"Demystifying Lag" not accurate

Discussion for GU League Players
User avatar
Cobra_Fast
Dev Monkey
Dev Monkey
Posts: 322
Joined: Sat Oct 07, 2006 11:51 am
Location: Germany
Contact:

Re: "Demystifying Lag" not accurate

Post by Cobra_Fast »

ts wrote:The voice communication is allowed with the mentioned tools and other tools are treated regaring my quote.
That's not what the whitelist says. The whitelist allows you to use any voice communication, but it recommends you to use mumble or teamspeak.
ts
Dev Monkey
Dev Monkey
Posts: 970
Joined: Fri Jan 14, 2005 6:26 pm

Re: "Demystifying Lag" not accurate

Post by ts »

Yes, any voice communication is allowed, provided the name of the tools used have mumble or teamspeak as official product name.
GU league: http://www.guleague.org/
An introduction to TCP: http://www.lafkon.net/tc/
User avatar
Cobra_Fast
Dev Monkey
Dev Monkey
Posts: 322
Joined: Sat Oct 07, 2006 11:51 am
Location: Germany
Contact:

Re: "Demystifying Lag" not accurate

Post by Cobra_Fast »

Okay let's take a closer look to whats written in the whitelist at http://my.bzflag.org/bb/viewtopic.php?f=103&t=13245 .
Voice Communication

It is explicitly allowed to use voice communication for playing.
This allows me to use my voice to communicate with other players (preferredly from my own team), so I'm probably allowed to use any tool that allows me to do so.
However, the application used must be designed to run well with online gaming, in that they have no or a minimum affect on connection quality.
So I have to make sure it doesn't screw my connection. I totally did that, as I tried out the most common tools and compared them.
The following tools are approved for league playing:

* mumble
* teamspeak
Here the whitelist recommends me to use mumble or teamspeak as such a tool, it does not say that these are the only tools legal to use (as stated in the first sentence about voice communication). In case you're not familiar with the exact meaning of the word "approve", which is used to state the condition, you can look it up at the Wiktionary: http://en.wiktionary.org/wiki/approve .
if you know of any other tool that you think should be on the list, contact us.
This gives me the opportunity to make the people who run the league review a new tool so they can put it on the list of recommended tools if it succeeds their testing. It does not mean I have to ask for permission to use another tool than the ones listed in the 3rd step.

So from my point of view I'm not doing anything wrong by using Skype as it doesn't affect my internet connection in a noticeable way. Feel free to test this on your own, but dont threat me with banning without having a single peace of evidence that Skype is definetely a tool that's not fine to be used in league play.
smoooth
Private First Class
Private First Class
Posts: 81
Joined: Sun Aug 05, 2007 6:13 pm

Re: "Demystifying Lag" not accurate

Post by smoooth »

I will have to say the whitelist is completely ambiguous and poorly written, both points of view are completely legit because the rules contain contradictory statements in multiple places. I have lobbied for the "rules" to be re-written and clarified for over 3 years. They probably need to two versions one written in english and one written in german. Perhaps one other language but these two should cover 95% of the GU players.

There needs to be a group assembled to "fix" the rules. Unfortunately that means someone who is an admin and is also an active player -- not many of those. There also needs to be fair representation from active players NOT just the admins or friends of admins to do all the work. This is why the league should admin itself and not be dictated by a few who frankly, don't even play that often.

smoooth
User avatar
october rust
Private First Class
Private First Class
Posts: 22
Joined: Sun Jul 03, 2005 8:01 pm
Location: Porto Alegre, Brasil

Re: "Demystifying Lag" not accurate

Post by october rust »

Hey there,

Someone just pointed me this thread (I'm kinda retired from the game here :)

This doc was a group project that started about February, and took us a couple months of dicussion to get it done. The goal is to offer useful information for players that want to learn how to deal with lag.

