-Coulum-
-
Content Count
1790 -
Joined
-
Last visited
-
Medals
Posts posted by -Coulum-
-
-
Well, the recent sitrep says body armor sim will be improved.
Good news, I am interested to know just how much this will change things. When you have the time to aim precisely I predict it will make a big difference. But in common firefights, I predict we will find many situations where units take several hits before dying (abeit probably fewer). Not that I am against that. modern body armour is very effective. My main gripe is the speed and behavior of the ai reaction.
I was talking about CQC and the situation I ran into was me coming through a door, putting 2 shots into a guy (thinking that should be enough), and having him turn around and kill me.
Doesn't sound unreasonable to me. For unarmoured targets let alone armoured. I do think the speed at which hit ai and players react is too fast. But if you assumed 2 shots was enough and moved on, then you deserve to get shot. Shoot it till its dead. Humans are surprisingly resilient. Unless its too the head, I wouldn't rely on any number of shots instantly taking someone out of a CQC fight.
-
Enter-able buildings are important. They are a huge sources of concealment on altis. I suppose the jungles of tanoa could help make up for that lack... will the level of entrance planned be on par with chernarus or takistan though?
I hope the reasoning isn't solely because they can't fill interiors with furniture. I'd be very disappointed. Know there was some fuss about it when altis was released. I equated the serious complaints to thankless whining. I predict that more than 90% of players would rather fully enter-able but empty houses rather than locked houses. Probably more. All the freedom on altis has really changed the game. Maybe I am mistaken though...
I do sincerely hope BI reconsider. I understand if its just not within their capabilities.
-
1
-
-
the throwing of smoke grenades.This is the first actual behavior decision implemented. Hopefully more will follow.Sweet, will have to switch over to dev branch again this weekend to give it a go. Honestly I am skeptical because most ai mods with smoke did more to reveal their location than hide it IMO. Hopefully vanilla will do it right.
Looking forward to the "more to follow". Thanks Ai guys for the continued hard work!
-
I am also interested to see how the jungle turns out, especially with the ai.
Solution to these problems is unknown as there aren't really any games that pull off the ability for AI to appear realistic in both dense cqb areas and long range firefights up to 800m that I know of.I believe the thing to tune is not spotting distance, but spotting time. Ai should still spot you at 300m in the jungle (when in LOS). Just it should take alot longer. Like 5 seconds on airport vs 5 minutes in jungle. Line of sight should not mean instant spot (pretty much now). Rather line of sight starts a countdown. When complete the ai spots you. Length of countdown depends on combo of target's terrain, uniform, stance, movement, distance. Ai lose line of sight, countdown counts up until its regained.
That, combined with your mentioned spotting animations/investigative behavior during "countdown" would go a looooong way toward making the ai less terminator and more human.
With the current system I'd suspect Player would be last man on totem pole to know what and who the hell his squad is shooting at and where.Agreed. You would only see the enemy when they're all dead, or when you're the last man standing. Not so much because of the ai's shitty communication to player (I am all for improving) but rather Ais' overly effective communication among themselves. That's why players feel so clueless. One ai spots something, soon the squad spots it. Player is still wondering who's 9oclock. Imagine the ai each had to spot enemies for themselves. Add in the "countdown" idea^. Players would have a fair chance. Ai Firefights would be longer.
Anyhow my two cents regarding ai in jungle, I am eager to see what comes. I am hoping BI's decision to go with jungle is because they are confident they can make changes to the ai that will make that environment enjoyable.
-
I mean, is the size of the bullet the determining factor in the loudness of its sonic boom?I am no expert but I believe yes. The loudness of the crack is dependant on the volume of air being displaced. Bigger bullet is pusshing aside more air... But take that with a grain of salt, maybe someone else is more knowledgeable.
-
Gunshots are waaaaay too quiet now over distance. In a firefight between AAF and CSAT happening 400m from me, all I can hear is the Mk200 and some really soft pah pah pah that is the Mk20. It sounds like the fight is about 1km away. The dropoff distance was too short before this update, but now it's super short.The new cracks and indoor attenuation are really good though.
Pretty much this. Gunss are way to quiet over distance. Plus bullet impacts are too quiet in my opinion.
