You'd still get LAN pings though if that's what you're asking.

---

Added on November 12, 2009, 2:34 pmWhen I have time I'll record some gameplay of sniping with ~400ms on perpendicularly moving targets to illustrate the point.

And yes, like you said, it's very susceptible to hacks, except it's much easier to code an aimbot that just shoots at targets directly on your screen using the game client than it is to mess with the packet stream.

Regarding other player's positions - what you see on your screen is a snapshot that the server had at a certain point in time. This is key... this means that other than your own (and only your own) position client side info is identical to what server side info was at a certain point in time (<your latency>ms ago) and is not wildly different (like WoW ahahahahah)

This post has been edited by Falk: Nov 12 2009, 02:36 PM

would this mean something like

onscreen:

X <---enemy
|
|
|
O <-- you

reality:

X <---enemy
|
|
|
______O <-- you

This post has been edited by MYNAMEISJASON: Nov 12 2009, 02:48 PM

Hmmm, ok let me try to illustrate a scenario to show you my doubts about this (I'll be using very general approximations, so please don't butcher me on the timestamps):
Player A has a ping of 10ms
Player B has a ping of 100ms
Both are connected to server X

Timeline:
Time 0:
A is at location x1,y1
B is looking at x1,y1 (Effectively A)
X takes snapshot of A @ x1,y1 and also B looking at x1,1
Since A has a lower ping, his data reaches X far earlier before B.

Time 1:
A moves to location x2,y2
B fires at position x1,y1 (A)
X is waiting on response from both clients

Time 2:
A's information reaches X first and it updates A's position to location x1,y1
B's fire command at x1,y1 is still enroute to X

Time 3:
A is at location x1,y1
B's fire command reaches X and updates that B is firing on x1,y1


Ok, here are the issues with this:
1) If it is very client side oriented as you say, which "snapshot" should it take? A or B? There's no way that it can prioritize client snapshots because something like this would happen. If it takes A's, then that's roughly how lag prediction works now where it favours the lower ping clients. If it takes Bs, then A is obviously at a disadvantage since though he has little lag, his actions can be overridden due to another's time delay.
2) How does the server synchronize the time? With this, it seems that the server's own timestamps are very reliant on the slowest client in the game, which makes it doubly unfair to everyone else.


This is just my very rough understanding of how latency is handled in most games (And I'm not saying you're wrong), I'm just curious about the whole thing as it seems very different from what I used to know.

is it like 2 guys in CSS line up, aim the head, and shoot at the exact same moment, and 1 guy wins cus he has lower ping?

Er, did you mean A is at x2,y2 for Time 2 and Time 3?

I think H@H@ made a mistake there too

Ah yes, my bad... but you get my point right?

http://www.mediafire.com/file/njdtmovzmiz/snipeantilag.mp4

This was recorded at 30fps. It's a clip of a headshot on a perpendicularly moving target, then played again at 1/4 speed.

http://img402.imageshack.us/img402/5862/frame48.jpg
Frame 48 of the clip, when I pull the trigger. Note how I'm firing directly on target, not leading ahead to compensate for latency

http://img94.imageshack.us/img94/74/frame55.jpg
Frame 55, the exact moment I'm awarded the kill. Incidentally we can infer that 7 frames at 30fps --> my ping is ~210ms (lol)

http://img25.imageshack.us/img25/3571/frame41.jpg
To calculate the distance target moved in 7 frames, I rewinded to frame 41 (48 - 7 = 41) and took this screenshot

http://img22.imageshack.us/img22/1699/comparisonl.jpg
And here I lined up Frame 41 (orange) with 48 so that we can see the distance the target moves in 7 frames. If we assume he continued moving along that locus, we can approximately infer that the location his head is when I was awarded the kill at frame 55 with the blue circle.

Yes, it's unfair. If my ping was 400, he'd have moved twice that distance before dying. Even if he'd gotten behind cover.

Just so I'm clear, exactly when you pulled the trigger was there a firing sound or only when you were awarded the kill?

