ETPro 3.2.6 - stuff to improve

[deej]

Over at Crossfire, MadScientist has written a great article on stuff that could still be improved / changed in ET. Perhaps it's something the etpro developpers could have a look at?

The discussion can be found here.

[ouroboro]

You shouldn't have shown me this. I just wrote a fucking book. No, seriously. I'm actually embarassed.

[ReyalP]

I love how people think their personal favorite gameplay tweak will "save" et. And then write huge books about it. Face it, no matter how we tweak ET gameplay, the playerbase is gonna be about the same. Unless you really fuck it up and drive a bunch of people off.

The argument that current ET is unskilled is just utter horseshit. Just look at the difference between the top teams, the run-of-the-mill teams, and the average etpubber. If there weren't skill in ET, then the only difference would be those who knew the game mechanics and those who didn't.

It sound like madscientist wants a different game, where everyone is super accurate under all conditions. To him, that is somehow the peak of skill I guess. That's nice, but it ain't ET.

[HarryHirsch]

F1

[madscientist]

It sound like madscientist wants a different game, where everyone is super accurate under all conditions. To him, that is somehow the peak of skill I guess.

May I ask what's wrong with that notion?

I'm not even remotely trying to change ET's fate and we all know its player base will keep on diminishing no matter what. The only thing I did was to write a simplified article about various bugs, missinformation and how a true competitive gameplay would have to be. It's up to the players to decide whether they want the changes and fixes. I'm only pointing at them.

ReyalP,
You asked me about proofs of the turn spread bug over at xfire.be but I remember we had a discussion already about it which ended in that having more FPS as mouse rate would indeed result in a reduction of spread. But I can't seem to retrieve any of the posts since xfire.be recovered from a downtime without any backups.

We also had a discussion about knockback creating prediction errors and thus favoring the player who hits first.

[ReyalP]

It sound like madscientist wants a different game, where everyone is super accurate under all conditions. To him, that is somehow the peak of skill I guess.

May I ask what's wrong with that notion?

Nothing at all. It would probably be a fun game, but if we start trying to cater to every random players idea of what would make ET a fun/better/more skilled whatever you want to call it game, things would get ugly very quickly.

You asked me about proofs of the turn spread bug over at xfire.be but I remember we had a discussion already about it which ended in that having more FPS as mouse rate would indeed result in a reduction of spread. But I can't seem to retrieve any of the posts since xfire.be recovered from a downtime without any backups.

Mouse rate < FPS seems to have an effect on spread, which it shouldn't.
Low FPS (down to 60 or so) gives less recoil than high FPS

Spread is still somewhat affected by FPS (independent of mouserate), where very low FPS (<30) give you more spread/per turn rate, but AFAIK the most glaring problems have been fixed. If this isn't the case, please remind me of where it goes wrong.

Those we have sufficient detail/reproducibility to work on. If there is anything else in this area, please be specific.

We also had a discussion about knockback creating prediction errors and thus favoring the player who hits first.

Yes, knockback and view bounce give a disadvantage on being hit. This as designed. It's fine that you think that this is bad gameplay, but the fact it causes a prediction error doesn't add to the argument.

[ouroboro]

I do recall testing the turning spread / mouserate business a while back and it is indeed true, or at least seemed real enough that I was convinced. I used to have a demo but it's long gone.

Anyway, it shouldn't be that way. A haxed USB port = smooth sexy leetness and shouldn't be punished.

But that nospread business is just garbage and you would see that if you'd only setup haxed etmain server and let the playtests prove it.

[Ragnar_40k]

Over at Crossfire, MadScientist has written a great article on stuff that could still be improved / changed in ET. Perhaps it's something the etpro developpers could have a look at?

Sounds like another "I'm to 1337 for feature XYZ, please nerf ET" thread to me.

When I read such posts I always think "Chess must be the better ET!". Here some feature that make chess überleet:

• Its turnbased - so no advantage for low pingers and no lag.
• Highly hardware independent - you don't need a computer at all to play chess, you don't even need a board - just memorize the current position.
• No wallhacks possible: no walls - no wallhacks, death simple. All that fancy clipping stuff from ET makes chess players laugh.
• Settled and easy rules which didn't change for decades. So no room for exploits.
• Impossible to cheat - come on, when you try to make a forbidden turn you only look like a complete idiot.

[Dersaidin]

Chess is pretty awesome. ET is prolly better though.
I could probably dispute you on every one of those points for chess though :p

Of everything he said the turning spread and FPS dependant recoil are probably the only issues.

Knockback is the only other one I would consider to be fixed, but it is part of the game, and I'd be just as happy to leave it.

[madscientist]

AFAIK the most glaring problems have been fixed. If this isn't the case, please remind me of where it goes wrong.

