[FS Zenity]

crazyyhedgehog, on 12 July 2015 - 05:03 PM, said:

Good job!

But I have a few questions now.

If we are revising weaponry, spread/recoil/damage I have to ask:

1) Will we finally be able to do double damage? if two players line up - we can shoot through a player and hit the player behind him with reduced damage?
2) You are fixing up LR300 and M4 to make it clearly more superior, then I'd like to demand to remove penalty on SR8 to even it out (you can keep the knock back, remove zoom out and moving / jumping penalty). I have played Call of Duty 4 (promod / competitive) for years and let me tell you, nobody complains about scope being superior. Here is an example:

http://romtruck.pw/videos/cod4/

3) And yes, I will always complain about SR8 being stripped down its underwear... TwentySeven hated me for a bit ;)

4) GOSU was an amazing test @ 3.7(.7) which was fun. No, it was a lot of fun... Ah memories...

Cheers,
]BoS[HedgehoG

I'd say it's too early to be concerned about any such details. We won't even have the exact same weapons in all likelihood. However, from what I can tell, the general feeling in the team right now seems to be very pro movement and anti movement penalty. We may experiment with some ways to turn up the heat. :) My personal feeling is that hitting somebody in mid-air while scoped and strafe jumping at 60km/h is already challenging enough as it is, so it does not need to be discouraged by adding artificial penalties. However, we also certainly don't want sniper rifles to reign supreme in a fast paced action shooter, so it needs careful deliberation.

As for GOSU, does anybody remember what it did exactly and why it wasn't adopted in the end? From a quick google, it seemed to have been rather controversial at least. It also seems to have increased weapon damage which I don't think I would agree with. You already die extremely quickly in URT, and I wouldn't want to turn it even more into instagib. :)

[crazyyhedgehog]

Zenity, on 12 July 2015 - 05:48 PM, said:

I'd say it's too early to be concerned about any such details. We won't even have the exact same weapons in all likelihood. However, from what I can tell, the general feeling in the team right now seems to be very pro movement and anti movement penalty. We may experiment with some ways to turn up the heat. :) My personal feeling is that hitting somebody in mid-air while scoped and strafe jumping at 60km/h is already challenging enough as it is, so it does not need to be discouraged by adding artificial penalties. However, we also certainly don't want sniper rifles to reign supreme in a fast paced action shooter, so it needs careful deliberation.

As for GOSU, does anybody remember what it did exactly and why it wasn't adopted in the end? From a quick google, it seemed to have been rather controversial at least. It also seems to have increased weapon damage which I don't think I would agree with. You already die extremely quickly in URT, and I wouldn't want to turn it even more into instagib. :)

This was for Urban Terror 3.7 and this file is server side only. Check it out for yourself!

http://romtruck.pw/q...zzzSidGosu7.pk3

GOSU increased damage on some weapons, such as LR300 and M4 were most noticable and completely removed SR8 penalty (yes you could snipe full strafe speed and be accurate) this made a huge impact on CTF and TS back then, it was so much fun.

I miss those days, playing a game that was truly competitive. Now if you tell me some interesting stuff is coming, I might want to get my butt back on QA team :)

This post has been edited by crazyyhedgehog: 12 July 2015 - 11:03 PM

[iCu*posssesssed]

ive been using only sr8 now for about 2 years. the skill level has completely dropped off the last few years and honestly the lr/m4 just arent challenging in that any noob can spray and get good hits. and thats why i use the sr8 now, for a challenge. this and being able to do things within strict paramaters that only a few out there can achieve. ive recorded so much "wow wtf" gameplay this last 6 months, that i feel completely removing the movement penalty removes the "sweetness" of sr8 kills. randomness and high skill shots are absolutely necessary to have with the sr8. its challenging, and very rewarding once you really "discover" the weapon.

