This comment was made by
Reya1p in a reply to 
Madscientist discussing ETpro antilag. As you can see in the quote, ET's has a few weird bugs concerning netcode and knockback. 
Antiwho?
First of all, how does antilag work?
Instead of leading your aim, you just play the game and the server will calculate whether you would have hit if you ping had been 0. So when your packets arrive at the server 100ms later, it will just check if you would have hit if the packets had arrived 100ms earlier :P Antilag is a good thing as it will make you aim more accuratly.
Antilag and you
So what does this mean ingame. First we got some minor 'bugs'. Sometimes you get shot behind walls. This is normal as the guy actually hit you before you were going around the corner. Add the 200ms lag and antilag: you die around the corner. The lagger aimed correct, antilag made sure his hits got registred. Keeping this in mind, running toward cover before you die might be a good idea. Along with flying in the direction you were running to, you can try to die behind the corner, maybe get an easy revive as the medic should have no trouble finding cover.
Second, all weapons that aren't hitscan (not nades, riflenade, panzer, mortar = 'weapons that move') are NOT antilagged. SMG and luger might show tracers but these are just an effect not the actual bullets, they hit instantaneous. This means that your ping will somehow affect how they shoot. I never noticed this, and tbh when playing high ping its still easy to hit rockets in quake. All hitscan weapons just become useless, as Quake doesn't use antilag.
Last but not least: Laggers. Curse them, flame them, they have been around as long as ET existed. Antilag evens the battlefield, giving both ping 20 and 40 a somewhat equal chance on the battlefield. As a low pinger it sometimes feels like you are at the disadvantage, but lets kill another myth: low ping > high ping. Still, why do we hate polish players? Lets say the lagger has packetloss. A packet gets lost, get resent. The lost packets + the new packets arrive at the server where they are processed. Although antiwarp stops too many fire commands from beeing accepted at the same time it might be possible more more hits are being registered in a certain amount of time, more then the ET firerate allows for.
Packetloss, lag, high ping. We all hate playing versus such players. But the truth is laggers do not get any advantage. You will 'feel' them differently as their hits and the effect of those hits might be delayed, sometimes even come in tiny bursts. But lagging does not give one an advantage. Low pingers are still way better off.
Prediction Errors
Knockback causes prediction errors. It's a common misconception that prediction errors and antilag are related. They are however unrelated. Prediction only applies to yourself, everyone else is interpolated between server snaps, known points, or they are extrapolated. Prediction errors cause your aim to be off, no matter how accurate you are aiming. Knockback causes a prediction error for you when you get hit, because the server knows you got moved, but you don't :'( Some might call this a bug, I prefer to call it a feature. Just like screenshake is a part of the game, a gameplay element that gives a decisive advantage to teamplay, strategic positioning, covering fire and crossfire.
Prediction errors get worse if:
-your ping gets higher
-there's a height difference between you and your opponent
-if g_knockback gets bigger (default at 1000, ESL uses 800)
-if you are being hit more (orly?)
Conclusion
So lets say I'm fighting a guy who is standing higher as me. If I were hitting my opponent would have a hard time hitting me, even if he is aiming right on target. So what could you do if I you knew this?
At close range, preshooting is a good answer. Although you tell your opponent exactly where you are, when executed properly preshooting can save your ass in a lot of 1v1's. Publics really dont encourage you to trust your gamesounds as people will be running around randomly forcing you to check your compass everytime you hear a sound.
At fallofdamage range it is wise to either get the first hit or 'retreat'. If you are getting hit, try to get out of your opponent's track rather then firing back and hitting shit. Ofcourse this might not always be possible, but its nice to keep in mind as an option.
Last but not least the RTCW rule 'shoot at the defenceless guy first, then finish the guys shooting at you' isn't really worth much in ET. When you are getting fired upon, shooting down the helpless targets is not a good choice. Shooting back to give your opponent knockback is the only solution. Even if you manage to kill one guy, getting a gib is impossible, you prolly be dead before you even got a kill.
Bibuy
This collumn wasn't meant to bash antilag. Although I misunderstood it myself at first, antilag has nothing to do with prediction errors. Do not suggest lowering g_knockback even more as its a very big part of the game and reducing it would overpower medics. I tried not to make this a technical rant :P When writing this collumn I was convinced antilag, knockback and prediction errors were related. Luckily some people actually know what they are talking about :P
Read more about antilag and knockback
Shoutouts:
#cpma.et
#k1ck.et
#inm = old #international-night-mixes rip
Reya1p for the quotes used
If your connection is 512 or higher, setting max packets to 100 and always having a stable ping of 48 has always been the optimal setting for RTCW, CoD, ET.
