Jump to content


  • Content count

  • Joined

  • Last visited

  • Medals

  • Medals

Community Reputation

119 Excellent


About instagoat

  • Rank
    Sergeant Major

Profile Information

  • Gender
    Not Telling
  1. Tanks - Damage improvements

    Which is why I said explode, not burn out. For example, an ammunition explosion (such as a HE shell detonating when struck by a AP fragment or a HEAT jet) in a T-type tank is sufficiently strong to rip out one side of the hull and throw the turret dozens of meters aside. It should only happen if a shell explodes, not when the propellant burns, but being in the way of a blast like that will do hell to the equipment on the outside of your tank if it happens right next to you. A slight distance should already greatly diminish the damage. However, I just saw a chain-reaction with a couple of nyxes standing next to each other. One got hit, blew up, and then one popped after the other. This shouldn't really happen. The vehicle fire simulation in Arma is flawed, though, it is very simplicistic. Maybe screwing down the damage is a good fix.
  2. Tanks - Damage improvements

    An explosion sufficiently large should also damage the tank, not just the crew, especially vulnerable auxilary parts. Blow off slat armor and ERA fixtures, destroy gun turrets and tracks. Is there any real world data suggesting how large an explosion needs to be before the crew inside feel it? A tank blowing up is a pretty huge explosion, typically.
  3. Bad Weapons Dispersion

    Thanks for the quick answer Kju, much appreciated!
  4. I have a question. I found that AI unit weapons dispersion is way off the charts for me, and when I came back from my hiatus to check out tanks, the problem was exacerbated as it seems. I have made a thread here: Is anybody else affected by this? It only hits AI units for me. All mods are turned off. Player weapons accuracy is completely fine, which is making me think this is a game issue.
  5. Anybody else get this? Only AI vehicles are affected. This is a tank shooting at a 3km distant target. Player weapons are perfectly on target. The dispersion near the target is around 10 meters in radius(!!!). At the distance of the tank (920m), the dispersion is almost 5 meters in radius, as an estimate. To get this number I used the 10x10 meter cubes and stacked them. Dispersion was large enough to miss a single floating cube reliably at 3 km distance. Arma's copyprotection for illegimate copies used to be that player weapons were affected by progressively worsening dispersion. My copy was bought way back in early access phase. For a couple of patches, only AI has been affected for me. I didn't play Arma a lot recently, and I used mods so I thought maybe something was wrong there. Now tanks are coming, and I want to get back and see how they work. I turned off -all- mods for this. What is going on? I am on Dev Branch.
  6. Tanks - Damage improvements

    One major problem with A3 vehicle ammunition is ricochets. I have tested the Kuma, Slammer and T-100 extensively, firing from distances of 500 - 2000 meters, and found very odd behaviour when encountering angled surfaces. One example here. I found that the APFSDS behaved much like a rubber ball when hitting angled surfaces, for example I had this happening: https://imgur.com/a/LHSX4 AP from slammer at about 1km, ricochetting off the armor, without doing damage twice. This shouldn't happen! Modern APFSDS rounds are designed to be very resistant to ricochets, in fact they are designed to nose into armor they encounter. Apparently the angle at which they bounce is approaching 80° to normal. Anything less and the armor needs to be thick enough to stop the projectile from plowing through it. Tank rounds need to be -way- less bouncy. What often happens is that the projectile shatters, and the tracer goes flying off while the penetrator in the front of the round continues through the armor.
  7. I tried running this and found that CBA appears to break the addon. Running only this, and CBA it says that watcher.sqf is missing, and crashes the game when trying to enter a mission. Is this a known problem?
  8. Loved some things, really disliked others. Positive: Nice pacing in the missions, good challenge: no mission really felt unfair or broken. Good tour around the island, with nice scenery to be seen all the time. There's freeform gameplay with the scouting missions, and objective based story missions inbetween hubs. Overall design of the assets helps the immersion. With the new sound design, the new features and assets it is even better. Negative: Reason for CTRG being on Altis is pure schlock, like something out of a 1960s B movie or a pinball table theme. So many reasonable cyberpunk dangers to bring into play, and they picked what they picked? This broke the immersion for me quite a bit once I realized what was going on. ORBAT is useless: killing enemies in missions, or destroying or stealing equipment does nothing to this. Feels like it was meant to do something at one point and then left in when it was realized that it wasn't doable. Smells a bit like harvest red. This also goes for the scouting missions: the only thing this does is recover equipment, which is a nice bonus, but overall they feel pointless because they do nothing to the enemy. Generally I liked it more than I did harvest red, or the Armed Assault campaigns. Still not quite OFP, but I think that campaign wouldn't fly today either.
  9. Tanks DLC Feedback