i would say sr8 could use perhaps a tiny boost as the m4/lr is still over powered comparatively speaking, but removing the movement penalty is bad idea. right now it requires skill and understanding of how to use it for it to be effective, but a good-great sr8 can be the difference in team survivor... and i know because i played a long time with the best sr8 the game has ever seen, tarquin of iCu*/horsies (ok maybe snoppis or dracostian could be considered as good, but didnt play as long or win as many titles either). removing the movement penalty in my eyes makes it a very overpowered weapon.

and i really dont want sr8 to become a silly weapon, either. i feel the current structure is fine. hitting shots in mid air might seem fun at first but it would tarnish the weapon. my recommendation for sr8 is that the bolt time is decreased a bit (50-100ms?) so there is less time between shots and/or the ability to switch to secondary/sidearms is also a tiny bit quicker. this keeps the weapon in tact, to be only truly a dangerous weapon in the right hands. id also like the sr8 and other sniper rifle bullets to be able to puncture through enemies and pass through other enemies. that would also make more sense and be slightly more beneficial for the gun, and create more highlight reels.

but honestly the main issue with the sr8 isnt the sr8, lol, it is the fact that the lr/m4/ak are overpowered a tad comparatively speaking. would be nice if there was some damage drop-off / inaccuracy after certain distances (that is a lot of data editing, i know)

This post has been edited by posssesssed: 13 July 2015 - 12:46 PM

[FS Zenity]

Ah yes, I've been thinking about adding more damage falloff. I think that's a better way to balance the range of a weapon than to add more spread. Right now URT has a rule that the first bullet is always 100% accurate (when standing still at least), so to make any weapon less dangerous on long range would require either reverting this rule or adding damage falloff. The latter seems more interesting to me.

Automatics will (or rather should IMO) always be more generally effective than a sniper rifle, but the perfect balance for a sniper rifle is that it adds something to a team that automatics cannot provide, so you usually want to have at least one in the team. Snipers in URT also benefit from being able to carry a rather powerful secondary weapon (though we might want to reconsider or adjust this feature).

That said, it's not just about balance, but also about making you feel awesome using that weapon. I'm not afraid of making things "silly" and balance it in another way instead, as long as it's the stylish kind of silly and still rewards skill. If it would fit into an over-the-top hollywood action movie, then it would fit in Urban Terror.

Just a random thought came to mind, what if the movement penalty of a sniper would not actually be random, but instead offset the aiming vector by a predictable amount based on your velocity. This would still be impossible to control perfectly (without hacks), but would reward experience in that you can get a feel of where the bullet will actually land.

[Mod Iye]

i personally would just keep the movement penalty the way it currently is. system wise at least. maybe you could just lower the amount of spread a little.
IF you want to buff the sr8, just remove or lower the no scope penalty, because that one is ridiculously high atm. like, you stand right in front of someone, stick the gun into him and look at their center torso, but can still miss the shot. surely that partially is a problem of the current engine, but definitely is something that shouldn't be adopted into HD directly.
For a "static" movement penalty algorithm: i wonder how that would fit the PSG, or whatever the autosniper will be in HD. like you could fire 1 shot, and then readjust, and hit all following shots. Also, i just ususally stop for half a second to fire the sr8. so something like being able to fire on the move, with a greater chance, just doesn't seem to attractive/worth it to me. Like i would still stop to fire.
its still only a chance, because, as you said, we probably wont be able to perfectly predict it, same as the dragshots you can do, or the firing just somewhere around your enemy and still hitting with a good chance, just out of a "feeling" for the game mechanics),


And for damage falloff. well. maybe? 5% at 100 meters? maximum?
Like you dont want to make every auto/nonscopable/whatever gun useless at long range. you might even add some to the sr8, but like 5% at 1000 meters then :)


all the questions about the sr8 basically come down to one point IMO:
do you want to make the game aim based only, like a good CS player could just hop over, adjust his mouse settings and start top fragging everywhere, or do you want it to be more of a specific game skill based thing, like good aim helps ( a lot ), but there is still that edge you have from knowing the game & its mechanics by heart.

just my 2 cents. :)

tldr1: actually, sr8 doesnt even need a real buff. imo. slight buff would help it overcome the problems you get into when not 110% focused. Also: what about other snipers?