Ping = roundtrip time between you and the server. cl_maxpackets doesn't change that round trip time, the utilization of your link and the number of routers (or hops) between you and the server determines your true ping.
My point is that different values for cl_maxpackets in ET mess up the ping calculation, the ping which is shown on the scoreboard. Your actual ping should not be different between cl_maxpackets 30 and 100.
So saying that you have a stable ping of 48 because of cl_maxpackets 100 is false since your true ping might be lower.
For example: I have a ping of 19 to 21 ms on our warserver. There are 9 hops from me to the warserver so I'm pretty close to it and I guess the routers are from good quality as well (I pass Level 3 as transit provider in Amsterdam, they have a pretty good network). Yet when I set cl_maxpackets to 100 my ping is displayed as 48 and it's a fix value all the time...
Lot of facts and good stuff.Its nice at you still keep talking about this stuff and its always nice to talk about these kind of things at irc.
Seeing the 'remove antilag' comments must make Bani turn over in his ETpro grave... Antilag makes you aim much more correct. In RTCW it all came down to the ping. The fast firing rate and no antilag made even the smallest ping difference give you a huge advantage / disadvantage.
ETpro was never meant to be an all out shooter. It's a strategic teambased game where you have to aim . . . sometimes. If you ever played CPMA you will know when you are actually playing a game where you have to aim 10 minutes non stop instead of shooting 2x4 seconds every 30 seconds ...
Overpowered medics, screenshake, knockback, grouped spawns, etc. It all ads up to make ET a game that is team and strategy oriented. It might be fun to have 'an anti anti lagged cup', but I see no reason why we would turn back the clock and use a netcode that isn't meant for an online game. It would not give you the perfect ET, it wouldnt even improve the game only remove the one feature that made ET so popular among avid gamer.
It's a feature.
Screenshakes are waay more annoying in ET than they were in RTCW, atleast to me. (Perhaps this is because RTCW was smaller) So imo the annoying amount of screenshakes in ET are a feature, but an unbalanced feature.
optimize prediction is a client cvar. Feel free to turn it off, and enjoy exactly the same gameplay with lower FPS.
fixedphysics just makes jumping highs more consistent. While really hardcore TJers may prefer normal physics with the right FPS cap, it doesn't have any major gameplay impact.
we will ask you about your opinion, concerning antilag, in half a year.
nP :)
If there wasnt this antilag programm, hitboxes would be no more moved on the right?
b_screenshake [0-1]. The ETpro coders almost fed him to the lions :/
b_screenshake [0-1]. The ETpro coders almost fed him to the lions :/
i actualy suggested total prediction error removal, that means no knockback, no screenshakes, no spread. Meaning it balances calsses more but also makes aim even more relevant then it is now. Thats why i was almost fed to the lions
Tbh, i realy want cpma for et, with no screenshakes, no konckback and no spread+ birghtskins, more focus on jumping and stuff like that. Lucky for most people i'm the only guy who wants that :(
Thing is i think of etpro as a pub mod for skilled players, not a competition mod
i hate that after trying out madscientist's mod i have to play some game that has so much lotto about aiming.
you have to agree with arQon on one thing - i forgot the comment, but i remember the point which is something like: any game that has headshot hitbox and spread is a noob game.
you get a much nicer game without spread in terms of killing, although even easier gibbing could be a problem
i dont know, i mean i have been playing et since release and im used to it, and i get more annoyed with lame spam based tactics in wars but playing on pub without spread i seemed to get mainly 2 or 3 headshots per kill without massively lower accuracy, where with spread i sometimes just plain suck even with the crosshair on the opponent it seems
i think aim at the moment is partly down to movement - in terms of minimizing spread increases - i remember nonix watching me play at my pc and he said something along the lines of 'aim is fine but its this hand which is bad' - after thinking about it i realised that he meant the difference in his movement and my own.
oh and also movement can provide a 'smaller' body hitbox or remove it further from the head box so youll get less lotto body hits due to spread leading shots away from the target at times even if you were aiming on the head
explain please ?
Crouch and prone obviously change the size of the hit box. Hit boxes are axis aligned, so depending on the relative positions between you and the other player, there is more or less area exposed. However, if you are in a roughly face to face fight, this is always going to be pretty equal.
Also, ping is just one aspect of your connection. You can have a good ping and still have bad hit registration, lag, etc.
1. We make the game fun for a bigger public, by leaving this antilag on en letting things be like they are we give more new people a chance to kill others and have fun playing this game.
