dialup latencies and bad cable included, given the huge mix of dialup-and-cable-and-DSL users sharing the same servers of that era. Combined with all that massive motion blur of 60Hz displays. And the low two-digit framerates of that era. And the 125Hz mice. Heck, PS/2 mice were still very common too -- sometimes as low as 30Hz (33ms poll interval) for some of the competitors. The slower systems and rendering pipelines adding lots of lag. Combined, it was not unusual to have a 500ms latency amplitude between the best-latency and worst-latency pairs of all the gamers on the same system. Servers weren't universally filled with primarly high speed Internet users back then, yet.

Today has evolved to the point where there can be far less than 50ms of latency noise on some servers -- depending on the mix of gamers playing, everyone's generally on highspeed nowadays and if it's a same-city server targetting same-city gamers using a latency-equalizing matchmaker algorithm. Then the latency noise plummets like a rock by an order of magnitude compared to 15 years ago in some cases.

Thus, textbook rewrites. Sometimes re-measurements need to happen to rewrite the textbooks in the era where we have 240Hz monitors, 1000Hz mice, 128-tick servers, gigabit fiber. This is where little silly nuances -- even minor ISP's traffic shaping behaviours at the low-few-millisecond timescales -- start to produce hitreg-feel / latency-feel differences -- certainly no longer hand-wave-dismiss stuff anymore with the tighter/lower noisefloors today for a gamer that spends over one thousand hour per year playing these games and becoming well attuned to tiny feel differences.

We now encounter sub-10ms button-to-pixels latency on a well-tuned gaming system (that's even less than the 13ms button-to-pixels we encountered in our GSYNC-101 tests of CS:GO VSYNC OFF). Suddenly, a lot of statistical noise fall off. The formerly-obscured error margins start to reveal themselves -- like small-amplitude ping jitter at many ISPs!

In other words -- textbooks of "my numbers are fine" of 15 years ago need to be rewritten, so actual hard measurement data is strongly encouraged. The "numbers are fine" talk based on 15-year-old thresholds, do not cut it around Blur Busters anymore.

So.... Measure, measure. Don't judge the numbers based on a 15-year-old standard, just spill good hard data (preferably graphs rather than just a ping text paste) for others to analyze. While some measurement utilities are good, some measurement utilities need to display more digits of accuracy, those "round ping off to nearest millisecond" even is a problem for Blur Busters, so the programmers need to up their measurement-tool game too. Interesting human-relevant statistical noises reveal themselves when we show things at the hundred-millisecond timescales nowadays.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

nick4567
Posts: 118
Joined: 10 Mar 2019, 19:55

Re: internet latency

Post by nick4567 » 07 Jun 2019, 19:35

i get around 1-2 ms ping fluctuations as recorded by the cs go netgraph the problem is that i see nothing unusual when i get hitreg issues and the rare times i dont get hit reg issues nothing is changing as far as ping volatility
Last edited by nick4567 on 07 Jun 2019, 19:43, edited 2 times in total.

open
Posts: 223
Joined: 02 Jul 2017, 20:46

Re: internet latency

Post by open » 07 Jun 2019, 19:36

All comment are useful, the problem unfortunately was your attitude IMO and oversimplification of the problem without fully understanding it. I welcome any advice about testing side of things, but then again this it not something that can be fixed on the clients side unfortunately. And not everyone is capable of or even willing to do more advanced testing and looking into this issue.
Network congestion and internet performance fluctuations in relation to gaming can happen regardless of latency and packet loss issues and WITHOUT ANY serious ping spikes.
That is your first sentence in your first post and it is a gross misrepresentation. Everything that you talked about in your theories and walls of text is describing how packet loss and latency manifest. A packet being UDP doesn't magically make it unable to be understood as lost or delayed when it gets lost or delayed.

This whole chain of conversation could have been avoided if you were able to clearly explain that the issue you are talking about is with 3rd party tools to 3rd party servers (like icmp based ping to google.com) vs ingame packets. Yet you chose to say that it is not related to latency or packet loss. Be more accurate. And try to understand others better as well.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: internet latency

Post by Chief Blur Buster » 07 Jun 2019, 19:45

nick4567 wrote:i get around 1-2 ms ping fluctuations as recorded by the cs go netgraph
The existence of hitreg issues doesn't appear/disappear but the hitreg feel can change -- e.g. how easy it is for you to try to compensate for hitreg issues etc.

