1 15 16 17 18 19 Previous Next 580 Replies Latest reply: May 16, 2013 5:14 AM by deamonomic Go to original post RSS
  • 160. Re: LagComp Video Discussion: BlackOps2 Edition
    ovenbakedmuffin

    Staticon wrote:

     

    In MoeJoeWins's split-screen delay test video, it's very easy to see a slight delay in the ammo counter going down upon a shot firing (the usual 133ms delay); even in realtime, this delay is noticeable to me. However, in DenkOp's video where he tests his latency via the ammo counter trick--in an actual networked game no less--the delay in the ammo counter going down is miniscule, perhaps even less than the slight 133ms delay we observe in all our local LAN/splitscreen tests. How is it possible that an actual game running on the internet, assuming DenkOps didn't have host and assuming that his latency was somewere in the ballpark of at least 20ms, that the delay in DenkOps ammo counter going down was seemingly even less than the supposedly constant 133ms of delay in every situation?

     

    Perhaps my eyes are deceiving me, and I don't have the tools or know-how to slow down the video and accurately mesasure DenkOps's actual delay, but it does seem like it's less than 133ms to. Fwiw, I feel confident that I can perceive 133ms of delay, and I can most definitely perceive 166ms or more delay. I've been suffering with the ~200ms of input latency present in the PS3 version of Battlefield 3 for nearly a year, so I have some frame of reference as to how to gauge a certain amount of delay. Judging delay or latency only becomes difficult for me when it is in the 50-75ms or less range.

    I'll quickly check this for you using some recorded footage.

     

    //Tested once

    Player vs Host:

      0ms ping

    • Ammo Count: 184ms
    • RedScreen: 184ms      

      800ms ping

    • Ammo Count: 867ms
    • RedScreen: 967ms

     

    Host vs Player:

      0ms ping

    • Ammo Count: 33ms
    • RedScreen: 0ms

     

    Player vs Player:

      400ms ping

    • Ammo Count: 651ms
    • RedScreen: 617ms

     

    As you can see there's a difference of up to 100ms / 6 frames in this single-sample test. GUI calls occur at a different rate than game states are sent by the host. It's not an important game mechanic and therefore not significant. Ammo count remains a valuable tool to compare your delay, but it won't be exact.

  • 161. Re: LagComp Video Discussion: BlackOps2 Edition
    playerone2

    You say only the developers have access to the networking code?

     

    Are the variables in the "lag comp" system determened and "corrected" at the servers. or do you think the server just sends the data or the amount of latency each player has to the console. and the actual amount of "compensation" to add is in the game itself.

     

    I may be way off with that. but if thats the case havent people been hacking the games and modifying the code for ever.

    --Couldnt someone figure out which lines of code decide how much to add or delay. and change it all to zero and start from there?

    Or is that way different then just bringing up the hitbox or changing the amount of aim assist or whatever they do to make aimbot wallhack and whatever other mods they do.

     

    Just curious.

     

    BTW, I think your video may have actually made me a better player.

  • 162. Re: LagComp Video Discussion: BlackOps2 Edition
    nuttin2say

    When you look at the above situations, can you also look at player performance leading up to the particular instances?

     

    Because what I see in the OP video I see in-game all the time ... and I can predict with some accuracy when it is going to happen and when it happens has nothing to do with how many bars each player has in the lobby.

  • 163. Re: LagComp Video Discussion: BlackOps2 Edition
    ovenbakedmuffin

    playerone2 wrote:

     

    You say only the developers have access to the networking code?

    ...

    That's correct, only devs. If they were to leak it that would change, of course, but who would risk their career when absolutely no one would pay $5 for BO2's code?

     

     

    playerone2 wrote:

    ...

     

    Are the variables in the "lag comp" system determened and "corrected" at the servers. or do you think the server just sends the data or the amount of latency each player has to the console. and the actual amount of "compensation" to add is in the game itself.

     

    I may be way off with that. but if thats the case havent people been hacking the games and modifying the code for ever.

    --Couldnt someone figure out which lines of code decide how much to add or delay. and change it all to zero and start from there?

    Or is that way different then just bringing up the hitbox or changing the amount of aim assist or whatever they do to make aimbot wallhack and whatever other mods they do.