-
It worked in 1.40 with the same MODs, so it need to Work in Patched version too ! Its also only Weapon and Radar Mod wich have nothing todo with the AI.Still it would help alot more if you could find away to reproduce it in vanilla. I personally haven't been able to as of yet but I am keeping an eye out for it.
-
Alrighty, thanks for the clarification.Sry, not in this case. This was mainly about using a correct zeroing. But a lower accuracy with sights without magnification should be achievable via aiDispersionCoefX/Y properties in the fire mode class of a weapon. If you find an odd situation in the game regarding this let us know, we can try to improve it. -
•Added: AI Suppression commands: getSuppression, setSuppressionAwesome!
•Fixed: Weapon attachments properly affect AI aimingDoes this mean ai with iron sights will be less accurate than those with magnification? Sounds promising!
So there's going to be a data lock. Is "AI running to cover when suppressed" and "AI telling other AI groups about enemies" still going to be added?Doubt its gonna be improved any more before/for 1.42.
-
True thatShould be fixed nao
Sweeet!
Several tweaks and balancing. Thanks for your feedback! Situations where AI's were few meters away from each other, couldn't hit and depleted all ammo should be history now.- Fixed AI firing too much or even only off the target. A suppressed AI has much higher probability to miss the target, but that doesn't mean it can't hit it at all
- Increased the time AI needs to recover from being suppressed.
- AI can occasionally suffer a lower rate of fire when pinned down.
- AI should also usually take some time before opening fire after turning (depends on the target's distance)
Good stuff. Still think Its a bit too difficult to suppress with 5,56 and even 6.5's. You have to keep a very high rate of fire up and be putting your bullets within 2 metres of them. Such close shots should be more effective, regardless of calibre.
Also I am not a big fan of the decreased rate of fire when pinned down. Basically you unload on an ai and he just sits there and stares at you for a second or so. It looks silly/glitchy, especially when you are within 150m of the ai. If he were seeking cover or something it would be good, but he is just sitting there. He should still be firing. If anything, assuming he had no where to hide, being under fire would make him more trigger happy as he shoot's in panic.
However, for longer range engagements, I can see this mechanic being more acceptable. If it could be made so distance to the shooter effects how often ai suffer the ROF penalty (or tweaked so more distance is required (200m+)), I think it would look and work much better.
Keep up the good work devs!
- Fixed AI firing too much or even only off the target. A suppressed AI has much higher probability to miss the target, but that doesn't mean it can't hit it at all
-
According to the latest weapon config guidelines documentation, this is already the case for attachable bipods, while presumably the weapons with integral bipods already or will have theirs already accounted for. (A weapon with built-in bipod just needs hasBipod = true; and the name of the relevant memory point in its config, while the attachable bipod's config has an inertia value and its ItemInfo subclass has a mass value).No I mean when you are actually deployed, you should have more inertia than when just rested or shooting normally. To me inertia represents how quickly/easily you can change your aim. It is easier to change your aim by simply rotating your upper body then it is shuffling around a pivot point. So deployment should have even more inertia than regular offhand aiming.
so the deployment without bipod should be less effective than with bipod - more sway , same recoil as resting , less stabilityYes I totally agree.
-
Cover finding when suppressed could be under development still ? My post was mainly to highlight that cover is part of the feature set for the DLCWell I hope your right. And it probably is being developed for the long long term. But My understanding of "ai will better react to incoming projectiles and take cover" is: ai will now perform its um, poor, cover finding techniques when bullets passby. Rather than before when passing bullets wouldn't trigger the cover finding techniques. The cover finding isn't any different, just when it is triggered.
But maybe I am being pessimistic. I hope so. It would be awesome for ai to better use cover. And with the new resting surface detection and deployment animations I think there is alot of potential to make ai able to better detect and really hug cover. Twould be awesome. But frankly I don't think its in the scope of marksman.
Regardless the suppression stuff is a great improvement already! Though I hope they can fix the ai's extreme inaccarucy when suppressed before it hits stable.