It's the nuances of the feel -- the intensity of hitreg problems, the pattern of hitreg problems, the ability to compensate via shootahead/shootbehind tactics, etc -- they can all change with various changes of variables. Seasoned gamers in esports get familiar with hitreg behaviours of a specific games and partially compensate for the more predictable hitreg error factors by various tactics. Often there's no pattern nor order to it. You can see the hitreg pattern change, but you can't make hitreg perfect.

If you're the LPB (...low ping bastard...) on a high-latency-variability server competing against a lot of erratic-latency gamers -- you ARE going to get a lot of erratic hitreg problems even if your Internet connection is perfect. You're going to make a lot of successful-looking headshots that doesn't even register. No way to solve that except simply changing your gaming tactics -- some find sniping stationary enemies or people running towards/away from you is more latency-reliable than trying to hitreg sideways-moving targets (e.g. strafing enemies) in such a situation. Depends on the type of competition you're doing, the game, the map, etc.

It's all crapshoot in those borderline situations. On those high-latency-variability situations (especially if everyone has high latency variability, not just you), it just ends up wasting ammo. Gaming tactic changes ends up being warranted.
nick4567 wrote:nothing is changing as far as ping volatility
Doesn't mean your enemies didn't change ping volatility. Hitreg feel changes occur even in situations where only the other enemies have changed their ping volatility (heck -- even as simple as someone else in their same connection, beginning as Netflix stream). Minor hitreg feel changes are more visible on well-optimized systems.

You can't control or predict other people's ping volatility. If everyone's on a LAN it's a lot more predictable and levels the playing field more. But the whole Internet is a crapshoot of ping variability that is completely unrelated to your ping variability. For many, the player-matchmaking engines of many games (including CS:GO) tries to group players of similar-latencies automatically in the same game. That helps. But...

Even if there's only three players on one server:
Person A to server
Person B to server
Person C to server
server to Person A
server to Person B
server to Person C

If only one of the six has any changes to ping or ping volatility (suddenly increases or suddenly decreases), that can create noticeable hitreg-feel pattern changes even if your latency or latency volatility is fine. Remember upstream volatility can be different from downstream volatility. And that volatility may have completely nothing to do with your connection. And it begins to create weird changes to hitreg feel anyway. Then it's a "suck-it-up-cupcake" situation.

Tons of people are on cable Internet connections now, and some of those are hugely asymmetric (Me: I'm on a gigabit cable Internet connection that only has 30 megabit upstream -- occasionally I get a very ugly 50ms volatility on the upstream on fast.com and speedtest.net especially when other members of the household suddenly grabs my upstream).

During evening peak when everyone is online, lots in my neighborhood are sharing my cable node's upstream. And that means hogging 1Mbps versus 25Mbps of my upstream creates a 50ms-amplitude ping volatility shows up like a christmas tree even at 60Hz in hitreg feel -- the simple mere act of someone else in my home suddenly uploading a YouTube video, a Dropbox sync, an iCloud sync, or other mudane everyday upstream-hoggers -- clogging my upstream temporarily in the middle of a game.

Toto, this ain't Kansas anymore of your yesterday's rare bittorrenter of 15-year-ago. Everybody's become much bigger uploaders nowadays thanks to those boom of cloud services, and it's wreaked havoc on random upstream ping variability in ways not foreseen 15 years ago, on top of the soup of firewalls, security monitoring, NAT, trafficshapers, average increases/decreases in upload:download asymmetry in your neighborhood, as well as also all the massive Pandora Box cauldron of latency-error-margin contributors. Hitreg DEFINITELY feels different today than 15 years ago, full stop. It's just the way life is. Deal with it. :D

Even if you have perfect 0ms latency with 0ms volatility, your hitreg on me will definitely totally change in the middle of a game.
Just because somebody else is now screwing around with my Internet's upstream ping volatility. My downstream is amazingly dandy, but upstream is :barfemoji: -- a gigabit connection....with....only 30Mbps upstream and 50ms upstream ping volatility in an upstream-hog situation. </I wish I had symmetric gigabit>

Now, there's also a wildgoose-chase effect to be considered. If you move to one neighborhood, you may be matchmaker'd with your high-ping-variability neighbours in a server's game matchmaking algorithm (even if you upgrade to a perfect Internet connection, it doesn't mean everyone else has the quality if low-ping-variability connections are rare in the city). So yes, it can be a situation of "It's the network fault, but not my ISP's fault". Totally beyond your control. And then you move to a good city or neighborhood with lots of gigabit fiber, then you end up being server-matchmakerd with lots of consistent-ping neighbours via your server's latency-equalizing matchmaking algorithm.

