LagComp Video Discussion: BlackOps2 Edition

Call of Duty Black Ops II Xbox 360

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to asianman12

asianman12 wrote:

Then how does it explain my experience then? I'm using the most common Cable providers in the country to play Call of duty.  I get host sometimes but not all the time since the people I play with usually have much better upload bandwidth.

Aside from being host.  What am I doing right?

You always pull out the same comments, "Why don't I have a problem? or What am I doing right?"  like that will dismiss that fact that there is an issue; Rather than asking yourself : "Why don't I have a problem?"  It might be better to ask why others do, which is the issue at hand. 

For you to be able to understand anything related to any issues with this game you unfortunately have to leave all the fanboyism aside.

Level 2
Likes: 8
Posts: 28
Registered: ‎05-12-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to m4xburrito

Not being a fanboy.  I am going about it this way.  Obviously I am immune to the problems in some way. Or at least not as vulnerable to it as of now.  Think of it like a virus.  I'm immune to the disease so shouldn't people be looking at me to see why I'm not having the problem?  I probably have the cure to the disease and never knew it.

Ever think about it that way?

Level 25
Likes: 281
Posts: 2313
Registered: ‎14-07-2011

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to Nubdub

Nubdub wrote:

As a computer repairman I see this often, DSL is inferior in every way to cable and fiber broadband. Almost always bad pings and get half the speed of what you pay for. DSL is ancient technology compared to other broadband types. This would explain the "good" host and client he gets which is actually causing other players to get over comp'ed giving him the preception that there is not a problem.

The main hurdle for a DSL line is distance from the exchange, after that it's interference the copper lines are subjected to and their general condition and so on. You know you have issues if you analyse the modem data and find you are losing sync with the exchange and/or generated thousands, even millions of FEC/CRC/HRC errors every hour as well as loss of signal, framing and so on. Presumably your aware of such factors? I fortunately do not suffer from any of this. The rest is down to the ISP quality and your internal network.

Also you seem overly focused on ping rather than the more important are of stability, essentially variance in the line ala jitter and packet dropping. I'll use pingtest to provide a general insight into my apparently "inferior" line. If the line is indeed crap as Nubdub suggests then results from simple line tests should reflect this:

Pinging from UK to almost 7000miles away to Singapore I get the following:

http://www.pingtest.net/result/75247490.png

Zero jitter and zero packet loss. Of course i'll never get matched that far away but it shows the line quality/stability over a rediculous distance. If I  pick closer to home (I'm in the UK) but still... lets make it to the Netherlands across the sea rather than mainland UK:

http://www.pingtest.net/result/75247651.png

Predictably I get zero jitter and zero packet loss and a ping that is well within "gaming range" even though it is still overseas. I generally get matched with UK gamers but also occasionally German, Netherlands and French but mainly UK players. Hopefully this highlights your presumptions regarding my experience and the line quality, both in ping and the more important aspects you failed to mention - jitter and packet loss, is wrong.

There is clearly a problem for many here, I am posting my setup and experiences with BO2 to hopefully pitch in and shed light on the matter. I suffered severe problems with MW3, I have seen the gloater posts from back then and have been there, it sucks! Do not make such sweeping assumptions Nubdub, just ask.  

Level 7
Likes: 55
Posts: 233
Registered: ‎11-11-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to RatedM2

I'd still like to see a local pingtest if you'd be so kind. Maybe I'm way off about adsl or cable connections having more latency hence more problems.

http://www.pingtest.net/result/75253634.png

http://www.pingtest.net/result/75253594.png

http://www.pingtest.net/result/75253413.png

http://www.pingtest.net/result/75253542.png

http://www.pingtest.net/result/75252269.png

Level 16
Likes: 127
Posts: 833
Registered: ‎21-09-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to Ss78_

Ss78_ wrote:

You might want to take your name out of the link.

So you're using adsl with low upload. Do you even get host with 1.5up? I know that BE is a good provider but with such a low upload would that not mean your experience is mainly as a client. My client experience is far better than my host experience but because of my location in NL and my 50meg fibre upload it's not often I will be client hence my experience is mainly bad with the matchmaking also playing a big part in this too.

The name isn't mine, I did notice that though.

I am mainly a client player yes, although I get host around 30% of the time. I know exactly when I'm hosting whether it's the start, during (migration) or end of a game because I can check my phones browser (connected to the router) and it shows host data transfer rates on-the-fly of circa 350kbps both up and down - this confirms my status as host. I do not verify it by quitting games ala host migration screen, as this could simply be that I'm being outplayed. Normal client play indicates circa 30kbps transfer rates up and down.