-
I see resting as almost useless now, and I can barely even tell it's there without some sort of visual and/or audio indicator. It's nowhere near stable enough to depend on for accurate shooting in combat; deployment will be used 95% of the time unless in a VERY time-sensitive situation. If anything I would make resting even more stable (and add indicator of some kind), or just leave as is.IMO deployment should take longer to set up (especially for bipods) and should have some sort of inertia or rotational speed limt to preventing quicker snap shooting. That would make resting a valuable option while making the mecahnics more realistic.
In reality deployment helps stabilize your weapon, for sure... but it makes rapidly changing aim harder/slower. This is not the case in game. There is little penalty to being deployed when there should be. In reality if you want to shift your aim 45 degrees while deployed with a bipod on a sandbag wall, you literally have to sidestep or at the very least lean an uncomfortable amount to do so. You are not going to have steady aim while you make that movement and it is going to take longer than simply rotating your arms/upper body (as you would unrested/deployed). In addition it will take a moment to regain the steadiness provided by the bipod. None of this is represented ingame. Just the sway/recoil bonus. You deploy your bipod and you have a huge range of motion with minimal inertia or limitation to how quickly you can turn. Plus you can set up your bipod very quickly, even while aiming down the sights, while keeping this steady aim the whole time - yeah right. Try deploying a bipod while looking down a magnified optic. Your aim is going to be bouncing all around.
Obviously I believe that, although the positives of bipods are represented in game very well (accuracy and recoil), the disadvantages (less manoeuvrability, slower manoeuvrability and set-up time) are pretty lacking. It makes bipods/deployment very overpowered, and makes resting a seemingly useless thing. There is no reason not to deply you bipod if you have cover nearby.
Make bipods have an inertia penalty, take longer to setup, and take you out of "aim down the sights" while setting up and now you have a realistic reason for choosing not to deploy and taking advantage of weapon resting. You also have more realistic limitation to deployment itself. My opinion. Interested to know if you guys can relate. Deployment is awesome. But I think it comes with lot of realistic pros, while ignoring many realistic cons.
-
Well suppression is great, but its a bit to harsh on the ai right now because it makes them shoot wildly left right high and low of you. A suppressed ai simply cannot hit you at all. It makes it a bit unfair for the ai, and takes away alot of fear and risk for the player because he knows his suppression is always 100% effective, and he will never be shot by someone he is shooting at rapidly.
Once that gets fixed though I think things are going to be hella awesome.
Very much so, and AI will find COVER too as part of suppression (I was under the impression that may not be in scope but apparently is) !Well in theory. For the most part I have not noticed any exceptional use of cover by the ai. I think its pretty much the same old. However I have noticed that they will avoid certain routes if you hose those routes down with fire. Overall I think cover taking and ai is something that still needs alot of work.
-
Is it possible due to the supression effect, the opposite force starting shooting with dispersion creating also a supression with ours forces? In Combined Arms Showcase the blufor starts to be without ammo too fast, looks it is created a effect movie like this:http://i.imgur.com/WRFXYFi.gifYeah that's basically how it works. One shoots at the other, causing inaccuracy, the other returns inaccurate fire but still enough to suppress the first. And then a stalemate kind of ensues. This is alot like reality though, but you have taking cover and hiding also thrown in the mix (AI don't really do that). Its why outmanoeuvring the enemy, having artilery/airsupport/armour, or just bringing bigger more accurate fire power is really important.
However suppression related ai aiming error is a bit broken now. Basically suppressed ai simply can't hit. Like not even a remote possibility. They always shoot to the far extremes of their "cone of fire":
Some unwanted behavior making the AI fire only in the extremes of the imprecision cone slipped in :/Leading to situations like your gif, or like MAXZY pointed out:
Once that is fixed things will probably be alot better. Currently its pretty much unplayable IMO. Way too easy.
-
I like what the guy in the vid suggests. I think if the transparent housing wasn't just a photoshopped image (as in the vid) and instead actually reacted to lighting and what not it would look reasonably good. Probably should have added more blurred to make it look more similar to what you see at 2:23 in the vid. That would actually be most immersive to me.
-
Some unwanted behavior making the AI fire only in the extremes of the imprecision cone slipped in :/Well it is good that this is not intended behaviour. I look forward to seeing how it works when this is fixed.