    I read over the damage model research done by olds and the summaries found especially here: Arma 3 Damage Description, and I did two firing tests against vanilla vehicles using the vanilla damage model, one of which I already published here. The other one is in spoilers below, for those who are interested. What the tests showed pretty clearly is that the damage model is rather antiquated in its feel, and the way it is configured diminishes its strengths because of the legacy components as I see them. The biggest problem with that is the HP based component damage and the bad simulation of ammunition types against armour. This includes non-simulated HEAT jets, extremely ricochet-happy fin stabilized KE projectiles and on/off modelling of blast damage. Another problem is the lack of complex internal components, and very basic damage model for the components that do exist. The components of the vehicle that are made obvious in the damage display are tracks, hull, turret, gun. That's it. I keep repeating myself, but there is one game that does this (in a, pardon the pun, transparent fashion) rather well, and it's warthunder. The rounds are all individually modelled hitboxes, so are fuel tank, horizontal + vertical turret drives, and even components like optics and radios. The gun alone is made of three parts (vertical drive, breech, gunbarrel) all damageable individually. The amount of vehicles available in that game at release is comparable to what we have in A3, and old vehicles could progressively be updated. It pains me that my favourite game has less accurate vehicle damage models, than a mediocre PVP tank game. The problem with A3 is really not the outside of the vehicles, but the inside, and the inside only. In the Kuma test, the shots penetrating on the loaders side more often than not went straight through the turret bustle, where the ready rack is located. This has blowoff panels. The projectile going through the door would leave a hole, but most of the energy would be dissipated through the roof of the rack - instead, the crew bailed, and the tank completely blew up a couple of seconds later. Damaging a critical component seems to set off some sort of countdown timer, and then the the truck/tank/plane blows up. Arma also simulates no fragmentation at all, or spalling. So if the shot with its infinitely small size missed the gunner, it missed him completely. Crew are very tough to hurt, while their vehicle is extremely easy to hurt. There are large, empty volumnes inside the tank only occupied by people. Putting an inert round (ie KE) through the lower side of a Kuma should injure the crew, but not cause the vehicle to explode. If a fire is caused, the fire extinguishing system kicks in. So, in summary, improvements on the vehicles would be most welcome as follows: Make vehicle components individually damageable and increase amount of these systems. Also model in 3D space their shape and location. These should be for example: guns (single piece model is enough), gun drives and stab, optics, engine, fueltanks, drivetrain, tracks, all crewmembers. Armor layouts should also be constructed of 3D components instead of flat surfaces on the outside of the vehicle as they appear now. For example, frontally, large parts of the Kumas armour outboards of the turret and hull are inert, but in the current model they count as penetrating damage if something manages to get "under the skin". That also means different armour types according to the location and design in real world terms. This requires modelling ERA, NERA, Standoff panels, Fences and such in more detail. Inside damage model should include component fires, fire suppression systems, spalling and fragmentation, and possibly (not really necessary imo) overpressure. Ammuniton, especially HEAT, needs to be modelled in detail. APFSDS is extremely ricochet happy, and behaves more like a ball bouncing off surfaces, rather than an arrow. Shot traps don't matter with modern tanks, otherwise everybody would be really careful in avoiding them. Once internal components can be damaged and destroyed, the hull HP damage can be removed entirely or at least tweaked so it is the last resort to destroying a vehicle, rather than the first. I hope I'm not being greedy after the additions to the fire control systems along with the jets. I also made a rough layout sketch, like this: The optronics on the countermeasure launchers obviously should disable the automatic launching of CM, not the gun vision. The sizes of the components are -extremely- wonky. The CITV for example is a much larger device that sinks well into the turret. The main gunners sight is also much larger, but this is just an approximation and illustration, so I left it as it is. The size and location of the ammunition is more or less correct, though. I did not add the fuel tanks, most of which seem to be located above the tracks and in front of the engine (?). Hitting those would obviously cause fuel spillage and fires. They are also NERA. Swedish STRV-103 used fuel tanks to augment its armor. There is a video of firings against the S-tank here, which is enormously informative: click I also made a frontal sketch like this, but I think that's redundant, so I won't show it here. Unless there is interest? Below my firing test against the Slammer (using T-100.)
  10. Weapon tests