Although, this is obviously not supposed to be a "complete solution" for all other gameplay issues - just effective stuff to go around one of its most common/dealable issues (lag), willing to eliminate the advantage between players that know how to fully deal with it in the league level, from players that don't. Abnormal and unpredictable things, like jitter and packet loss, have no much room in it, since those are not part of the normal gameplay, and neither there are ways to go around. Jit/ploss induce unplayable conditions, and the only workaround, in case of an official match, is to pause the game until the popcorny tank stops jumping around.

Unfortunately I don't have any more time ATM to discuss this subject further, but I just would like to drop a note about one of Smooth's comments, where lag is associated with "advantage". This sort of statement is very hazardous to the leagues, and may estimulate new players to induce lag in purpose, thinking that by doing that they would have a better performance. That's not just naive, but would also escalate lag issues, and spread around BZFlag a whinning attitude over an issue that can actually be overcome with knowledge.

Anyways, what we've wrote up there is preety much the best we could put up together with the limited time and experience we had. It certainly has flaws, and stuff to get done, and anyone with more practical experience and a deeper knowledge in the subject may feel free to improve it.

cheeers
User avatar
macsforme
General
General
Posts: 2069
Joined: Wed Mar 01, 2006 5:43 am

Re: "Demystifying Lag" not accurate

Post by macsforme »

Okay...

Smoooth, it would be nice if you had actual knowledge of the code or technical citations of benchmarks or other tests to back up an issue you point out or an accusation you make. Now, I appreciate that you observe discrepancies over time and you are able to form a theory from them, but without understanding the causes and effects the theory is meaningless.

The fact is that there are many variables affecting computer hardware, computer software, network equipment, and equipment at the ISP level that a user has no control over, so that you should not always expect to see the exact same network quality from all players when matching.
There is actually no justification for a player to have variable jitter or even jitter over 3ms.
90% of all jitter is created by the operating system or a computer that is poorly operated.
As the saying goes, 87.3% of statistics are made up on the spot. I have never seen so much as a patch from you nor do I remember any time you have discussed the code itself or how anything works under the hood. I would be very interested in hearing how you benchmarked a variety of client versions, operating systems, computer hardware, network equipment, internet connection types, and locations to arrive at these estimates; or how you analyzed and evaluated the code thoroughly and the results gave you these numbers.

In my personal experience, the ISP has more to do with it than anything as long as a player has reasonably modern hardware. This game has been around for 16 years and the requirements for running it have not changed much. For instance, on my cable internet connection from a well-known ISP in the heart of a metropolitan area, while on a wired connection (even with no router), I can still get an average of 5-9ms of jitter to most servers. I have tried many various machines that are all more than capable of running this game. Distance to the server and time of day also affect it. Lag and jitter both seem to drop significantly at night, and I generally have more jitter on European servers than on local ones. While lag is clearly affected by distance, increased distance also generally increases the number of factors that can impact the quality of your connection, so jitter can be impacted too.
When you hit the lagstats button it's just a snapshot of lag at that moment.
Again, foundation please. I do believe lagstats averages your network latency over a period of time, not at any given instant. That is why the lagstats show 0 +/- 0 for the first few seconds after you join the game, and why a large spike will cause you to show high lag for several seconds after your connection has returned to normal.

As for the rules, perhaps you could make specific suggestions for improvements and point out the contradictions that are so obvious to you. Nothing can be done with such general criticism you gave in your post.

Ts, I read the whitelist and by the wording of the "Voice Communication" section I did not take it to mean that any voice clients not listed are not allowed. Many players (including myself) have used Skype in the past while matching. If you think the wording should be clarified (or if you think someone should be banned, as you mentioned), that discussion should probably take place elsewhere.
smoooth
Private First Class
Private First Class
Posts: 81
Joined: Sun Aug 05, 2007 6:13 pm

Re: "Demystifying Lag" not accurate

Post by smoooth »

Constitution,

No sweat we can break this down. First, I'm not getting on the offensive here, but look what you just wrote
"As for the rules, perhaps you could make specific suggestions for improvements and point out the contradictions that are so obvious to you."


then in the very next statement you send TS a message that says
"Ts, I read the whitelist and by the wording of the "Voice Communication" section I did not take it to mean that any voice clients not listed are not allowed. Many players (including myself) have used Skype in the past while matching. If you think the wording should be clarified (or if you think someone should be banned, as you mentioned), that discussion should probably take place elsewhere."
Basically you, an admin, and TS, another admin have just expressed confusion and probably disagreement over what "voice communication" is actually allowed. So don't write me like i'm an idiot when I say it's not clear that even two admins can't see eye to eye on what the rules actually are supposed to be. That is an EPIC fail.