Excuse me, but isn't a bug that gives zero turn-spread "glaring"? I've already setup an experiment anyone can perform over @ xfire:
-set your mouse @ default USB 125hz
-load a high FPS map (eg. TE_Valhalla or CTF_Multi)
-set cg_crosshairPulse to 1 or cg_drawSpreadScale 2
-set com_maxFps to 125
-Move your mouse around "fast" and notice the huge Spread Scale
-set com_maxFps to 0
-Move your mouse around "fast" and notice that there's barely any spread depending on your maximum FPS.

Spread scales are evaluated on the server and sent to the clients, and for my part this is a considerable advantage for the guy playing at 250 FPS along with 125 Mouse Hz.

Yes, knockback and view bounce give a disadvantage on being hit. This as designed. It's fine that you think that this is bad gameplay, but the fact it causes a prediction error doesn't add to the argument.

If that is your thinking, we should have no use of antilag. If you REALLY wanted to keep knockback in the game, at least acknowledge that it's flawed designed and that it should be unlagged together with player movements? I only figured that disabling knockback would be easier than modifying much of the game code to implement movement antilag.

[ReyalP]

AFAIK the most glaring problems have been fixed. If this isn't the case, please remind me of where it goes wrong.

Excuse me, but isn't a bug that gives zero turn-spread "glaring"? I've already setup an experiment anyone can perform over @ xfire:

Please read the whole post FFS. I acknowledged that bug.

Mouse rate < FPS seems to have an effect on spread, which it shouldn't.
Low FPS (down to 60 or so) gives less recoil than high FPS

Spread is still somewhat affected by FPS (independent of mouserate), where very low FPS (<30) give you more spread/per turn rate, but AFAIK the most glaring problems have been fixed. If this isn't the case, please remind me of where it goes wrong.

Those we have sufficient detail/reproducibility to work on. If there is anything else in this area, please be specific.

The "glaring" is in reference to the remaining FPS vs. spread issues, not the whole topic.

Yes, knockback and view bounce give a disadvantage on being hit. This as designed. It's fine that you think that this is bad gameplay, but the fact it causes a prediction error doesn't add to the argument.

If that is your thinking, we should have no use of antilag. If you REALLY wanted to keep knockback in the game, at least acknowledge that it's flawed designed and that it should be unlagged together with player movements? I only figured that disabling knockback would be easier than modifying much of the game code to implement movement antilag.

Huh ? That makes no sense. The fact that knockback causes a prediction error has nothing at all to do with the value of antilag. Bringing up bogus strawmen doesn't help your argument. Running into other players causes prediction errors too... perhaps we should disable antilag because of that

Knockback throws your aim off in a way you can't compensate for. Prediction errors are a dime a dozen in ET, it's not a big deal. Yes, they are worse for people with high pings... that's gaming on the Internet for you.

In any case, you don't need etpro to disable knockback, since you can already set g_knockback 0 and knock yourself out, as it were

[madscientist]

Bringing up bogus strawmen doesn't help your argument. Running into other players causes prediction errors too... perhaps we should disable antilag because of that

You do have a very -ve feedback on everything I say? I refered to movement antilag, never said anything about aiming antilag. Movement antilag is the process of unlagging movements (doh). This is how it works:
As with aiming antilag which compares each client's view angles with the stored player paths, movement antilag compares each client's position with stored positions of other clients and movers. This means that each client new position is exactly what the server allowed them to in the past.

If you believe that this might result in players not-clipping through each other, you are right but there can't be ANY stationary players non-clipping. So, moving players might no-clip, but as soon as they stop moving the server/client code will move the client that came temporally last on the spot as in a medic revive (you might no-clip through the dead body, but as soon as you revive him, both players will move to their respective positions).

[ReyalP]

movement antilag, never said anything about aiming antilag. Movement antilag is the process of unlagging movements (doh). This is how it works:

Perhaps we are talking past each other. When I talk about antilag, I mean the specific technique of backwards reconciliation for hitscan weapons.

There is nothing like what you describe as "movement antilag" in ET or ETPro, or any other Q3 engine game that I am aware of.

Anything which influences a players position in a way that isn't predictable on the client (either a completely deterministic map object, or commands generated by the client itself) is going to cause prediction errors for that client (it is more correct to call them "prediction misses" since they are part of normal operation, not really an error.)

Getting shot is obviously unpredictable. You could try to fudge things so that the client who mis-predicted gets to shoot from where he thought he was, rather than where he really was, but that is a can of worms I wouldn't want to open.

Anyway, as I said before, if you don't want knockback, convince your favorite league to implement g_knockback 0. Don't ask etpro to change the default simply to promote your personal preference.

I'm not trying to be negative, but I think you are making more of an issue out of knockback than it really is. I do appreciate your bringing up the spread bugs again. They are something that should get fixed.

[ouroboro]

I do appreciate your bringing up the spread bugs again. They are something that should get fixed.

\o/

That's good to hear, because despite the turning spread business, I doubt you'll find many players who know about mouserate who don't increase it. It's so much smoother and more pleasant, that it's worth the increased turning spread for most people.

Still, if it can be fixed, all's the better.

[mortis]

I don't increase the rate, haxor!


j/k