I cannot say why your line gives you a bad hosting experience, I can only give you info on mine but matchmaking may play a part for your location, perhaps hardware (firewall, conflicts), nature of ISP (hops etc). A 1.5meg upload is 3 times the required amount to host. I suggest capping your upload to 60kpbs (to account for spikes) as this will mean you won't be host. Hopefully your router supports it, if it doesn't deffo consider investing in one that does. 

Level 7
Likes: 55
Posts: 233
Registered: ‎11-11-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to RatedM2

RatedM wrote:

Ss78_ wrote:

You might want to take your name out of the link.

So you're using adsl with low upload. Do you even get host with 1.5up? I know that BE is a good provider but with such a low upload would that not mean your experience is mainly as a client. My client experience is far better than my host experience but because of my location in NL and my 50meg fibre upload it's not often I will be client hence my experience is mainly bad with the matchmaking also playing a big part in this too.

The name isn't mine, I did notice that though.

I am mainly a client player yes, although I get host around 30% of the time. I know exactly when I'm hosting whether it's the start, during (migration) or end of a game because I can check my phones browser (connected to the router) and it shows host data transfer rates on-the-fly of circa 350kbps both up and down - this confirms my status as host. I do not verify it by quitting games ala host migration screen, as this could simply be that I'm being outplayed. Normal client play indicates circa 30kbps transfer rates up and down.

I cannot say why your line gives you a bad hosting experience, I can only give you info on mine but matchmaking may play a part for your location, perhaps hardware (firewall, conflicts), nature of ISP (hops etc). A 1.5meg upload is 3 times the required amount to host. I suggest capping your upload to 60kpbs (to account for spikes) as this will mean you won't be host. Hopefully your router supports it, if it doesn't deffo consider investing in one that does. 

Again if it's correct that it's worked off averages since you have adsl your line is average and the oos would be minimal for you. You also might have good matchmaking too.

I'm looking into my matchmaking results first before I resort to the qos no host solution.

Level 16
Likes: 127
Posts: 833
Registered: ‎21-09-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to RatedM2

RatedM wrote:

Agreed the connection related stuff is vauge enough  that it required me to watch it a few times to relate what he was saying to the tables. ...

To clarify the charts, they show the delays that player/host experiences. Both for inflicting damage, and for the disparity between the player's actual position and what the player sees when running to/from cover. For the host, the ping would be the ping of the player the host is in combat with.

Just as with everything in lagcomp, this increases in complexity the more we consider. For example, the PvP situation shows a player at X ping versus a player at 75ms ping. The results would be different if it were versus a player with 200ms ping. I simply don't have time for another 300 tests at every ping. Also note that the host may inflict damage immediately, behind the scenes, as with every other fps (except MW3). Or it may not. There is no effective way to test that.

RatedM wrote:

...

From what I gather he is essesntially saying that:

  1. Being host owns.
  2. Having a properly optimised connection works in your favour as both client and host.
  3. Running out of cover is advantageous, running to cover is not.
  4. As long as the matchmaking does it's job by including players with closely matched pings then the game should dish out a roughly equal experience for all (bar host who maintains an advantage). The tradeoff is that this ideal scenario feels like a MW2 "3 bar" experience at best.

If I am wrong in the short summary above perhaps the OP could clarify. ...

2) Having an optimised connection works in your favour as client. Not as host. The greater delay the players have, whether because of them or because of the host, the greater advantage the host has by comparison.

As for the point about host, note that the host does not have the large delay when running to / out of cover. Consider the host experience completely different than the player. Its advantage is purely related to when 'bullets' hit the target - about 150-250ms earlier. That also means the kickback will affect their aim 150-250ms earlier.

However, player deaths are synchronized across all players, meaning they will not actually "die" until both the host's console/pc has decided they are dead, and the player with the worst connection in the lobby is able to receive that information. This can result in the experience of taking more ammo than usual to get kills, and appear to be lag for the host.

From a programmer's standpoint, lagcomp is the core of an online FPS. While each individual aspect of lagcomp is exceedingly simple to comprehend, it is the vast effect it has which makes it complex. There is absolutely no margin for error.

RatedM wrote:

...

Points 1,2 and 3 are rather obvious. Point 4 is interesting. I wonder what the "Good connections get screwed!" and "Being host gets you screwed!" brigades think about this video?

I was part of that "Good connections get screwed!" and "Being host gets you screwed!" brigade, when it was true (MW3). It's simply not in BlackOps2. While there is a noticeable "advantage" for laggy players as to when, exactly, they see other players, part of my testing was to determine if Treyarch accounted for that. They did. And as 'bullets' are delayed by that same advantage, it means nothing compared to a less-laggy player.