2. We make the game more 'skilled' (ofc this skill thing is arguable but you guys know what I mean). To do that a few things can be changed, first turn off the antilag OR make some changes in its code (I don't know what is possible in this) and I know then people with high ping will get a harder time hitting opponents, but this is the way it is in most games + imo people with a better connection to the server may have a little advantage + the antilag we have at this moment is so good/bad that people who lag can even have an advantage on people who do not lag, for instance the hits from a lagger are not coming in the same rate as from someone who doesn't lag wich makes the knockback get worse wich makes the lagger harder to hit. Also bad connection and antilag combined makes him not only warp (or how you call it) so he is harder to track but also makes his own screen NOT warp so he can hit without any problems. This means in ET a lot of laggers (not all of them) have an advantage, and that's not good imo :P
Also to make the game more 'skilled' I think it should be more as RTCW, not just because im a fan :P but in general if the firerate from smg's is higher the gap between a bad aiming player and a good aiming player will be bigger plus you could add more stamina and maybe other things.
Ofcourse now people will say 'ok so we make the game playable for more people OR make it more skilled, well in this case more people is better' BUT we have mods :D and one of those mods is called ETpro, imo a 'pro' mod means more skilled than the original game, well if the ETpro team made these changes it would actually be more skilled, and the not competitive players would be stuck in ETmain :P
This solution maybe far from perfect but its a start and I think its just something we could give a try :)
You can use b_wolfrof (on the server, obviously) to get RTCWish ROF in ET.
Completely untrue. Most likely you've never tried playing with 150+ ping. You might be harder to hit, but that's more than compensated for by the fact that it's harder for you to hit other people as well. You think it's easy to hit stuff while warping?
If your ping is reasonable, the effect takes quite a bit less than normal human reaction time to happen.
Screenshake (another topic) means that once you get hit, correctly tracking is gonna be much harder anyway. The game is designed to give a disadvantage to the person who gets hit first. Whether this is good or bad is subjective.
In a head on fight, you will just get pushed back, which isn't gonna affect your aim point much. It gets worse if you aren't on the same level as your target, but I suspect the number of encounters that are decided by knockback alone are quite rare (and most of those are PFs being knocked back enough to vap themselves, which is a Good Thing :P)
1) The effect you mention which sets off your aim point, resulting e.g. in shooting just above the head although you were aiming correctly
2) (Especially on stairs:) You get pushed into the air (losing ground contact) for a short moment, resulting in maximum spread
Crouching doesn't help with that, aiming at the middle of the body only helps to a certain extent, in that the total hit angle might get larger and even a bullet with maximum spread could still randomly hit, and trying to run forward (even forward + sidewards) to reduce or eliminate the effect gives you a tactical disadvantage against an opponent who is only strafing sidewards.
Antilag doesn't make the laggers hits all arrive at once. As stated before, all it does is change the position that is used for hit detection purposes.
i have had that feeling of a few bullets hitting at once at times as well when you seem to go from 80hp to dead - actually it seems so abrupt it could even be my client recalculating after bad prediction when recieving a few lost packets?
Just an assumption. Could it be that because of the packets arriving later, they would get processed during a later server snap? This would result in more hits beeing processed in one snap so you could do more damage in one snap then the ingame firerate actually allows you to?
Actually its a question for reya1p :P
but its not really about the server its about loss of packets and them being resent causing errors in the prediction from clients to be corrected based on server snaps that were sent with data which is forced as an update AFTER the clent has rendered what it thought would happen
This seems to be more confusion about prediction. The only time you see another players position predicted (extrapolated) is if a snap from the server to you is delayed or lost. The server never extrapolates or predicts player positions.
If your connection is fine and you are watching someone with a bad connection, they will stop when they lose packets. When communications resume, they will go again. Without antiwarp, they may move very quickly as things catch up. With antiwarp, they should not.
Also ET never specifically resends packets. It has other mechanisms for dealing with loss.
If this happened, it still wouldn't be related to antilag ;)
usercmds (which ultimately result in bullets being fired) can get bunched up. This inherently implies that some of them are late rather than early. Antiwarp is responsible for smoothing out bunched up commands. It can't do this perfectly, but it should take out large bursts. The kind of thing you describe is far more likely to happen do to problems between you and the server. If you lag for a second, you can happily predict yourself along and then suddenly discover that you were dead a while back.
If people can show cases where people can cause major warps with antiwarp enabled, we'd like to see it.
and even with tracers on you arent seeing whats happening on the server side of things so thats no use in forming a judgement for your average player
but i do think that the nature of networking in europe is different, with more old phonelines involved, generally lower bandwidth for the average user, and slightly different standards in telecommunications between countries you are going to have more chances for physical problems with internet connections, even if the protocols used are exactly the same as in the states
I agree that quite a few OMGWTF moments are caused by spread.
