How long is it going to take the devs to nerf the projectile speed of rockets back to 1000 ups from 1200? They originally increased the speed to 1200 because of the delay in firing rockets, there is no need for them to be so fast anymore
Average dm still looks like this despite the hitbox changes
Galena is best girl.
Slash, Sorlag and Nyx fags need to be put in camps.
Bout when can we expect 2v2 patch?
i want 2 fug the slash
Here twitch tv/videos/222584320?t=24m50s and twitch tv/videos/222584320?t=26m40s they say that they target February for instagib and are also looking if they can add 2v2 ranked as well but it didn't sound as likely as instagib. And I guess the February patch will be at the end of the month again. CTF should follow after that if I remember correctly.
>twitch >tv/videos/222584320?t=24m50s >February for instagib
FFFFFFFFFFFFFFFFFFFFFFFFFUCK YESSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS HOLY SHIT THIS IS AMAZING
>Instagib this month. That sounds neat.
It's february already Tim. You said there would be patch Tim. Where's the FUCKING PATCH TIM!?
what is special about this game mode? i read that people die in 1 hit from every weapon? if thats how it works, sounds kinda dumb. like the kinda thing thats fun for a match or two then gets boring real fast
Everyone in that thread seems like an entitled little child. I don't want to whiteknight ID at all, but they're just being ridiculous.
>I don't want to whiteknight ID at all, You are not fooling anybody SyncError. Just fuck off, nobody likes you.
Joke's on you, I'm Tim
Whats wrong with gurkan?
guess theres really not a whole lot to talk about unless theres a patch or tournament or something going on. quake is like 30 years old at this point; pretty much anything you have to say about it has probably already been said and has a youtube video dedicated to it.
end of month
reminder gurkan killed /arena/ with his cancerous posts
Rockets feels good as they currently are IMO Also interesting that you qualified it with "average DM". I basically only play custom game Duels and LG and Rail are as scary as rockets, if not more (depending on champ/situation)
why is this image so fucking funny
funny how that works. every now and then ill see a normal-ish looking image and it just sets me off. now that youve said the image is so hilarious, i couldnt help but chuckle when i saw it, despite it not really being funny.
you think this is funny HUH? this is SLOSH, show some respect
so stupid. i never noticed how weird her face looks
S L O P
Yeah and what's up with nox?? What were they thinking
S L A G
looks like the villain from final fantasy 7. someone turn the hair white
no fucking kill yourself sage
i wish i could
take it back, slosh is cute the weird part is her gigantic giraffe neck because russkies spent all their modeling effort on her tummy
so how does it usually go for a player starting from nothing to getting good? is it just gradual getting better, or does there come a point where all the practicing and stuff kinda just "clicks"? i woke up today and suddenly im not total dogshit at this game anymore, even finally won deathmatch once.
we all have our days but for me i always get gradually better with any game.
feels real good improving. even just getting 3rd place is great. have not had this much fun since titanfall 2
what would you guys consider the best piece of advice/tip/strategy youve received that helped you improve at quake?
It's hilarious to hear people complain. There are things in QC that are broken as fuck, but they don't know what it is, and just keep coming up with bullshit that isn't even true to justify why they're losing. Reality: you suck, you're a +forward retard, as the game improves you're only going to do worse, git gud, stop playing Slash, and shut the fuck up.
>There's more than just x weapon, you just can't aim
I live in Japan, and I just want to say the Que times are fucking ridiculous, like multiple hours sometimes. It's a shame because it's really fun but still.
I'm not surprised, Japan loves its fighting games and weeb games only.
>144hz monitor hasn't even been repaired and started shipping back yet. 60hz is killing me. I wanna play quake again.
This is largely because the LG isn't powerful enough to moderate the rockets, and the HMG has too much bullet spread.(Bloom).
They need to buff the LG back to 150 starting ammo, and they need to reduce the HMG's bullet spread, and increase the fire rate.
it's due to projectile prediction, and maybe some other netcode factors. you can't dodge rockets- it's just up to the attacker to miss. Rocket pop combo just overpowers LG with the attack so heavily favored.
I'm getting my new optical keyboard tomorrow, and my monitor is gone for the time I want to take to test it. This sucks.
Well part of it is that the LG doesn't have any ammo when you pick it up anymore so you can't prefire it anymore or you run out of ammo. So they need to rebuff the starting ammo to 150 again. Maybe give it a 50~75u range increase. To solve this problem, without a buff they need to buff the HMG by reducing it's spread, so that you can combo HMG distance and LG swap.