     

    Just curious.

     

    BTW, I think your video may have actually made me a better player.

    That's not really what happens.

    There are many ways to do it, and while I could describe them all, it would take a rather long time. I doubt I could explain it fully in layman's terms. Regardless, lag comp is code. It would take an incredible amount of time to reverse engineer. And being that it is complete garbage, no one would do it. Without access to the actual code, nothing is certain apart from reproducible data.

     

    Aimbots/wallhacks/etc are often directx overlays. That's one of the simplest and least detectable forms, anyways. They all share one quality; they attach to the program and change what is displayed on your screen. They do not change the game's code, they merely hook into the program and run alongside it. Changing the code is far too easy to detect.

     

    Referring to the hacked lobbies in MW2 for example, those are changes to variables such as gravity, done on the host. Players cannot do that. Such variables can be modified, but to change entire sections of code to the extent needed to hack via lagcomp is near impossible.

  • 164. Re: LagComp Video Discussion: BlackOps2 Edition
    nuttin2say

    Forgive me here ovenbakedmuffin, apparently I'm seeing something that is being overlooked here. Well, not entirely but it's not really being discussed. I've read through a third of this thread and it is touched upon, but not discussed.

     

    Is the lag compensation that is gone over in the OP video dynamic in nature or static, assuming the connections to the host are static?

     

    If you've got different gamers looking at the same issue and coming up with similar activity but different measurements, couldn't it be they are all correct because the process studied is dynamic?

  • 165. Re: LagComp Video Discussion: BlackOps2 Edition
    ovenbakedmuffin

    nuttin2say wrote:

     

    Forgive me here ovenbakedmuffin, apparently I'm seeing something that is being overlooked here. Well, not entirely but it's not really being discussed. I've read through a third of this thread and it is touched upon, but not discussed.

     

    Is the lag compensation that is gone over in the OP video dynamic in nature or static, assuming the connections to the host are static?

     

    If you've got different gamers looking at the same issue and coming up with similar activity but different measurements, couldn't it be they are all correct because the process studied is dynamic?

    If BlackOps2 uses a form of lockstep networking (which is a fair assumption due to synchronized player deaths / actions), then the most lagged player determines the delay in many situations. That is one explaination for inconsistency, unless you find yourself playing with exactly the same players at all times.

     

    As for a dynamic lagcomp based on a variable, such as the 0.8-1.3 K/D issue being discussed, I have not overlooked it. I simply cannot (reliably) test for it. Consider the difficulty in testing something that affects each console individually, or each region of the world separately. My viewpoint is that it is quite unlikely they would go through all the trouble for no reason.

    >Speaking directly to the K/D variable, my 0.0 K/D alt shows the same results as my 1.54 K/D main.

     

    One thing is certain, however; the lagcomp is not fully dynamic. My results have been within 1~4 frames of one another for their respective ping ranges over the 2+ week period required to gather them.

  • 166. Re: LagComp Video Discussion: BlackOps2 Edition
    Jamaro85

    Not exactly on topic, but I'd like to see a bunch of people who have recorders go up against one another in a free for all game, regardless of skill.  Share the videos with one another to compare what the different perspectives were showing live in the game during heads up encounters.  I think that could make for some interesting footage.

  • 167. Re: LagComp Video Discussion: BlackOps2 Edition
    nuttin2say

    Thanks for being understanding in regards to bringing me up to speed with where the discussion has been. I could have gone through pages 7 through 15 of the discussion, but within pages 1-6 and 16-17, what I mentioned was only touched upon once but there are more thankyou posts than I cared to go through. Not that you don't deserve them and not to be a prick, but I want to get to the heart of the issue.

     

    In any case, the fact that the kdr variable is within the discussion raises a red flag to me. It means that I'm not the only one questioning when the lag comp issue comes into play.

     

    And that question is play because to me, a "lag free environment" is within a single console, not necessarily on a LAN. Did you observe these disparities within split screen on a single, unconnected console?