Some are caused by unexpected effects of poor map construction:
http://www.collectivecomputing.com/~reed/et/etss/ggcliping-sw_oasis_b3.jpg
That said, it is certainly possible there are still problems in the hit detection. The problem is nailing them down.
I suspect part of this is due to the way humans seek out patterns...
When you have a good run where all your shots connect, you try to find something to explain it. Equally, when none of your shots hit, you look at /players and try to find some netsetting to blame it on.
(you in the general sense)
It most certainly does neither of those things.
I find it fairly obvious when there's loss. Lost snapshots show up as red in the lagometer. Lost usercmds aren't displayed, but they have a distinctive feel, and it certainly isn't good.
So there is nothing that can explain the phenomenon where the hs are abundant even if your aim is off?
My experience is that a steady high ping can feel like more shots are landing than a steady low one. Real ? Placebo ? I don't know. I've never found a truly laggy connection (PL or lots of spikes) to feel better than a smooth one.
A bad connection will make hits more lotto, in both directions (i.e. getting hits when you don't expect, and not getting them when you do). That's the nature of the internet.
Another thing that adds to the confusion is that hitsounds come from the server, so on a laggy connection, they can be significantly delayed.
lower maxpackets will make it look like your ping is lower on the server
and just because your ping is equal it doesnt mean you are sucessfully recieving the same amount of packets as guys from the other side of the world.
"cmd"
In the console type:
"tracert yourISP.com"
"tracert serverip"
For me its
"tracert telenet.be"
Usually gives you some info on where you are getting network problems.
my trace route to www.sapo.pt (my isp) and to TELENET 2
I use 100 maxpackets everytime, but if i use 40 i cant notice a big difference on ping (less 5 or 10 in nl)
Anyway some ppl like finnish, can have 20ping or less in their own country servers (maxpackets @ 100]. Is impossible in portugal :< i can only have +/-30 ping with 15 maxpackets... 100 maxpackets = 48 ping @ portugal xD
Anyway, if you are being routed from Portugal - England - Germany - Holland, it would explain thet crappy ping. Call your ISP and tell them to fix their fucking routing.
The ping in ET is fucked up, your ping is the same regardless of what maxpackets you have, be it 30 or 100 - your 'real' ping is still the same. Check your ping at the command prompt with the "ping <host>" command to see an accurate ping.
i will send a email to my isp noobs , thank you
Ur ping to isp is not that good. Most users(dsl,cable) ping under 10 to isp thats why they can have so low ping .
man, u shouldn't talk about things that u got no idea about... and who is that ex team mate of mine?
I doubt there is such a guy
Afaik (*) the reinforcement time offset is between 0 and 15 seconds, meaning effectively worst case 30 second advantage in initial respawn for one of the teams (e.g. Allies 1st spawn after 5 and Axis after 30 seconds on Radar, or Allies after 20 and Axis after 15 seconds at the other extreme). Then do the same in SW round 2 and tell me this isn't a deciding factor in close games. :)
*afaik = reyalP will probably correct me :)
(Sorry for OT. :P)
edit: ooops perhaps i should read the whole comment before answering next time :o
thought u were talking about missing spawn etc
Best case for allies on Radar:
Allies 1st respawn after 5 seconds (then every 20 seconds)
Axis 1st respawn after 30 seconds (then every 30 seconds)
Worst case for allies on Radar:
Allies 1st respawn after 20 seconds
Axis 1st respawn after 15 seconds
You see the difference? :P
This is really complicated whats are the "bad" spawntimes, depends on maps ofc.
I feel that I play better when lagging, but then again you cant guarantee the skill of the israelis or polish I played vs, anyhow sometimes I barely had to aim on them to get hs.
so antilag improves my aim, and makes me harder to hit, what more can I ask for :D?
I can also remember playing on servers with high ping and not being able to hit shit, I think that packetloss is a big factor.
2.55 = more hits
2.60b = barely any hits
I still remember switching to the new patch. :(
Also, you can throw in any theory you want, but high pingers > low pingers, and antilag is there to "blame".
Following up kiitos here, I remember playing FF on our local (Dutch) server (where I ping 7-10), 20% acc anyone? And then their (German?) server (where I ping 40-45). I owned.
If high ping is better why none of the euro guys want to play on servers where they ping more than 90?
you whine when you play vs "high" pingers and you whine when you have "high" ping so it's kinda redicules.
Have you ever heard of unlagged for Quake?