Rail->HMG->LG Should be the strat method we use, a lot like in Quake live duels. But we can't because the bloom on the HMG is too much and there's no zoom sense for Rail to let people dial in their rail zoom, and the HMG is boarderline useless.
Even if rockets were 1000ups, I still think it's going to be a problem with the LG being weak. and HMG being useless. That's also why the nailgun is so strong right now, because it has so much ammo over time, and the LG isnt powerful enough to surpress it.
But anytime you talk about buffing the LG back to it's proper place, people flip their fucking shit, and forget that LG is part of the trinity, it's supposed to be stronger than nails, shotguns, MGs, for it's range.
150 cell LG on pickup is too much. I already barely ever run out of cells in Duel with the 100 pickup. Picking up ammo needs to be a thing. Rocket and Rails run out often, too >Even if rockets were 1000ups, I still think it's going to be a problem with the LG being weak. and HMG being useless. HMG once was far more accurate and it overshadowed the LG I like where it's at, it's stronger than the LG, but from up close >But anytime you talk about buffing the LG back to it's proper place, people flip their fucking shit, and forget that LG is part of the trinity, it's supposed to be stronger than nails, shotguns, MGs, for it's range. LG is still very strong. Watch dahang (or someone with good tracking) play. Mathematically the strength is still all there, they did NOT alter its DPS
What do you mean by buffing LG Barring Q1's superweapon LG, this is the strongest it's ever been. The cone detection is very forgiving compared to a trace (how it is in every other Quake) And it's still 7dmg a pop with the same ROF it ever had
I don't think you understand. Prefiring doesn't make a lick of difference, unless you have the reaction time of a sloth, because you're certainly not going to get more than a single tick extra before you can track. You won't hurt anything that's not on your screen, it's all client side, the attacker has the advantage on very corner. They get to come around the corner see you and fire before you'll ever see them on your screen.
here play with this simple model: gabrielgambetta .com/client-side-prediction-live-demo.html You can sorta recreate the supposed netcode environment of QC. Prediction, Reconciliation, and Interpolation set for both Player 1 and Player 2. Supposedly 120 server updates a second(fast! but I haven't checked in long while). Now image the dots represent a simplified scenario Blue dot approaching red dot = attacker coming around a corner. Red dot moving away from blue dot = defender dodging visible projectile.
Now there isn't a line on the screen but in your head draw a vertical line straight down between the two dots. When blue dot crosses that line on player 1 view, that's the moment in which he fires his rocket. Notice the delay before player 2 view's blue dot crosses said line. Red dot doesn't get to react until the blue dot HE sees crosses the line. At which point Red dot starts moving to get out of harms way. But player 2' view does not matter, only the player 1's view, because it's the shooter's view that is checked for a hit. That's right, the dodger's effective time to action isn't delayed by just his latency, or his latency+his attackers latency, but his attacker's latency+his latency+his attacker's latency+his latency again. Quake is a lot faster than these dots, but the delay is the same.
for kicks set latency for both to less than 45 and turn off all features
YESS THE THREAD PICTURE
I think you guys are missing a critical point here. This is QC balance were talking about, not other quake games. The weapon strats are still similar, but they balance setup will be different.
Right now the LG isn't strong enough to moderate the rockets, and it's not just because of the rocket UPS speed. It's also because of the type of encounters that the rocket launcher favors in QC specifically. Unlike previous quake games, the champion setup and hitbox system blatently favors rockets over anything else except Railgun at long distances. It's not just that Rockets are fast it's also that they hit far more reliably, despite their odd off reg problems. Especially with how large the hitboxes are now in respect to the rest of the game, and how the smaller hitbox champions now die so incredibly quick to rockets.
Where as LG's overall state against the new hitboxes even with the cone detect system still doesn't really change how fast it kills. PLUS ontop of all of this, QC's stacks and limited health pick ups again favor rockets over anything else. The LG and HMG are not strong enough at this point in time to stave off the massive over use of rockets. What first needs to happen is the 1000ups nerf needs to happen and then see if the LG is still lacking against rocket engages.
If the hitboxes werent so big, and the light champions so low health, rockets would be less effective, but the climate of the game simply favors the rate of fire, and burst damage of the rockets over anything else.
Who needs to worry about nyx picking up a health bubble when you double tap her with rockets and she's instant death. 3 Rockets against slash/galena/ranger/visor, and it's way easier to hit direct rockets against scale, sorlag, clutch, keel.
It's not just the rocket speed, it's their reliable damage you can do ontop of that in a short span of time.
This is also why the hitreg nailguns are so powerful now, and are being so favored.