I agree with Falk on how the lag compensation is done, its the same as Killzone 2, back in MW1, host only gets the info on when and where you shot, processing the hit on hostside, where as in MW2 players send both shot and hit info to the host. But this can possibly create another problem, where 2 lagging players are shooting each other and in each of their pcs it registers as their opponent dies. I've not experienced this yet in MW2 but in Killzone 2 many times it has happened that both me and my opponent dies.

Yeah that's my fear as well. Its incredibly broken and really makes hacking easier on the PC.

When I pulled the trigger. Sorry there's no sound, I CBA'ed because it takes ages to set it up. You can also see the bullet tracer (those yellowish lines) in frame 48, partially obscured by the bottom part of the scope reticle. That's how it's been always - it goes BLAM, and then a split second later you get the PAK sound and 'X' of a registered hit. That's also why in a lot of Promod CoD4 frag movies you see the +5 come out after different amounts of delay (nowhere near 210ms, obviously) from instantly on LANs to as the scope is zooming back out on online scrims on ded servers.

Uhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh it's EXACTLY the same as it was in MW1. That's how I picked up on it so fast. You think I researched MW2 net code in two days? :V

This post has been edited by Falk: Nov 12 2009, 03:45 PM

Are you sure its the same? I played on Aussie servers with MW and I remember having to do a lot of leading with my shots to get them to hit.

This post has been edited by H@H@: Nov 12 2009, 03:46 PM

Could possibly be different on Xbox, i didn't play the PC MW1 online before, but on the Xbox MW1, if it lags you have to shoot a where you think the enemy is moving to, and not right on the target.

I am one hundred percent sure it's exactly the same. I can go do a fraps of the same thing in CoD4 if you want.

The beauty of the antilag is that it's so unnoticable that along with placebo effect you don't really notice it unless it's pointed out to you (like what I've done - I doubt regular users would even notice the delays on shots at full speed like the video I uploaded)

Also, regardless of ping it's always better to lead targets anyway just due to human psychovisuals and motor reaction times. That may contribute to what you were experiencing.

edit: As an addendum, ACC season 1 had plenty of matches where 1 team experienced pings of 100ms throughout. (They were usually played on home-away configuration for fairness for the first two maps in these cases). You have no idea how much 100ms actually would affect snipers and long range assault shots if this feature wasn't present in CoD4 (Perpendicular motion would be about as much, if not more, as my clip if they were sprinting). However the main disadvantage was face-to-face CQB reaction times, when both parties opened fire at the same time. Firing on an unaware target merely resulted in delays on hitreg, no extra leading required.

This post has been edited by Falk: Nov 12 2009, 03:55 PM

When I say lead, I'm talking "Playing UT on 300ms servers" kind of leading. I wasn't born yesterday, I am well aware of the difference.

Sorry then, didn't mean to insult your intelligence.

---

Added on November 12, 2009, 4:00 pmfff I just realized my maths is horrid. 7 frames at 30fps isn't 210ms, it's 7 x 1000/30 = ~233.33ms.

Not that it really matters... heh.

This post has been edited by Falk: Nov 12 2009, 04:00 PM

Its ok, just wanted to get a rough idea of how the lag issue is handled.

If the community remains strong and that one friend of mine isn't dissuaded, I might pick this up myself.

It's still a very strong pub game despite my gripes about the competitive/hardcore side of things, and probably will get better when better connectivity starts overriding the shitfest that's IW.net. That doesn't mean I agree with IW's decisions on a good many things but there you have it.

The point I'd like to convey though is that the antilag and client prediction is as far as I've seen identical to CoD4 (and yus, I used to pub on GamingSA as well!). This contributes to a lot of the guns feeling very, very similar - I didn't suddenly become jesus at the game in two days. So don't let "omfg haxor antilag" be one of the cons if you're comparing it to the previous game. Antilag along with the higher pings on casual netplay, though... maybe.

This post has been edited by Falk: Nov 12 2009, 04:13 PM

Wow, that was a falking awesome read. Gonna read through it onces again latter after work.

thanks Falk.