With the same reasoning I can say that ET doesn't use antilag.
"uh, antilag does not cause prediction errors. Please ... prediction has NOTHING TO DO WITH antilag. NO FFS!!!"
"Prediction errors get worse if your ping gets higher"
"Antilag fixes aiming so you can aim at where people were in the past rather than aiming ahead of the player relative to your ping"
Antilag has a greater beneficial effect for higher pingers.
Prediction errors have a greater detrimental effect for higher pingers.
Ergo they are related, just not in a technical manner. Now ask yourself what is this community?
Common perception is that antilag fixes everything for the high pinger. Prediction errors state otherwise. Antilag has been seen by the community as the "OMG L33T LEVELED PLAYGROUND" provider for high pingers. It is NOT!
We can go on and on trying to explain why you don't hit people at certain angles, ping, servers, time of day, phase the moon is in but the base facts remain, there are MANY a time when you may be aiming perfectly at an opponent and be it whatever you will almost not land a single bullet. Plus you missed one of the other biggest problems of how bugged crouch is (Hi Butchji).
Higher ping leads to a greater amount of bullshit in games such as being killed in 0.3 seconds from 8 body shots, not being able to react fast enough to get behind a wall, high ping is A BIG advantage for people who know where an enemy is walk out crouching and shooting the opponent CAN'T react if they do react they're either dead behind a wall 1-2HP in less than a second without even a headshot.
what about when low ping corner camps and then busts out shoots.
your dead (the high pinger) before you even see a person...
all anti lag does is make it so you don't have to do prediction based on ping....
so please play high ping and then come back and say yea sweet high ping is awesome....
I've played with high and low ping the point is neither have that great of an advantage but what IS true is that high ping leads to unfair/stupid/bullshit frags; low ping does not however so naturally low ping should be favoured in competition so ban laggers.
Agree :P
The high pinger cannot shoot the low pinger until the server tells him that they can see each other, and with a higher ping that means "later". Also remember that deaths are not unlagged, so if the players have exacly the same reaction time and get the same hits then the low pinger will win every time (the high pinger will think that he hit the low pinger enough to kill him before it "rewinds" to the point where he himself was killed).
Why not use the magic ball to predict if people will hit or not. AntilAg sux0rs bigtime, mkey?!!!1111
Really, any ping under 100 should seem pretty lagless, so how is it possible that it's worst with 70-90 unstable pingers? Maybe it's netlimiter or packet loss or whatever, but it's not lag, at least not in the sense of ping/latency.
Your conclusion is that antilag is bad and causes "bullshit", which is inevitable. What I'm saying is that antilag is overall for the better, and that it's effects are generally exaggerated. It most definately doesn't give the high pinger an advantage.
I dont like your reasoning. Personal experience gives you somewhat of feeling where the answer lies, not the truth. Antilag = most of the time you hit ok, sometimes spread adds some lotto.Laggers = 'wtf hax' aka "they lag". They feel different compared to normal plays who have ping 48, end of discussion.
ps: you have a lot of frustrations my friend. I dont care playing laggers, but at least I dont deny a thing anyone of us experiences every day: antilag is nice, but also gives you teh 'wtf hax' from time to time.
About personal experience, it's just that solid proof of unhittable laggers who aim and play well is lacking. That's where I mix personal experience and logic to form an opinion... I'm certain the etpro team would like some proof too, since that allows them to try to fix any potential problems.
Play with emule.
BTW Hi Matias...
however, i had the same discussion few weeks ago with another random crossfire user, I already know u dont accept arguments :)
I can get a 100mb connection for like 60 euros a month.
go play cod2 with more than 100ping and change your mind.
BUT WHATS THE PROBLEM ON CLANWARS? HOW MANY LAGGERS DO YOU SEE ON CLANBASE MATCHES OR PRACCS OR WHATEVER? i see NONE! pings from 15 to 98. NP at all. If it happend that I found Israeli opponents (ping 200 on their server), I just disconnect and I find another opponent.
also never heard about anyone whining on me for hitboxes and stuff (polish connection, rly bad one)
so, ALL WHINE ABOUT LAG COMES FROM FUCKING FFA
->
AGAIN, TRYING TO FIX SOMETHING THAT ISNT BROKEN
EDIT: Many guys cant understand. You feel that you play better when highping+lagging. But you have no proof. It's the feel. Maybe its because you know you have high ping so you focus more on game than when you have 48 and therefore have more headshots and stuff. DUNNO but till you give eptro coders logical and stable evidence, how can you scream about forcing antilag off....?
PS I don't care what you do since I don't even play ET anymore.