I will be more than glad to point out all the contradictions in the rules, as well as write new ones. Count me in on that.

Now onto the jitter statistics. Absolutely those statistics are made up BUT they are representative of my experience with jitter. Jitter is, in my experience, VERY OFTEN is created from the own users computer via WIRELESS, Background Programs, Voice Communication Software, Video Card problems, FPS issues, etc, etc. My point is that people like to blame jitter on their ISP when more often then not it's their own computer. Sure you can have your opinion and say otherwise and I can't dispute it, but that is what I FIND, from monitoring, trial and error, experience, to be true. Can ISP cause jitter, absolutely -- If you're in dense metropolitan area your chances are increased. I would love to see a network test going from a wired computer to the first hop (ISP) and generating 10ms jitter. That would be amazing to me. Please post if anyone has a screen cap of that.

Yes, lagstats averages over a period of time, in fact that is the only way you can get an "average", but the point is that it is a discrete amount of time and later that jitter may be increase or decrease and is not representative of the quality of the connection by hitting "lagstats" one time. You have to "hit lagstats" several times, the more the merrier, to get a more accurate representation of what's going on with the connection.
User avatar
macsforme
General
General
Posts: 2069
Joined: Wed Mar 01, 2006 5:43 am

Re: "Demystifying Lag" not accurate

Post by macsforme »

smoooth wrote:Basically you, an admin, and TS, another admin have just expressed confusion and probably disagreement over what "voice communication" is actually allowed. So don't write me like i'm an idiot when I say it's not clear that even two admins can't see eye to eye on what the rules actually are supposed to be. That is an EPIC fail.
I didn't say you were an idiot, and I certainly see your point here. However, this is a matter of conflicting interpretations and not an inherent contradiction. People do not always see eye-to-eye about the meaning of statements, especially across language barriers (which you pointed out... perhaps the wording should be changed to be more explicit). My point was that when people complain that the rules suck and are contradictory, nothing happens because we really don't know what they mean by that. However, a well-formed argument with examples and suggestions for improvement are always welcome. On the other hand, there are reasons why the rules are what they are, and quite honestly I have heard about very few complaints or confusion about them from anyone except you. Please keep that in perspective.
smoooth wrote:Now onto the jitter statistics. Absolutely those statistics are made up BUT they are representative of my experience with jitter. Jitter is, in my experience, VERY OFTEN is created from the own users computer via WIRELESS, Background Programs, Voice Communication Software, Video Card problems, FPS issues, etc, etc. My point is that people like to blame jitter on their ISP when more often then not it's their own computer. Sure you can have your opinion and say otherwise and I can't dispute it, but that is what I FIND, from monitoring, trial and error, experience, to be true. Can ISP cause jitter, absolutely -- If you're in dense metropolitan area your chances are increased. I would love to see a network test going from a wired computer to the first hop (ISP) and generating 10ms jitter. That would be amazing to me. Please post if anyone has a screen cap of that.
My issue was that you were taking your own experience with jitter and applying it to the general population without saying how you were able to make that jump. I will gladly concede that most of the things you mentioned can cause jitter... however, the absolute statements you made that "90%" of jitter is caused on the local machine and that there is "no justification" for having jitter over 3ms... where do those come from? I gave an example of my own machines which clearly do not create much jitter on the client side, and though you may experience people creating jitter on the local machine I don't see how you can apply that to the player population percentages that you did.