And, if you buff the light champions they become a serious problem. The best way to deal with all this is a stronger LG, and HMG. But not a huge increase, just a little bit. Just enough to make people pull out the LG when an engage comes too close.
You could simply say that QC's networking (and this is a symptom of modern FPS networking) sacrifices visual accuracy to preserve the "integrity" of the act of shooting across multiple latencies
Carmack realized that being able to trust what you see was very important for an action game, so he devised the "you only see yourself fire after the server accepted you did" system. But most people are too bothered by the idea of having weapons function slightly differently across servers (and even latency oscillations on the same server influencing the delay before a weapon is fired) to actually vouch for the oldschool kind of networking, even if it is more fair.
It still isn't 100% foolproof. If I have 20ms and you have 100ms, I send a packet to the server requesting a rocket to be fired across a vector, the server solves it and it's time to send everyone a packet updating everything with what happened - I will get the packet in a round trip (40ms - 2*20ms) and you will get it in 100ms. From your perspective it will most likely look like you just got a packet saying that you died and an explosion effect will be drawn on the impact point - You will not be able to dodge that rocket either, as your ping disallowed for such a close range borderline maneuver
Still, the game would still feel more fair in all other situations (not dependant on miliseconds on the double digits). But again, people don't want variable delays on weapon fucking up with their practiced timings
Seems you don't enjoy Quake. Rocket is supposed to be the workhorse weapon. It is meant to be good at most situations Play Quake 1 to see what a RL that dominates in most situations actually feels like. QC's RL is taking after QL's, which were also sped up from VQ3's (because a lot of people felt Q3 was too hitscan-heavy, including the id devs) LG, Rail, SSG...The other weapons are meant to be less self-sufficient than the RL. You need to forget this idea that whichever weapon you pick should put you back into the game 100%. Even inside the trinity they were never equal, always changing a bit game by game
>It still isn't 100% foolproof. If I have 20ms and you have 100ms, I send a packet to the server requesting a rocket to be fired across a vector, the server solves it and it's time to send everyone a packet updating everything with what happened - I will get the packet in a round trip (40ms - 2*20ms) and you will get it in 100ms. From your perspective it will most likely look like you just got a packet saying that you died and an explosion effect will be drawn on the impact point - You will not be able to dodge that rocket either, as your ping disallowed for such a close range borderline maneuver
That's wrong though. He gets a stream of packets at server update rate, it doesn't matter how much latency there is, it'll still look consistent from rocket firing to death by rocket. Anyway, the difference in reaction time in your scenario vs. lag compensation scenario is (100ms+2*20ms) or 140ms vs. (2*100ms+2*20ms) or 240ms.
>it'll still look consistent from rocket firing to death by rocket Even then it won't...Or at least, with movement prediction it won't - Lagger may, on his screen, be able to see himself dodging the rocket that will kill him, but it won't matter, because to the server he is already dead - the target of the rocket was already hit, he is seeing it late through the lens of movement prediction. The rocket will seem to miss him yet do the requisite lethal damage if he moves away
Netquake-style is the only way to truly eliminate these inaccuracies (if you want to go even further, delay mouse movement by your latency - see who wants to play that), but who wants to move with WASD delayed by your ping
we're comparing prediction+latency compensation vs. simple authoritative server with regards to projectiles specifically
YES, prediction makes it look like fuck. like right now you will have your rockets erased, your rails un-fired, and your corpse sucked back through the tele to die on the other side. Pretty much for every single one of the deaths you have.
It's either playing on LAN or changing everyone's preference to unpredicted netcodes, so devs starting tending to that I personally don't mind Netquake-style netcode. I occasionally play co-op via Quakespasm - It's a delight to play at low latencies (the cutoff is like 60ms, though), it really feels like SP over the internet. And there are modern NQ servers with raised tickrate, so you can also do advanced movement in them, which the original low tickrate of the Quake server executable disallowed Quakeworld-style networking, even if it became the norm (and now we're further away from that, with shooting occurring client-side in most games and just the results of said shooting being uploaded to the server), was basically a compromise to allow for people with high latencies (56k or even less) to play against people with T1 lines and such. Internet bandwidths (and coverage) have become a lot better, so maybe devs should start looking into brutally server-side networking again, for the sake of accurate action
>so maybe devs should start looking into brutally server-side networking again, for the sake of accurate action
they really should. 120hz updates and sub 50ms latency standards
>has red LG skin
oh god what is wrong with you people! Authoritative server hit detection is a fucking garbage experience! You're talking about "trusting what you see" but server hit detection makes that even worse! You have to aim where the server thinks the enemy is - but I don't know that yet! The position of the enemy hasn't reached my machine yet, and the server requires me to just magically know where he is? It's garbage! You guys have forgotten how much it sucks to lead your rails
And your argument that modern latencies makes it ok doesn't hold water. Lower latencies improve client side hit detection just as much, I'd argue even better. At 50ms you will absolutely notice the delay between your click and the rail firing. But at 50ms you will not be able to notice if you made it completely behind cover before getting railed.