In those situations it is sometimes easy to blame your ISP, when it's simply the area/county/province/city/country overall Internet infrastructure just screwing around with hitreg because everybody's ping variability is continually changing at different times of the day. Even if you can theoretically get the perfect Internet connection, your neighbours may have ping jitter that constantly changes in amplitude (at different parts of the day for them), screwing around your hitreg-feel in realtime. Even if their own 2ms-amplitude jitter worsening to an 8ms-amplitude jitter as Internet Peak approaches, creates noticeable hitreg-pattern changes, even if it's not your connection. All totally beyond your control.

One city (or country) feels like playing on a LAN, a different city (or country) feels like playing in a zoo soup of dial-up-and-cable-and-fiber customers. It's often somewhere in between though. You're lucky if you live in a nearly pure fiber optic country where nearly all your neighbours are on symmetric FTTH, all the city backbones have ultra-low variability, and that the game server matchmaker logic is automatically pairing you with your fiber neighbours. Then it feels like you're playing on a LAN, and the hitreg behaviours start feeling more predictably compensatable for everyone. (In this situation, 1ms-to-2ms volatility-changes punches much more noticeable hitreg-feel / hitreg-pattern changes above the noisefloor)

The millisecond matters, even at the network layer level. But oftenso, it's beyond your control -- you have no control over the other players' ping variability.

It's just the way it is, it's just the way the cookie crumbles.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

nick4567
Posts: 118
Joined: 10 Mar 2019, 19:55

Re: internet latency

Post by nick4567 » 07 Jun 2019, 20:05

i mean i adapt as best i can to the hitreg issues i do the things u mentioned subconciously almost but when the problem is at its worst nvm the hit reg it feels like i dont even get a chance to react this isnt just hitreg this is people seeing me way before i see them sometimes i guess if you think of it like 0-10 0 being worst hitreg and 10 perfect hitreg mine probably fluctuates anywhere between 2-9 when its at a 2 even if enemy isnt looking at me ill have a tough time killing them and my mouse feels like dragging through mud peeking corners is a no-no because i get insta rekt no chance to react at all and when its at an 8 or 9 it feels super responsive super smooth i can get clean kills flick shots 1-2 bullet kills wheras when its at a 2 sometimes a whole clip wont kill
Last edited by nick4567 on 07 Jun 2019, 20:48, edited 1 time in total.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: internet latency

Post by Chief Blur Buster » 07 Jun 2019, 20:43

nick4567 wrote:i mean i adapt as best i can to the hitreg issues i do the things u mentioned subconciously almost but when the problem is at its worst nvm the hit reg it feels like i dont even get a chance to react this isnt just hitreg this is people seeing me way before i see them sometimes
While servers try to equallize the playing field, Hitreg asymmetries fluctuate massively for many reasons that are well beyond your control. Are you playing with chosen people, or playing with automatically matchmaker'd people?

Also, remember, it could be mudane stuff such as players running cheats at different times of the day -- an epidemic in some parts of the world -- but this might not be applicable here -- just throwing this possible cause out there.

Nontheless; have you tried any unorthodox moves such as getting a gaming VPN service to gaming servers in a different city in the world? The hitreg-feel might actually be preferable, especially if the latency volatility pattern of that particular city/server is more favourable for your situation (e.g. less ping volatility, even if higher average ping). The important thing is getting your ping consistency of the VPN as consistent as possible (something that at least you have more control over, than other player's ping volatility). Sometimes it's more fun to play at a very slightly higher ping if you find consistency (less jitter for everybody) -- say, a city that has better infrastructure that will produce more predictable latencies that screw around less with hitreg.

You might increase your own latency by 5ms or 10ms, but then everybody else's average ping volatility might fall by more than that amount; making it more than worth that latency trade-off. In some cases, you might even have a decrease in latency, especially if the gaming VPN actually routes you to an amazingly great low-latency server. Regardless, your latency volatility (ping jitter) should at least decrease.

This might not help your hitreg, but it's an unturned stone to try, especially if the blame is actually your neighbours in the same city having high ping jitter creating nasty hitreg asymmetries (real or perceived/training).