Also, I hate to make an ad hominem statement, but quite honestly I would give more credence to your statements if you either used more technical explanations as basis for your arguments (with benchmarks or code citations), or if I knew you actually were familiar with how the code works. BZFlag is an open-source game with freely-available code that many people have read and modified. It saddens me when people make accusations claiming a high likelihood of validity when I know there is another explanation for what happened. This is not to say by any means that non-programmers cannot contribute to evaluating game behavior, but rather that an understanding of what is actually going on under the hood helps a lot when someone claims that A happened because of B.
smoooth
Private First Class
Private First Class
Posts: 81
Joined: Sun Aug 05, 2007 6:13 pm

Re: "Demystifying Lag" not accurate

Post by smoooth »

Well I think admin take for granted that people know that rules actually exist. No one complains because people really don't know what the other players are doing behind their connection. I talk to people privately, i hear, and watch, and am able to notice things, as do others, because we've been playing for so long. For example, I had often suspected GS used a one button triple but never knew for sure until I heard him discussing one day. Presuming that conversation I heard was real, GS and I had a followup discussion about it and he made a valid point. In the whitelist it allows for setpoint and other mouse tools to be used, apparently one of the features allows the use of a pre-recorded triple. So what GS is doing appears to be legal but probably no one else knows that you can use a single button triple. If they did, they might complain such as me, or maybe not because most people don't take this game very seriously. He also pointed out 2.0.2 was legal, but probably not that many know it has a fullscreen radar and other bad bugs. If they did know, they might even go downgrade :(

As for the code. I do not look at the code. I never have. I could easily look and tweak but I'm more interested in playing the game. In fact I'm leary of people who do look at the code because I think it is too much temptation to tweak things yourself. You don't need code knowledge to see how the game plays. That's like saying you need to understand how a car works to drive one. Most people can feel when a car is out of alignment or if the car is in reverse without ever know how a suspension or transmission system works. I also need to know very little about bzflag code to understand how it interacts with operating system concepts and network layer etc . All that info, from BZflag, is passed to the operating system so therefore if the system is poorly running and has bad processes or overloading network programs, then will cause problems for bzflag as well. I'm not blaming anything on the BZflag code, my statements were completely based around how peoples individual systems can hurt the performance of the game via the operating system. Having outdated hardware, rogue processes, network taxing applications, poor memory management, list goes on and on. I don't consider myself a "non-programmer" but even that has its limitations. I actually think looking at the output of the game is as accurate as looking at the high-level C++ code, sometimes one may even be surprised how that high level code is translated at the compiler and lowest levels....
Jacko H
Private First Class
Private First Class
Posts: 123
Joined: Wed Apr 22, 2009 3:15 am

Re: "Demystifying Lag" not accurate

Post by Jacko H »

Where do I find watever it's called
Jacko Productions:
World War III, Kill Or Die, Nazarath, Warlock Adrenaline, Rabbit Season, Firing Range (Just Be Simple), Death Island.
User avatar
blast
General
General
Posts: 4931
Joined: Fri Mar 21, 2003 3:49 pm
Location: playing.cxx
Contact:

Re: "Demystifying Lag" not accurate

Post by blast »

Apparently using Google is too complicated?
"In addition to knowing the secrets of the Universe, I can assure you that I am also quite potty trained." -Koenma (Yu Yu Hakusho)

Image
tobylane
Private First Class
Private First Class
Posts: 130
Joined: Thu Jan 29, 2009 4:21 pm

Re: "Demystifying Lag" not accurate

Post by tobylane »

So.. put Bzflag on a ramdisk and don't play anyone or use any server more than 50 miles away from you. And give up on sucky isps.
Jacko H
Private First Class
Private First Class
Posts: 123
Joined: Wed Apr 22, 2009 3:15 am

Re: "Demystifying Lag" not accurate

Post by Jacko H »