Copypaste from video comments:

"There are also those individuals who claim they do not suffer from lagcomp. I won't pretend to know everything, as I'd need the source code for that. And if I had that, I'd fix it for them. But with that aside, I have no reason to believe that for these players it is anything other than simple aspects of FPS gameplay. Playstyle has a significant contribution, as a "good" camper simply won't be seen in time for lagcomp to matter. A difference in reaction time would also easily explain it."

As other forum members have mentioned, it is also possible the experience is adjusted based on factors related to other players. This is true of MW3, and it is possible BO2 may have adapted such calculations for their own purposes. It simply isn't as obvious, nor easily tested for as a result.

There's no denying the BO2 online experience is better than MW3 (at least as of August2012). But that's not exactly a selling point.

Level 6
Likes: 83
Posts: 105
Registered: ‎15-08-2011

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to ovenbakedmuffin


ovenbakedmuffin wrote:

Just as with everything in lagcomp, this increases in complexity the more we consider. For example, the PvP situation shows a player at X ping versus a player at 75ms ping. The results would be different if it were versus a player with 200ms ping. I simply don't have time for another 300 tests at every ping. Also note that the host may inflict damage immediately, behind the scenes, as with every other fps (except MW3). Or it may not. There is no effective way to test that.
   .
2) Having an optimised connection works in your favour as client. Not as host. The greater delay the players have, whether because of them or because of the host, the greater advantage the host has by comparison.
 
As for the point about host, note that the host does not have the large delay when running to / out of cover. Consider the host experience completely different than the player. Its advantage is purely related to when 'bullets' hit the target - about 150-250ms earlier. That also means the kickback will affect their aim 150-250ms earlier.
  
However, player deaths are synchronized across all players, meaning they will not actually "die" until both the host's console/pc has decided they are dead, and the player with the worst connection in the lobby is able to receive that information. This can result in the experience of taking more ammo than usual to get kills, and appear to be lag for the host.
  
From a programmer's standpoint, lagcomp is the core of an online FPS. While each individual aspect of lagcomp is exceedingly simple to comprehend, it is the vast effect it has which makes it complex. There is absolutely no margin for error.

I was part of that "Good connections get screwed!" and "Being host gets you screwed!" brigade, when it was true (MW3). It's simply not in BlackOps2. While there is a noticeable "advantage" for laggy players as to when, exactly, they see other players, part of my testing was to determine if Treyarch accounted for that. They did. And as 'bullets' are delayed by that same advantage, it means nothing compared to a less-laggy player.
  
Copypaste from video comments:
"There are also those individuals who claim they do not suffer from lagcomp. I won't pretend to know everything, as I'd need the source code for that. And if I had that, I'd fix it for them. But with that aside, I have no reason to believe that for these players it is anything other than simple aspects of FPS gameplay. Playstyle has a significant contribution, as a "good" camper simply won't be seen in time for lagcomp to matter. A difference in reaction time would also easily explain it."
  
As other forum members have mentioned, it is also possible the experience is adjusted based on factors related to other players. This is true of MW3, and it is possible BO2 may have adapted such calculations for their own purposes. It simply isn't as obvious, nor easily tested for as a result.
  
There's no denying the BO2 online experience is better than MW3 (at least as of August2012). But that's not exactly a selling point.


So you're saying that the host can hit someone 150-250 ms earlier but it only counts when everyone has received the update from the host?

Host sees player 2 x ms later and gets 6 hit markers+ kill
Player 2 sees host x ms earlier and gets 6 hit marker and kill
Player 12 with highest ping to host gets update from host that who died first?

Is this what you're saying? I also don't really get what you say at 2) you say client has the advantage but the host does by comparison?

Whatever the scenario 150-250ms sounds game breaking to me.

Level 16
Likes: 127
Posts: 833
Registered: ‎21-09-2012

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to ovenbakedmuffin

It almost seems like for low ping, high bandwidth connections, we should be choosing "any" for our search option. That would give us more of an opportunity to get in more games that are farther away creating an environment where we fit in to the "average" category.

Level 21
Likes: 252
Posts: 1360
Registered: ‎16-12-2011

Re: LagComp Video Discussion: BlackOps2 Edition

in reply to ovenbakedmuffin

The guy in the video talked non-stop for 20min, and at no point did I disagree with a word he said.

Level 8
Likes: 59
Posts: 350
Registered: ‎17-01-2012

Studios