There are some games where its latency-compensation algorithm in some games weirdly makes it easier or harder for hitreg at different times -- often because somebody else's ping or ping volatility changed. Algorithms that partially handicap LPB advantages to equalize playing field. In principle, these automatic compensatory measure in an algorithm's flawed attempt to level the playing field, strange hitreg asymmetries can occur when the ping volatility ratio constantly shifts around. Netcode algorithms have a tall order to attempt to automatically level a permanently-continually-distorting playing field -- making sure that even slow Internet users on high-volatility have a chance. TL;DR: It's common to see huge surges of hitreg accuracy increases/decreases (for you or for them) that are beyond your control.

In online gaming on the open Internet -- the pursuit for lowest absolute latency sometimes neglect latency variability, which can ruin a good game moreso than absolute latency. Thus, I see great utility in playing the "gaming VPN" lottery; trying a few gaming VPN services.

People debate about the validity of gaming VPNs until the cows come home, but some swear by them. They're worth it...sometimes. And sometimes a huge amount. Your mileage will vary. I can see situations where gaming VPNs help, by getting you away being matchmakered with volatile-latency players in your area -- and instead matchmaker'd on a better, consistency-latency server with more predictable-latency player participants. In a different part of the city, or different part of country, or even another country that has better Internet infrastructure with low ping volatility 24/7. For you, this may creates huge differences in hitreg feel and hitreg consistency.

That said, your weak links still include your own Internet connection. Maybe it is not your weak link, but if it is the weak link then it's probably all your neighbours' weak link too. Some cities just have very crappy high-volatility infrastructure that will vary throughout the day. If you're not wanting to move but your Internet options are limited... Thus, in these particular "I'm stuck in my cesspool" situations, the combination of a premium-cost business Internet connection AND also a gaming VPN is sometimes an (expensive) shortcut to solve a server-wide ping jitter problem -- to allow you to bypass a local ping volatility cesspool in a city of few ISP options -- to more latency-predictable servers with more latency-predictable players ...as long as you can get a consistent low-ping-jitter over the gaming VPN to the gaming servers.

(But first, try some paid gaming VPNs without switching your Internet service yet. See how hitreg feel changes)

No guarantee though. Hitreg may still be crapshoot (pun intended).
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

open
Posts: 223
Joined: 02 Jul 2017, 20:46

Re: internet latency

Post by open » 08 Jun 2019, 02:28


mello
Posts: 251
Joined: 31 Jan 2014, 04:24

Re: internet latency

Post by mello » 08 Jun 2019, 03:00

Chief Blur Buster wrote: Now.... I have something to say about that. Unfortunately what happened 15 years ago sometimes needs to be rewritten.
I hear what you are saying. But i am still not convinced that this much has changed when it comes to networking. ADSL/DSL/VDSL lines work basically in the same way as they did 10-15 years ago, some ISP's have made some improvements in their infrastructure, others didn't. The only thing that has changed is the fact that literally everyone has access to the internet now, so there is even a greater possiblity of additional noise, interferences and bottlenecks happeneing even then before,at least in a certain cases, and all because the overall traffic is that much greater.

An although you may disagree with that, i still believe that when it comes to playing fps games online, proper hit registration and accurate enemy player positioning is all that matters. I already gave an analogy about what happens when a high skilled player is being bootlenecked by a bad internet connection. It doesn't matter how good he is and how great PC, monitor and his gaming gear is, even with the lowest input lag possible, if shots are not registering properly, it won't make a difference online. I have seen and tested it personally, you can easily achieve a tangible difference against other players, simply by lowering input lag as much as possible by using high hz monitor, very high and constant fps and high hz mouse. With proper skills and a good connection it seems almost unfair when playing on a public servers against most people. But when plagued by registration problems etc. it helps only a little, and only when first few bullets actually register on a enemy model. This is also exactly why, when this problem occurs players often resort more to spray and pray tactic in some games instead of trying getting headshots with a simple 1-2. I even run some tests by artificially increasing my input lag, simply by locking both fps/monitor Hz to 60 and setting mouse Hz to default (125Hz). This is why i will probably always say that internet network performance is all that matters, because this is exactly what i have seen and experienced myself, it perfectly reflects the reality.