Also after the play there is:
It makes it harder to get matches,
Abuse,
Cheat accusations,
if you use it it is lag abuse nd many more.
It isn't good to lag.
Jacko Productions:
World War III, Kill Or Die, Nazarath, Warlock Adrenaline, Rabbit Season, Firing Range (Just Be Simple), Death Island.
An SQUERRILz
Private First Class
Private First Class
Posts: 91
Joined: Wed Apr 25, 2007 2:08 am

Re: "Demystifying Lag" not accurate

Post by An SQUERRILz »

smoooth wrote:Although this is a great effort it lacks many points that need to be made:
smoooth wrote:Jitter is the absolute worst aspect of lag.
As has already been pointed out in some way, it seems like your comment is better directed as a general speech rather than relating to the lag article.
smoooth wrote:What is overlooked here...
I think jitter is reasonably acknowledged as a big baddy. It makes sense to state the known anyway, since some people will get it wrong. At this point I will not comment on jitter wrt advantage.
smoooth wrote:People with consistent lag can be visually adjusted for by shooting early
Really?
smoooth wrote:folks with varying lag and jitter can absolutely NOT BE.
I would like to see some mathematics here :) -- show us the relationship between jitter and the decrease in margin of error for the timing of a shot passing through the hitzone. Is 10 ms significant in this respect?
smoooth wrote:If you are serious about playing and you want to know how good you really are, you need a wired connection.
I think this shows how subjective you have made this address. If you have something worth saying, perhaps it is better received in a manner other than a rant.
smoooth wrote:Naturally, bullets will not hit you as much if you have wireless connection or one with constant jitter, it's a fact.
Got any other facts? Careful, so a constant jitter of 0 ms will mean bullets will not hit you as much?
smoooth wrote:In addition the opponent is at a disadvantage because your tank location is inaccurate, you get pauses on movements, blinking tank movements, and unpredictable hitzones.
Have you got any consideration for whether jitter has any negative impacts on the 'victim' client?
smoooth wrote:The problem with the game is that there is no positive re-enforcement to have a good connection.
Still sounds like a rant. So are you going to host 100 servers with limits of 100 +- 5 ms lag?
smoooth wrote:The other aspect of lag that is overlooked is that when people are laggy or jittery it opens up the opportunity to cheat without being detected.
Overlooked?
smoooth wrote:Often times I will become used to the lag of a certain player and then magically on one shot, the lag is off. Was this the result of a lag spike? or did they "do" something to manipulate their lag? It's very hard to tell.
How do you suppose someone can control lag to such a degree that they can make things happen to within 1 second of their wish? If this 1 second range doesn't suit your fancy, how often are people being shot at? Can people tell ahead of 3.5 seconds (reload time) that they will need a particular dodge? Has anything or anyone been known in association with a lag tool?
smoooth wrote:Often times I think people mean "Fluctuating Lag" when they say "Lag"
If sometimes you can't be precisely sure of an event, how do you know whether your miss protest is valid? Maybe these are just numbers to some people, a cue; if someone spoofed constant lag 400 ms as 400 +- 20 ms, they could easily get complaints when someone 'realizes' they had 'jitter'.
smoooth wrote:When a player has 100-200ms lag it becomes very easy for players to hit F5 or have other cheats that add delay to prevent a tank from dying. The reason it's easier to get away with is because it's a lot harder to guage where the tank actually is, which means it can be manipulated easier without detection.
Now I have something to ask, a question encouraged for anyone who may know:
So I'm assuming the F5 pause notification can be disabled. Would anyone like to offer scenarios in which they think F5 would benefit the user? Please be specific and use realistic, detailed examples (but not about wall-walking).
If you can, elaborate on the correlation between "add delay" and "to prevent a tank from dying". What's this got to do with lag, or is this just another thing you want to say about 'cheaters'?
User avatar
llrr
Private First Class
Private First Class
Posts: 129
Joined: Tue Jul 03, 2007 7:00 am

Re: "Demystifying Lag" not accurate

Post by llrr »

Lol.
User avatar
Cobra_Fast
Dev Monkey
Dev Monkey
Posts: 322
Joined: Sat Oct 07, 2006 11:51 am
Location: Germany
Contact:

Re: "Demystifying Lag" not accurate

Post by Cobra_Fast »

llrr wrote:Lol.
100% agreed.
User avatar
Green Manalishi
Private First Class
Private First Class
Posts: 593
Joined: Thu Nov 18, 2004 5:54 pm
Location: Effingham

Re: "Demystifying Lag" not accurate

Post by Green Manalishi »

looool@all the heated, elaborate discussions evolving around a simple, funny game about tanks jumping, flying, shooting colorful bullets at each other..
User avatar
Knox
Private First Class
Private First Class
Posts: 158
Joined: Tue Jun 12, 2007 12:40 pm

Re: "Demystifying Lag" not accurate

Post by Knox »

Green Manalishi wrote:looool@all the heated, elaborate discussions evolving around a simple, funny game about tanks jumping, flying, shooting colorful bullets at each other..
<3
User avatar
Mopar Madness
Private First Class
Private First Class
Posts: 169
Joined: Mon Jul 03, 2006 3:31 am

Re: "Demystifying Lag" not accurate

Post by Mopar Madness »

smoooth wrote:When a player has 100-200ms lag it becomes very easy for players to hit F5 or have other cheats that add delay to prevent a tank from dying. The reason it's easier to get away with is because it's a lot harder to guage where the tank actually is, which means it can be manipulated easier without detection.
Didn't Moonpix also show a way to calculate exactly where a bullet was in relation to the real place of a lagger using an equation where lag times a constant could be applied to trailing shot lines on radar? Provided the lag numbers were accurate, that equation should make it obvious if a cheat was used. That leaves lag fluctuations as your only point, and that also according to Moon's lag theory should be a 2 way street, the more your lag fluctuates, the more the total lag fluctuates, the harder time you have hitting people. I think the league officials are vary capable of finding cheats and exercise a lot of techniques to find cheaters. More reliable techniques than eyeballing it, guesstimating lag and shouting cheat every time something seems funny. In baseball or soccer if a coach or player fights with the ref/umpire or makes game calls, the ref/umpire is likely to card him or throw him outta the game. GULeague refs should be treated the same way. They know what they're doing. They ain't dummies.
User avatar
Zac
Private First Class
Private First Class
Posts: 359
Joined: Fri Oct 20, 2006 11:59 pm

Re: "Demystifying Lag" not accurate

Post by Zac »

"demystifying lag" accurate. everything taz wrote is more or less true. (put that there because you can't just claim it isnt accurate because you dont agree with it :P ). essentially smoooth, your argument (in my opinion) is a lost cause. there are just to many people who dont agree with you. i say, give up on this...play the game. and if you dont like lag or jitter, dont play it. (btw 2-4 jit is not going to make a significant difference in a match) the more you think about this, the more you'll convince yourself that something is wrong, when really. it isnt. :wink:

relax :turtle:
User avatar
Zac
Private First Class
Private First Class
Posts: 359
Joined: Fri Oct 20, 2006 11:59 pm

Re: "Demystifying Lag" not accurate

Post by Zac »

oh yeah, and btw, i thought this was a topic discussion on demystifying lag?? getting something off topic here :)
An SQUERRILz
Private First Class
Private First Class
Posts: 91
Joined: Wed Apr 25, 2007 2:08 am

Re: "Demystifying Lag" not accurate

Post by An SQUERRILz »