     

    edit - Oh, btw, there is no doubt in my mind that the lag comp variable is not "fully" dynamic. What I'm driving at here is that it only has to have a small amount of static - a very small amount - in order to be effective. If it is small enough, 99% of the community will never catch it. But players that are online 500+ hours a year playing video games will. In the mean time, and I dare say this, it brings a great sense of "fun" to that 99% because all of a sudden they can take on the big dogs and the game "feels" balanced to them.

     

    edit 2 - In fact, it doesn't have to be static at all. And it does not have to be based upon kdr necessarily, either. It could be coded trigger that reacts to ppm, kdr, total kills, total points, wlr - any number of single triggers or combination of triggers.

     

    I would call anyone that suggested this a complete nut job were it not for several things. First of all, BO2 is using a new engine and new code. That's per Ghandi's comments in the MW3 title update discussion thread. That's one thing. Second, people were screaming to have the host be given lag advantage for years and the logic behind the "community" demands seemed reasonable. The host has "paid more" for his quality connection; the host having the advantage assures that only one player in the lobby has an "unfair" advantage instead of multiple people, etc, etc, blah, blah. I can too easily see ATVI having TA and IW make such an adjustment with the caveat that there be a trigger within the system that prevented host from totally devastating lobby after lobby. What you are showing in that OP video fits the bill to a T.

     

    First of all, it becomes a relationship, as you say, between the host and the worst ping in the lobby. That gives the host an advantage against half the lobby and a disadvantage against the remaining 5 players. Teams being evenly divided, the game is "balanced." At the same time, ATVI can legitimately say that the host has the advantage. As an insurance policy, trigger lag if the host begins to dominate. Boom, I've got a game 99% of the community believes is balanced when, in fact, this would be the most atrocious thing to have ever happened in the gaming world.

  • 168. Re: LagComp Video Discussion: BlackOps2 Edition
    ovenbakedmuffin

    nuttin2say wrote:

    ...

    In any case, the fact that the kdr variable is within the discussion raises a red flag to me. It means that I'm not the only one questioning when the lag comp issue comes into play.

    ...

    I may have said this earlier (writing a lot of replies):

    I believe the .8-1.3 k/d is related to the fact that whichever player is moving from cover has the advantage. No one player will ever have this advantage 100% of the time unless every other player does not move for the entire match. In actual gameplay, you may find that you have the advantage 50% of the time. Even a bad player can get a k/d of 0.8 if they have a 200-300ms+ advantage half the time.

     

     

    nuttin2say wrote:

     

    ...

    And that question is play because to me, a "lag free environment" is within a single console, not necessarily on a LAN. Did you observe these disparities within split screen on a single, unconnected console?

    ...

    Another member posted a video showing splitscreen lag on PS3. I have confirmed on xbox. It is within the same range as online play, suggesting they simply didn't "remove" lagcomp code for local matches.

     

    As for the conspiracy theories, I think you are overestimating the ability of this dev team in regards to networking. They might want to create something so complex as that, but can they? Even I would go insane thinking of the number of possible situations to code for. It's much simpler to just make lagcomp work properly in the first place. Consider that, to give the host an advantage, all it takes is the very standard networking we see in nearly every FPS today. To remove that advantage if a host is dominating, they can migrate hosts when the game ends, as soon as they find a suitable replacement host. With hosts obviously having an advantage, more players purchase better connections, meaning more hosts. You end up with COD4 networking, and you've done <30mins work.

     

    Aside:

    They would lose the killcams and theater mode if they did what I just suggested. It would require a timeframe of about a day to get them fully up and running again.

     

     

    Jamaro85 wrote:

     

    Not exactly on topic, but I'd like to see a bunch of people who have recorders go up against one another in a free for all game, regardless of skill.  Share the videos with one another to compare what the different perspectives were showing live in the game during heads up encounters.  I think that could make for some interesting footage.

    It could be fun. 'Might not provide any actual data, but you're right, it could make for some interesting footage and new perspectives. If something did appear out of the ordinary, it could be used for the basis of a new line of thought and targetted testing. Unfortunately, while we're on the topic of conspiracy, I could not be sure these players did not manipulate their connection during the match.

  • 169. Re: LagComp Video Discussion: BlackOps2 Edition
    Nubdub

    Why would they lose killcams and theater by removing host lag comp?

1 15 16 17 18 19 Previous Next