On the other hand, when the playing field is equal (LAN) or kinda equal (Internet but with no bootlenecks) every ms matters, especially at the highest level. This is where lowest input lag solutions will always shine through. And for that reason i am all in support of what BlurBusters is and always been about.
Chief Blur Buster wrote: If you're so sure it's network lag, have you tried the Business Internet service of any of your ISPs?
Related topic on "many ISPs" -- sometimes one gets desparate. The support from business can be much better, even if you are paying 3x the price for the same Internet speed. In some cases, it removes a lot of traffic shaping algorithms from your connection, while giving you better technical support (e.g. quicker upgrades/replacement of a congested node if you're complaining about latency problems). There are even cases where a business connection bypasses all the traffic-shaping / microthrottling / securityscanning FUD and gives you better ping consistency especially in the upstream channel, even when sharing the same cable Internet node with consumer users. Other times it is no difference except that phone calls are answered faster with quicker fixes for Internet problems (And better node maintenance). These stories do sometimes borne fruit, so it's an unturned stone to try.
This perfectly illustrates what DSL Business Package was all about in the past. And all the things that you described were most likely a part of (at least in the past / depending on the service) and put under a Packet Prioritization umbrella.
Chief Blur Buster wrote: I know people who paid $100-$200/mo for a business Internet connection equivalent to an $50-$100/mo connection to solve a gaming problem. Overkill, but it worked for them when there were no alternate options.
Many people in Europe did exactly that when DSL was a thing back in the days. Now it is basically the same thing as ADSL for regular customers, the only difference being that it just offers a guaranteed sync speeds and fast acting customer support when needed.
Chief Blur Buster wrote: My downstream is amazingly dandy, but upstream is :barfemoji: -- a gigabit connection....with....only 30Mbps upstream and 50ms upstream ping volatility in an upstream-hog situation. </I wish I had symmetric gigabit>
Damn this is almost criminal to have only 30Mbps upstream in that scenario. Thank god i have a symmetric gigabit these days =)
Chief Blur Buster wrote: One city (or country) feels like playing on a LAN, a different city (or country) feels like playing in a zoo soup of dial-up-and-cable-and-fiber customers. It's often somewhere in between though. You're lucky if you live in a nearly pure fiber optic country where nearly all your neighbours are on symmetric FTTH, all the city backbones have ultra-low variability, and that the game server matchmaker logic is automatically pairing you with your fiber neighbours. Then it feels like you're playing on a LAN, and the hitreg behaviours start feeling more predictably compensatable for everyone. (In this situation, 1ms-to-2ms volatility-changes punches much more noticeable hitreg-feel / hitreg-pattern changes above the noisefloor)
Funny thing, because that was basically the case when it comes to countries like Sweden, Finland and Norway IIRC. They always had high end internet tech first here in Europe. And back in the day these countries have produced many legendary teams and players over the years. The idea is, that happened because they were able to learn the game and play constantly against each other in near LAN scenarios (very low ping times, low ping variance, no lags), while other countries in Europe where learning, training and playing against each other with +20ms / +50ms higher ping times. And when that happens, and when you actually play on LAN against one another, players that were more adapted to real life scenario (LAN) were performing better.
nick4567 wrote: i dont even get a chance to react this isnt just hitreg this is people seeing me way before i see them sometimes i guess if you think of it like 0-10 0 being worst hitreg and 10 perfect hitreg mine probably fluctuates anywhere between 2-9 when its at a 2 even if enemy isnt looking at me ill have a tough time killing them and my mouse feels like dragging through mud peeking corners is a no-no because i get insta rekt no chance to react at all and when its at an 8 or 9 it feels super responsive super smooth i can get clean kills flick shots 1-2 bullet kills wheras when its at a 2 sometimes a whole clip wont kill
This is exaclty what happens. As i said before, network congestion does not only affects hit registration, although this is mostly talked about. It lowers your perceivable reaction time, especially in peeking corner scenarios, and what happens in reality is that your enemy sees you first on his screen, before you can see him on yours. It it either a misplaced / incorrect enemy model positioning due to UDP packets being delayed / delivered late / delivered out of order, that which creates this phenomenon. Your reaction time and enemy model positioning are interconnected, and you either prefire when you hear the enemy and assume his position, or you aim and shoot when model shows up on your screen. If it doesn't show up properly, you react only when it actually does, and that is often too late, because enemy player already saw you first and is shooting at you. Back in the day when i was playing for fun and casually with bad internet connection i was quite often being accused of cheating and have been banned many many times on countless private servers, because of a sudden increases of performance and the ability to make an amazing plays at certain times. I was literally able to make players leave the server, because they were thinking that they are playing with a cheater. And all of that because when playing with poor network performance i was basically handicapped, and when that happens you are limited to the things you can do, you underperform, you often need to play differently, be more careful, you need to spray way more etc. and when bottleneck is lifted the gaming feels almost effortless.

There is not much you can do about that, and there is probably nothing else outside of what i and Cheif described.
Last edited by mello on 08 Jun 2019, 03:25, edited 2 times in total.

Locked