Your mental models of netcode are stuck in the past. You need to stop thinking that the server holds the source of truth. Each client holds their own truth. You need to stop thinking that an active game is represented as one instant of in-game-time. Each client's game is represented as a history of events. Every client tells every other client about their entire history of events, and those histories get merged together. The server's job is to pick one correct history when two histories don't merge cleanly. For every modern shooter, the server picks the shooter's history as correct. Because it's better that way.
Really, we should be debating how the server should be handling conflicting events. I think that getting railed before going through a TP should favor the target instead of the shooter. I think that when both enemies think they got a kill in a rail v rail battle, both kills should count instead of just the player with the lower ping. I think they already got nyx right, when she goes invis it always evades damage, even when the hit registers from another player.
There's no reason why rockets can't be doubly-predicted. When I see an enemy rocket on my screen, both my ping and the enemy's ping has already been taken into account, and the rocket has been fast-forwarded along its path by that ping difference. So even though the enemy's client is responsible for hit detecting that rocket, the rocket I see on my screen is quite accurate
>You have to aim where the server thinks the enemy is - but I don't know that yet! The position of the enemy hasn't reached my machine yet, and the server requires me to just magically know where he is? It's garbage! You guys have forgotten how much it sucks to lead your rails When you flick your camera to shoot someone, you also don't know if he will be there when you shoot. Even if you try to "confirm" the shot within your reaction time (which you will rarely have time to do in a fast-paced game), the enemy can still move unpredictably within your reaction time - getting away within the ~250ms or so you take before deciding to shoot your "confirmed shot" and actually shooting. Even basic strafing will evade a shot within that time
> At 50ms you will absolutely notice the delay between your click and the rail firing. But at 50ms you will not be able to notice if you made it completely behind cover before getting railed. At 50ms you will absolutely notice warping back with movement prediction on the client. Especially in an AFPS
It's basically a tradeoff between the integrity of shooting across multiple latencies and the integrity of movement. Most people think this is an obvious choise, since FPS are about shooting - But shooting with a delay, if you can compensate for it, while movement keeps its integrity can be an interesting alternative. If you're willing to put up with the delay, you can have far greater solidity in your ability to dodge projectiles on reaction, for example
>Lower latencies improve client side hit detection just as much, I'd argue even better This is impossible to argue, so no, you wouldn't There is an extra layer of abstraction in client-side hitreg that will make anything except literal 0ms have a greater disparity than server-side authority The gist of client-side hitreg is that it will not wait for the server to "simulate" your shot and give you the result. You will shoot and see the particle effect, the tracer, whatever - And the packet will go the server saying "I HIT X FOR Y DAMAGE". Then the server will acknowledge it (with rollback or not, depends on implementation) and send you back the effect of your shot (i.e. the enemy dying, the enemy taking damage, etc.). This is how it is in most games, with concession for cases in which it adds too much weirdness (e.g. in TF2 they predict hitscan but not projectiles). And with server authority, the actual shot simulation is made on the server and sent back to you (i.e. it's not only acknowledging the effects of the shot - according to the server you can still miss before it says you hit)
But back to what you said, that lower latency improves CSHD as much or more - Both of these are false. It will help facets of you gameplay, but far less than in a serverside model. There is a mandatory higher disparity between what the attacker does and what the defender sees (unless it's 0ms), see the latter part of this post, I can go in deeper if you can't understand but I'm starting to approach the charlimit Where I'm getting at is, for a server-side authority model, lower latency will remove roundtrip latency directly. You used to play at 100ms, now you do at 50ms. This is a literal 50% improvement at everything related to the game, dodging, reacting to people around corners, the works. Not so with CSHD - Your lower latency will only help your shooting register faster, as a lagged enemy can still hit on his client side and create massive disparity for you
