[sarcasm]
You could also alternately, have server-side ping leveling. Meaning, all updates to all clients are determined based on the reply of packet times.
The server runs through the array of players reading input, once all are received, it sends the output. This wold logically make the game fair, but vary the gameplay as your input/output latency would vary depending who you play.
[/sarcasim]

Logically, antilag sounds reasonable. But, take my point of view, I'm mostly a netQuake player and have 13ms on the servers I play (I use QW to play on European servers from the us). QW has always felt to me like what everyone here complains about with antilag. Where as, I get 10x more mid air rocket shots in QW than in netQuake. I've always assumed that in QW when I press FIRE with any weapon it flags that packet, thus noting my velocity, angles, and weapon and tracelines instantly to an origin along my aim. Calculating the velocity over travel time for the rocket if it will collide with the predicted origin of any opponent.

This feature may make using the sg/ssg/lg more LITERAL but from the point of view of the attacked there will be a lot of "I didnt even see that shot"...

How do other modern games handle hitscan weapons and prediction? http://www.gamesurge.com/pc/interviews/netcode.shtml

The problem with most modern games like cod or bf is that the weapons itself have massive spray so single fights in the game are really, really random and I don't even know how the massive lag on the servers affects the equation. Good aim is of very little help in these games, it's based more on the strategy and tactics.

other games have less knockback from other peoples shots. how many other games allow you to juggle players with lightning guns? generally it doesn't matter so much if you're hit, you just lost a little health, it doesn't result in you getting hit the next shot 0.1 seconds later too.
other games interpolate instead of extrapolate other players, so have additional latency built into the engine, can be as much as 50ms which isn't even shown in some games (+more with packetloss).
basically, other games can hide it a lot more. they still might have impossible shots, but you're much less likely to notice it, and its feels so much better to hide that extra latency.

if you want a fair fight, use minping, with antilag to cover it up a little.

Well no one has suggested that with antilag we could forget about fair pings. But especially in 4on4 even when we have equal average pings, we still end up in games like yesterdays final where ParadokS had 12ms and I had 40ms. Without antilag I could just as well have left lg in the water, where as now with antilag I could take it and bind the enemy quad into the wall, just like I could if I had 12ms.

Excuses, excuses? Anyway, you clearly do remember - so what are you so surprised about?


Compared to what? Your original question and comment made little sense, but this makes even less sense. How Spike's code is different from Spike's code? Well, what? It isn't.


Huh, what the hell are you on about now?


OK, then tell that to Medar?


Let's get a couple of things straight... There is no 'my' solution. At all. I haven't written a single byte of code towards this lunacy, unless you count a QW config to abuse Qqshka's brainfart. Prolag is a more honest name for what Qqshka/Renzo called 'antilag', which is as dishonest a name as is possible. There is nothing anti about it at all. In my parody post, I called Spike's implementation 'super duper anti lag', ridiculing the sensationalist way this obvious logical flaw was marketed as if a time machine had been invented.

But yeah, just to keep things off-topic: I write about what's wrong with Prolag, I get flamed by the usual Ezguys, I respond by making a parody of them... Yeah, surely it's me with the attitude problem.

Oh, and hey, what's happening here now? I call Medar's bullshit... I get flamed by the usual Ezguys.

2/10

Better luck next time.

Yes, most NQ players seem completely uninformed about how NQ and QW works. It's kind of funny really. It's also funny how they will complain about QW predicting the local player position, while NQ predicting the local player view angles is A-OK.

But no, Prolag has nothing at all to do with the local player position prediction done by QW.

No, you can do it even better. And the player on the receiving end will also have a harder time dodging.

Well, that is the main problem that I always thought this approach might have. I just seem to have hard time actually noticing this effect in-game. I have tried playing 80/12 ms povs but the results remain inconclusive. I mean the benefits of the feature are crystal clear, and if we are unable to show the negative effects in practice it's difficult to oppose this.

Well, Angryfish already pointed out that airshafting is easier with Prolag. The reason, or one of the reasons, is that the higher ping you have, the longer time you get to aim and shoot at the enemy before he gets knocked back in the view of the attacker, which is the view that is used to calculate if the lightning gun hit. Say you have 100ms ping and you start shooting at someone with the lightning gun, then you get 100ms extra to shoot at the enemy before he starts getting affected by lg knockback. But as usual a demo is worth a thousand words, so watch prolag.qwd and you'll hopefully get the idea.