It's now possible to keep the AI suppressed by putting a 6.5 bullet approx. almost every second closer than 2m from the AI.And that's the thing. 1 round a second is a high rate of fire to be maintaining. And for a 5.56 its going to need to be even higher. Teams will be running out of ammo in short order.
Basically I think that in the "quantity X quality X proximity" relationship you mention, proximity should have more weight/be more important. A shot that wizzes past an Ai's head should cause more than a single second of stress, regardless if its 5.56, 6.5 or 7.62. Even just making it two seconds instead of one would make it a tonne more reasonable. Just my opinion of course.
-
When your eyes are further away shadowing happens more easily than up close. There's two sides on this because the scopes aren't real 3D but only zoomed in.True but theoretically in game you are holding the sight at the same distance in both examples in the vid. Only thing that is changing is the power of the magnification. And more magnification means more sensitivity to that shadowing. But yeah, the problem stems from the magnified scopes not being real 3d sights, but rather just zoomed in/up close non magnified sights.
-
What do you guys think of what this guy says:
-
I know that its stil wip, but the bipods are looking good so far. My tweaks would be:
- Standing or crouched while rested should have more sway and recoil. Sitting and prone are pretty reasonable
- Weapons without bipods should be alot less stable while rested.
- Inertia should not be decreased while rested. Shooting from a rest makes it possible to steady your aim. It doesn't increase the speed at which you are able to aim. Should still have to use controlled steady weapon handling even when rested.
Anyhow just my thoughts on the resting so far. I haven't kept up with this thread so I am sure alot of it has been discussed already.
The dispertion value, should in my opinion be treated as the weapon's own inherent accuracy. It should have nothing to do with the shooter. Imagine putting a weapon in a fixed stand and shooting a group. There will be some dispersion because of barrel harmonics and such things, and that's what the dispersion value is for.Yep. Dispersion is for stuff that in reality is out of the shooter's control imo.
At 3:20 seconds, you can actually see him firing around the same ROF as in game, and it shows a target and it's dispersion. That grouping... Is actually better IRL than in game. Which is actually quite frightening...That is very impressive/scary. I do think that 9mm+ would produce considerably more kick though, and that being in any stance other prone would be much harder, even with bipod deployed. But never personally fired an MG. Also it can be hard to see just how tight his groups are from such a camera angle.
- Standing or crouched while rested should have more sway and recoil. Sitting and prone are pretty reasonable
-
So I don't know if this is the right place to bring this up. I think it is to do with inertia. The scope shadows that appear on the MRCO are awesome and really helps to convey inertia. I just which that it would work on the sighs with magnification. This shadowing is exactly what they need. It does seem to be there but it is significantly less sensitive - whereas in reality the opposite would be true, as magnified sights are more sensitive to that shadowing due to misalignment. Really hope that is changed. And that the shadowing appears after recoil + bringing the sights up as well. Here's a vid showing off the shadowing if you missed it:
-
I am for a blurred/transparent holo sight housings. It would be hard to pull off though with out it looking "gamey", even though it is more realistic.
-
I am playing on expert Ai. The aiming error is definitely too harsh on the Ai while suppressed. I tried to demonstrate just how bad they miss by when suppressed:
This is a CSAT Rifleman at 100m. As you can see, he's missing by up to a dozen metres and sometimes shooting the ground right in-front of him. I feel that's too much. In this situation I would expect the shots to be no more than 3 metres off target. If I were to adjust the aiming error due to suppression, I would make it around 30% of what it is now. Even then I think it might still be too harsh.
We actually tried to achieve the opposite - one random shot shouldn't rly suppress the AI. But a sustained fire is what puts you under pressure. With different calibers and target unit's skill the challenge is to find the right rate of fire to keep the enemy suppressed while not wasting all of your ammo at once. While in one situation three 5.56 rounds, within 2 seconds, in about 6m from the target are enough to make the AI combat inefficient couple of seconds, some other AI may just not even unless you keep spraying it or use other means of persuation.I understand what you're going for then. My concern is the "sustained" rate of fire you need to suppress is too high. Especially for 5.56's or 6.5s, keeping fire superiority will drain you of ammo very fast. Anything less than continuous bursts of fire seems to really have little effect. This does introduce a cool logistical challenge, but that challenge comes a bit premature if you ask me. I believe that realistically it is possible to effectively "suppress" an enemy with smaller rounds without using such a volume of ammunition. More should always be better, but slow steady accurate fire should be a viable suppression option as well.
