CS 1.6 first bullet accuracy (valve blogpost is posting wrong info)

[Labyr!nth]

Originally Posted by HeatSurge

So, the decals lie? That's weird/inconsistent that they coded the game to draw the decals in a place where they don't actually land server side.

Does cl_lw also work on CS:S and CS:GO? Can someone do tests on them with that turned off/on?

http://forums.steampowered.com/forum....php?t=2610958

For anyone else interested how and why it works in the way it does, read my OP edit.

[lpkane]

After messing with this map for about 30 minutes I honestly think the way the spread works in GO when you are crouched should actually be what the spread is when you are standing. If you try it and look at what happens it would correct all of the current problems people are having with it, single shots, bursts and long range effectiveness.

Give it a shot and if need be I will make a video.

P.S. It still needs to be tweaked for better grouping but this is a step in the right direction.

[HeatSurge]

Originally Posted by Vendo

Yes, the decals lie. Also: you cant do that online cuz of massive lag you´d get. Letting the client lie to you speeds things up and prevents lags.

( and you´d know all that if you´d have a clue about cs....)

That being the case doesn't necessarily mean it's the case in CS:GO though. Do the decals still lie in CS:GO? Also in Source? Maybe the shot going dead center was actually a bug in 1.6 and the designers didn't intend for it to be that way - was the center shot or the decals the design intention?

The inconsistency between the drawn decals and what actually goes to the server is really puzzling to me, and it seems like some kind of bug (either in drawing decals or the server registration).

P.S. I've said that having a clue about previous versions of CS is not relevant at all. This is interesting info though.

[nyran125]

i fully understand what you guys want.. for competitive play, you need bullets to go exactly where you want them or its too hard or its not fun enough for you ..

Some people might like that hardcoreness more realistic difficulty. Competitive people obviously dont.

I get it.

Good luck

Not being a competitive player, i dont mind it being really difficult.

[Krovvy]

If I remember in Source they do lie (correct me if I'm wrong), I don't know about Global Offensive.

[Labyr!nth]

They do yeah. Blood on the face of an enemy, when you haven't actually hit them etc.

[lpkane]

Originally Posted by nyran125

i fully understand what you guys want.. for competitive play, you need bullets to go exactly where you want them or its too hard or its not fun enough for you ..

Some people might like that hardcoreness more realistic difficulty. Competitive people obviously dont.

I get it.

Good luck

Not being a competitive player, i dont mind it being really difficult.

There is a huge difference between hard that can be controlled and hard that is random and can't exactly be. We still want it to be hard but at the same time something we can control, currently it's not. A player wants to know the reason he missed his shot is because he was not aiming properly, not because the shot has random hitscan that does not place it properly.

[Saijin_Naib]

I stated it on the last thread where the guy "proved" GO was accurate. You can't rely upon client-side decals, they mean nothing. Only server-side determination matters.

[HeatSurge]

Originally Posted by Saijin_Naib

I stated it on the last thread where the guy "proved" GO was accurate. You can't rely upon client-side decals, they mean nothing. Only server-side determination matters.

Is there a way to test the server-side determination in Source and GO as it is right now, with decals drawn from the server-side hit detection (instead of the client-side)?

I'm not sure this entire discussion is relevant to the development of CS:GO because if it was a certain way in 1.6 and Source doesn't mean that it should be the same way in GO, but it's really piqued my curiosity.

[nyran125]

this video here, should give you a pretty good idea that Valve is listening to you but they are trying new things with CS:GO and it REALLY sounds like they want to go away from 1.6 in this video.

http://www.youtube.com/watch?feature...&v=2fNH7DwLieQ

this video almost makes me think theyve turned the forums off lol.

NOTICE HE SAID THE WORDS REALISTIC? With quotation marks?

i recommend for you to have them up for a 1.6HD edition.

[rec]

Originally Posted by nyran125

Good luck

We'll need it, if this game isn't about skill any more.

[ido]

Thanks for taking the time to make this video!

You're right, cl_lw 0 turns off the client side prediction. But the server is authoritative. cl_lw 1 decals accurately portray the hits that happen on the server and not cl_lw 0, because with cl_lw 0, the client no longer runs the prediction of what the server is actually doing.

To demonstrate this, I replaced the target with a grid of 4x4 func_breakables and ran the same shooting test.

http://imgur.com/KT439

You'll note that all of the decals are in the center because the client didn't bother calculating the actual spread that happens on the server. But the actual bullets, the ones on the server, broke brushes all over in a pattern that matches the original image we posted.

Another way to confirm this is to have 2 players on a server, have the one with cl_lw 0 shoot right next to the other's head repeatedly. Despite the fact that the decals always appear next to his head, eventually you'll kill him with a stray bullet.

You should run the same experiment as I did and let us know what you get!

[Labyr!nth]

Originally Posted by HeatSurge

Is there a way to test the server-side determination in Source and GO as it is right now, with decals drawn from the server-side hit detection (instead of the client-side)?

I'm not sure this entire discussion is relevant to the development of CS:GO because if it was a certain way in 1.6 and Source doesn't mean that it should be the same way in GO, but it's really piqued my curiosity.

Yes. Type cl_lw 0 in console. And cl_lc 0 too, though I think doing the first command implies the second. This turns off all prediction and compensation, taking info straight from the server.

[HeatSurge]

Originally Posted by Labyr!nth

Yes. Type cl_lw 0 in console. And cl_lc 0 too, though I think doing the first command implies the second. This turns off all prediction and compensation, taking info straight from the server.

Can someone who owns all 3 versions of CS make a video (or 3 videos) with comparisons of server-side decals (with cl_lw and cl_lc 0) on the range map at the longest distance with the M4/AK? That would be lovely. I do have CS:S and GO but I'm kind of lazy :P . Maybe tonight.

Edit: Nevermind, ido posted something above which makes the above redundant/unnecessary!

[LuauDesign]

bump.