//edit Iye:
//regarding the client side hit detection:
well, at first i was like "WTF hacks incomming"
then i was like, hm. if the server checks it, it would be any problem.
next i started to think about it a little more. like, what if you do it 2 (3 with the server) sided?
(please note that i totally ignore latency and such in this "thought experiment")
like, the shooter says if he "thinks" he hit, the target says if the "thinks" he was hit.
if both agree, everything is fine, the server doesn't even (have to) check and we are good.
now, what if they don't both say the same, because you made the algorithm to be fast, and thereby a little inaccurate. (or maybe because one of the sides simply hacks).
you would need to decide which one to trust. but you cant do that hard coded. like a hack for not getting hit is just as good as one to hit "every" shot. or maybe your algorithm just differs. You cant trust the shooter by default.
Now you are back to a server side check. Maybe you could weight the trust in clients by average ping,...

Another case ( "1 client detection" )
Lets say the shooter tells the server if he hit, and the server only does a raw check.
Now someone with a ping of 300 hits me on his screen, but on every other screen he wasnt even around the corner.
Do you trust the compensation tools in the networking part? or is it just like, the client side hit detection removes this problem totally. because i think it doesn't. it would actually create a gap between the hit detection system and the visuals you see. but thats just what you tried to fix by using the physical asset has the hit mesh... ?

tldr2: very confused. much special cases. not so wow.

This post has been edited by Iye: 13 July 2015 - 03:03 PM

[FS Zenity]

Asking the target whether they think they are hit isn't really feasible I'm afraid. :) You already mention that you didn't take latencies into account, which would be a big issue (how long would you wait until you confirm a hit?) but it's also illogical because by the nature of the beast, the target will never think that they are hit, because their state of the game is very different to the one of the attacker. That's the whole crux that lag compensation / client side hit detection tries to deal with.

This is not an issue though. As I mentioned, if we really end up needing it, we could simply do full server side lag compensation for verification, only allowing for minor discrepancies. Should things go really bad for some reason, we'll just revert to server side lag compensation only. But there is no reason not to give client side hit detection a try first, and I am quite convinced that it will work out just fine anyway.

[KroniK]

Wow, Love the double client side check concept Iye, and letting the server give weight by ping isn't a bad idea.

I personally would like the game to stay competitive at fairly high pings (around 120ms). I hate how some games are super hard on players who have a ping any higher than 60-80ms. Its very hard for me to get those low latency connections because of my currrent location, so I'd rather have it balanced at a higher ping average.

[karnute]

Zenity, on 13 July 2015 - 02:31 PM, said:

Ah yes, I've been thinking about adding more damage falloff. I think that's a better way to balance the range of a weapon than to add more spread.

I agree. Some time ago, I propossed to change SPAS into a one powerfull bullet (instead of pellets) but with a lot of damage falloff with distance, and also the possibility to apply the same for all other weapons in UrTHD (with different parameters of distances and reductions).
In that post, I tried to explain the idea with excess of details, but essentially it was that every weapon can have a near range where it makes its full damage (0% falloff) and from that point and beyond the falloff starts to increase linearly so that there is a (long?) distance where damage is 0 (100% falloff), so beyond it there is no hit. This allows to define different rates of falloff by means of two distances ("near" for full damage, and "far" for no damage and no hit) and linear falloff between them.
If both distances (near/far) are carefully selected for each weapon (pistols short/medium, snipers long/very-long, etc.) it could provide a better way to balance than the spread.

[crazyyhedgehog]

Oh come on.If you want a challenge, use PSG1. SR8 needs to be more flexible. It should be a support/rush weapon, not a sit back camp and wait weapon. I am proposing to reduce no scope spread and reduce moving spread to at least what it was in 3 series. Which was no moving penalty while walking.

This post has been edited by crazyyhedgehog: 14 July 2015 - 01:50 AM

[TehReaper]

Awesome work! This is exciting!