I'll watch it but as a quick note, in my opinion airshafting becomes easier once the kickback starts to take effect, not vice versa. The kickback will slow down the movement and "hang" the opponent in mid air making him an easy target to keep shafting at.

That's very interesting demo. Certainly it doesn't seem right that you can hit almost twice as many cells with 302ms than 2ms.

You get 300ms' worth of extra hits which is 3 with lg.

I thought we all already knew this.

No, I don't think it has been clearly covered anywhere that you hit more cells the higher the ping.

blAze, I understand what you're saying, but imagine in a povdmm4 game; all you have to do is hit them a little when they're falling and if they don't manage to escape your first shaft hit, they'll start getting the delayed pushback. Now if they try to use the pushback to change direction and escape yet again, it won't matter because the moment you reconnect the beam onto them they'll receive more delayed pushback, and so on and so on. This is at least how i see the problem and it looks noticeable when watching high pinging, high skilled shafters like Bulat.

Hmm think i explained this totally wrong haha

Basically what I'm trying to say is that it just feels as though once you start hitting them, the pushback is delayed on your screen so the whole pushing them around the map with shaft scenario is a lot slower and easier to control.

Is this really grounds for dismissing antilag though? I'm not sure myself. What bothers me more is how this effects shafting and dodging when both players are grounded. I hate the idea that lg battles could all of sudden feel 'fake'

Yeah the problem with this is that the scenarios are seemingly endless. Human tests are so vague when even with the same settings the results and percentages can change so much. On the other hand a simplified test situation such as in the demo also doesn't give an accurate picture. If that was a real game, the opponent would already be +300ms on his way down from the balkony before you could even start shooting him, so it's not that clear that high ping would be an advantage. The opponent would have +300ms time to place a rocket in your feet, etc etc etc.

Interesting find I have to say. Even if the ping difference, positioning and the lack of movement is intentionally set up, it still proves a flaw. Something like this shouldn't happen in any situation.

Yea blAze, imo the important thing is not so much about who has the advantage, rather how the game dynamics are changed. But the airshaft thing does give the hpb an advantage from the looks of it. Lots of testing and more testing to be done anyway for sure.

Careful with the thought experiments... You need to look at it from both sides.

Imagine two players standing on each side of some wall... Let's say one player at window on dm3 and the other player at quad. If the player at window is lagged 300ms and he runs around the corner, he gets 300ms where the enemy can't start dodging in *his* view to spray lg at him. That means the low ping enemy is already 90 health down before he can even start dodging the high ping player's shots.

Well my point was exactly that playing this kind of scenarios in your head is a bit difficult because there is too many variables at play. We'd need some kind of simulations. I guess one point is that in our real life experiments hardly anyone ever plays with over 100ms pings, so in practice there has never been enough time for lg to shoot those extra shots. Would capping antilag at 100ms solve this? Would the low ping player see the lagged player 300ms earlier and start evasive movements immediatly so that when the lagged player first sees him he is already dodging?

3 shots at 300ms is an average in these scenarios. So you can't solve this by capping at 90ms. It would just reduce the average to 0.9 shots.

How can lg fire 0.9 times? I though it's about the fact that lg can fire 3 extra cells in 300ms, but I guess I'm missing something here again.

The situation doesn't end at 0.9s. The fight continues until somebody dies or whatever. So the time frame is 0.9+x seconds, where x is quite random. There is a 90% chance an "extra" lg shot fits in that frame.

This is mitigated by the fact that the low-pinger has an extra 0.9s at the start of the fight.

... in certain situations, and doesn't exclude all the other disadvantages the high ping opponent pushes onto him.

Oh, and let me point out that in the situation we're discussing, the low ping player *DOES NOT* have any extra time at the beginning of the fight.

Hmm. Let's say the 300ms player is right at the edge of the corner so that if he moves at all, he will become visible. Now with 300ms ping it takes about 150ms for the server to receive his movement command. The lpb player then receives a packet from server showing the hpw moving around the corner at around 150+13ms later, where as the hpw will see himself moving around the corner the whole 300ms later?

No. The player with the high ping sees what is around the corner as soon as he turns around the corner in his view. There's no lag on the local view in Quakeworld, and his client already knows the enemy is standing around the corner so he is shown immediately.