Just to clarify real quick, charlimit forced me to make that last paragraph read weird. When I typed: >Your lower latency will only help your shooting register faster, as a lagged enemy can still hit on his client side and create massive disparity for you What I meant by this is that shooting is decoupled from latency in CSHD models, so your lagged assailant can still hit you from positions that are completely absurd in your screen (including rushing you with a high latency, seeing you, shooting you - and by the time you receive info of him entering the room, a few ms later you're getting damaged, never having had the chance to react - Imagine this situation in the "old style" of netcode and you'll see that even if he can still spot you earlier due to predicted movement on his client, you will still have time to react to his shooting) >inb4 "b-but if everyone can shoot in an "unlagged" fashion it's fair!" Not so, because I'm not shooting in a vacuum, I'm shooting another player. So CSHD can benefit laggers in extreme manners in certain situations. Low latencies will benefit a player in most manners, of course, there's no escaping that - But no, it will not benefit a player more than it would in a server-side authoritative model
If you "confirm" a flick, then it isn't a flick. Your argument that no human action is ever fast/accurate enough to make a good shot is fucking dumb, because pros hit fucking amazing rails every day. But those pros would not hit those rails if the enemy was not actually where the screen is showing him
>you will absolutely notice warping back yea, now that you phrase it that way, you're right I have noticed it
>shooting with a delay, if you can compensate for it, while movement keeps its integrity can be an interesting alternative if you build that game, I wouldn't want to play it
>Imagine this situation in the "old style" of netcode and you'll see that even if he can still spot you earlier due to predicted movement on his client, you will still have time to react to his shooting You will have time to react exactly equal to his latency, which in most cases will be far less than a human reaction in miliseconds. But yes, it will feel more accurate than the same situation in a client-side model (in the sense you will have actual time to get out of the way AFTER you see him, even if it's not his full reaction time like you would have @0 ms latency) Well, all Quakes up to QL (which added unlagged to hitscan only) were successful with the networking style you're calling "antiquated". All UTs had it too, including UT3 - and Unreal Engine's default networking is still that very same one, it delays your shots by your latency These work well at low latencies. The further we get away from these models, the bigger the compromise gets to make latency "unnoticeable", but you always pay with inconsistency in some other area of the game. With CSHD models, it's definitely dodging/reacting to stuff
>if you build that game, I wouldn't want to play it I guess QC is your first Quake, then
>Seems you don't enjoy Quake. lol how did you come to that conclusion? QC is my favorite FPS game.
>missing the point that fucking much
Well, Quake's weapon balance was never meant to make every weapon highly viable by itself like it was CS or something (exceptions notwithstanding, the Glock is a poor match for a M4, but a Galil, a FAMAS or even a MP5 can mostly stand toe-to-toe with it) Most weapons in Quake are "combo" weapons, you're not supposed to walk around with a LG ready to do a complete kill with it (even if it can happen) RL is the weapon all others are balanced around in every Quake except Q2, so it should be something for you to expect
I don't think you're reading my posts correctly user.... You completely glazed over the whole combo and range situation I was explaining...
You type stuff like >Right now the LG isn't strong enough to moderate the rockets What the fuck does this even mean, then? This is not something you should be expecting from a Quake game. The LG, in general, is not meant to hold a RL pursuit back (even if it can happen). You need a RL of your own to better do that (and to better engage from mid-range) I'm trying to think of a situation in which LG can "moderate" RL and all I can come up is Quake 1, in which the LG was basically a full auto 30dmg railgun
>This is impossible to argue, so no, you wouldn't lol wtf do you think an argument is? I can and I will, watch this!
>But back to what you said, that lower latency improves CSHD as much or more - Both of these are false You claim this is false, but your only argument relies on a premise of mismatched pings. You're right that mismatched pings creates asymmetric advantages on both sides, but it's awfully mean of you to say that this one case completely invalidates my argument. When latencies are about 100ms on both sides and lower to 50ms on both sides, CSHD absolutely improves just as much. This is a literal 50% improvement at everything you listed, dodging, reacting to people around corners, the works. And with CSHD, my hit detection experience was already 100% flawless, and it continues to be 100% flawless with lower pings. I put a high value on that experience!
Mismatched pings is a real problem. I think that server behaviors should be different when the pings are unequal.
>I'm not shooting in a vacuum, I'm shooting another player. You kinda are shooting in a vacuum. For at least as long as it takes for your packets to reach the enemy client, you're the only person in the universe who knows what you're shooting at. You're smart, you understand this part. We're trying to find a way to resolve two flows of data that are created simultaneously, but cannot be shared until after some delay. With CSHD, I get to prioritize exactly the things that my human brain wants to be prioritized. Human brains are trained by human experiences to expect instant results to physical actions. CSHD gives me instant results - and so it feels better.