Mopar Madness wrote:Didn't Moonpix also show a way to calculate exactly where a bullet was in relation to the real place of a lagger using an equation where lag times a constant could be applied to trailing shot lines on radar? Provided the lag numbers were accurate, that equation should make it obvious if a cheat was used. That leaves lag fluctuations as your only point, and that also according to Moon's lag theory should be a 2 way street, the more your lag fluctuates, the more the total lag fluctuates, the harder time you have hitting people. I think the league officials are vary capable of finding cheats and exercise a lot of techniques to find cheaters. More reliable techniques than eyeballing it, guesstimating lag and shouting cheat every time something seems funny. In baseball or soccer if a coach or player fights with the ref/umpire or makes game calls, the ref/umpire is likely to card him or throw him outta the game. GULeague refs should be treated the same way. They know what they're doing. They ain't dummies.
In short, the method isn't accurate, even if it was the detection can't be incorporated into the client easily, lastly the admins are relatively powerless (either that or hopeless) when it comes to subtle cheats.
Zac (afk) wrote:"demystifying lag" accurate. everything taz wrote is more or less true. (put that there because you can't just claim it isnt accurate because you dont agree with it :P )."
There are numerous inaccuracies in the tutorial (believe me my critique is more thorough and unrelated to smoooth's). Everything is more or less true. [You can't just claim 2 = 3 is not accurate because you don't agree with it.] Are you in want of an education?
snick
Private First Class
Private First Class
Posts: 95
Joined: Thu Jun 02, 2005 2:26 pm

Re: "Demystifying Lag" not accurate

Post by snick »

Some aspects of the document are roughly accurate. Some aspects, however, are not.
I'm not going to run through a huge list, but I'll note in passing that this is inaccurate:
Lagstats are very reliable as they simply tell the delay of the gaming packets
related to the recent player movements/shots [...] lagstats can be taken into
account as a solid reference for lag reading and compensation.
BZFlag's lagstats are terribly flawed. They can take minutes to converge on anything
close to the actual lag values, they can behave erratically and unpredictably in the
presence of packet loss (sometimes packet loss doesn't show up as such, but shows up
as high lag instead), they can mask jitter spikes and there is no indication of lag asymmtery.

Etc.

I know this because I have accurately simulated lag on a locally hosted server so I
know what the true lag is and what the figures show. And very often the lagstat
figures are very wrong.
User avatar
Cobra_Fast
Dev Monkey
Dev Monkey
Posts: 322
Joined: Sat Oct 07, 2006 11:51 am
Location: Germany
Contact:

Re: "Demystifying Lag" not accurate

Post by Cobra_Fast »

snick wrote:
Lagstats are very reliable as they simply tell the delay of the gaming packets
related to the recent player movements/shots [...] lagstats can be taken into
account as a solid reference for lag reading and compensation.
BZFlag's lagstats are terribly flawed. They can take minutes to converge on anything
close to the actual lag values, they can behave erratically and unpredictably in the
presence of packet loss (sometimes packet loss doesn't show up as such, but shows up
as high lag instead), they can mask jitter spikes and there is no indication of lag asymmtery.
BZFlag's lagstats are NOT terribly flawed. They work perfectly. But most players do not understand that they can only measure lag and packetloss if there is data sent. People are pausing matches and stop any actions and remain sitting there for hours to see their lag go down what won't happen because they are not producing any traffic what could be used for lag measuring. BZFlag does not send packages on a regular basis (like every 100 ms), it only sends data if there is new data to be sent, so the player has to produce some traffic to have his new lag measured.
snick
Private First Class
Private First Class
Posts: 95
Joined: Thu Jun 02, 2005 2:26 pm

Re: "Demystifying Lag" not accurate

Post by snick »

Cobra_Fast wrote: BZFlag's lagstats are NOT terribly flawed. They work perfectly. But most players do not understand that they can only measure lag and packetloss if there is data sent. People are pausing matches and stop any actions and remain sitting there for hours to see their lag go down what won't happen because they are not producing any traffic what could be used for lag measuring. BZFlag does not send packages on a regular basis (like every 100 ms), it only sends data if there is new data to be sent, so the player has to produce some traffic to have his new lag measured.
No, you do not understand how the lagstats are calculated. They are not calculated
from movement and shooting data. The calculation is based on special "LagPing" messages
that are sent out at regular intervals. The result is used to calculate an exponential moving
average and it is that average that is shown in the lagstats. But the average moves so slowly
that changes in lag can take minutes to register fully in the stats.
Post Reply