Ideally first contact would require the heavy volume of fire it does now, but then after that, slower (accurate shot every couple of seconds) would be sufficient to keep the enemy in a stressful state. Does that stray from your guys' vision? Perhaps making the proximity of the passing bullets have more weight might help achieve this balance (initial shooting is less accurate, high ROF required, as enemy are spotted shooting becomes more accurate so lower ROF is required). Thoughts?
Can you?Nope. But I would give those high grasses a good hose down.
Regarding ai behaviour while suppressed, I get the feeling that its on the devs' wishlist. They know how it could be improved. Its just a matter of doing it.
-
Awesome stuff, really happy to have this feature in. Didn't get much time to test it fully but I would agree with the "its too easy/effective" comments. From the little I played my suggestions/observations would be:
- Generally the ai accuracy penalty needs to be alot less harsh. They should still be able to return "threatening" fire, and be accurate at ranges under 50m. At 250m they should not be missing by more than 2 or 3 metres when suppressed. I know suppression comes in different strengths but generally it seems to make the ai too handicapped. At 100m metres, they are not shooting even remotely close to me. There is zero chance they will hit. There should be. Make the error less.
- Ai should suffer more aiming error than it currently does from sporadic single shots, and less error than it currently does from prolonged fire. Aiming error should jump quickly but not continue to rise to ridiculous levels after prolonged fire.
- Similar to above, but for time. The time ai should suffer from sporadic single shots should be longer (2-3 seconds) while the maximum time they should suffer after prolonged fire should be shorter (like 5 or 6 seconds max.)
Why not just bring back the Arma 2 feature of decreasing AI rate of fire when suppressed? It's much more realistic than maintaining the same ROF with horrible accuracy. The AI is never going to be able to take cover properly, so it might as well suffer an ROF drop in addition to the accuracy decrease. It was probably too severe in Arma 2, but it makes sense that someone would decrease their volume of fire when taking alot of incoming rounds. That's the entire point of suppression tactics in the first place.I know what you're getting at - in reality suppressed enemy return less fire because they are hiding rather shooting. But I disagree that simply changing the ROF is the way to go about representing this. It should be done by ai actually hiding. I know that doesn't exactly work right now, but ROF is not the way to compensate. Because then ai are just going to be standing in the open casually/slowly returning fire. When in reality, you're caught in the open the reaction would be exact opposite... like CaptainObvious has says:
I beg to differ, the reasonable thing to do while suppressed is to seek cover and return fire, return lots of fire.If they can't/don't seek cover they should be shooting like mad men.
Looking forward to seeing this feature really fleshed out. I think it could pretty much be the end all be all to ai terminator/accuracy problems if done right.
- Generally the ai accuracy penalty needs to be alot less harsh. They should still be able to return "threatening" fire, and be accurate at ranges under 50m. At 250m they should not be missing by more than 2 or 3 metres when suppressed. I know suppression comes in different strengths but generally it seems to make the ai too handicapped. At 100m metres, they are not shooting even remotely close to me. There is zero chance they will hit. There should be. Make the error less.
The Struggle - Enter-able Buildings Vs. Non/Partly-Enter-able Buildings
in ARMA 3 - GENERAL
Posted
WHAT! A city of enter-able buildings devoid of interior furniture is way more useful than totally un-enter-able buildings. Huge difference. The former might take hours to clear out of enemies with intense firefights, CQC and house to house fighting,despite armoured, artillery or air support. The latter can be cleared in short order, mostly from afar, even more so with armored, arty or air support. Un-enter-able vs enter-able buildings is the difference between a nonfactor infantry platoon in a town, and entrenched pain in my ass infantry platoon making me pay dearly for every inch of ground.
All the enter able buildings in Arma 3 thus far has changed alot about the gameplay. It would be a real shame if it was changed on Tonoa.