    I did some more armour penetration testing in light of the upcoming DLCs, particularily tanks. This is to document a bit better how the armour behaves ingame. Own Vehicle: Slammer M2A1 Target Vehicle: Kuma in frontal aspect. Distance: 2126 meters Gun and Ammo: Slammer 120mm APFSDS All Green markers indicate hits. Numbers are the order in which the shots were taken. Numbers are to the right of the corresponding impact. Not all shots were taken against the same target. Everytime I blew the Kuma up, I reset. Below I will indicate shots taken against the same target by a number in brackets after the number of the hit itself. All orientations are made from the position of the target vehicle (left means left from the Kuma's perspective. I'm doing this simply because that's the way I wrote it down and I don't want to make transcription errors) Hits: 1: Perforated Turret ring, ricocheted inside the turret and was stopped by the inner wall of the hull next to the gunners seat, at the height of the return rollers. Tank exploded after this hit. 2: Perforated the lower slope of the turret cheek, ricochetted on the hull top around the turret ring and was stopped by the gunners position. Tank exploded after this hit. 3: Ricochet off the gunners hatch. Main gun was completely disabled (red) after this hit. (Why?) 4: Perforated lower slope of turret cheek barely inside the non-inert parts of armor. Ricocheted off the hull while inside the turret, exited the turret rear and ricocheted again off the engine deck. Tank exploded after this hit (Ammo behind gunner?) 5: Ricochet off turret roof in front of the loaders hatch. Main gun disabled. (1) 6: Perforation of turret ring, clean through and out the back, stopped by engine deck. This shot blew up the tank. (2) 7: Ricochet off the shot trap's slope downwards, perforated the turret ring and was stopped inside the fighting compartment. Tank blew up. 8: Ricochet off the shot trap, perforated and overpentrated hull roof over the ammo compartment, exited through the floor of the hull. Tank blew up. 9: Perforation of turret roof well in front of the loaders hatch. Normalized lightly on the armor, went slightly downwards and exited the turret back and grazed the slat armour at the back of the tank. Tank intact. (1) 10: Clean shot against center of turret cheek. Shot was stopped inside the armour, slightly in front of where the gun mount would be. No damage to vehicle. (2) 11: Turret cheek perforated underneath gunners optics. Shot ricochetted downwards of commanders position and was stopped by the engine firewall. Vehicle destroyed (3) 12: Hit squarely on the right hand side of the gunners optic. Shot was stopped by something right behind the optics surface. Gun damaged. (1) 13: Perforated the turret just to the left and high of 12, shot ricocheted on entering the fighting compartment and exited the back of the turret. Turret clearly damaged after this shot. (2) 14: Perforated the gunners optics again, at 6'o clock from 13. Bounced through the fighting compartment and was stopped by the back wall of the turret. (3) 15: Gunners optics again, clean into the turret. Ricochetted through the commanders seat and was stopped short of the back wall of the turret. (4) Vehicle still intact! 16: Perforated the gunners optics 2:45 o´clock of 14, in through the fighting compartment and stopped by the comanders position. (5) 17: Impact in front of the optics on the cheek armor. Shot was stopped at the surface of the armor, no damage to vehicle. (6) 18: Impact 6´o clock of 17, shot ricochetted off the armor and over the optics, making no further contact with the vehicle. No damage. (7) 19: Square impact on the thick part of the turret cheek, just outboards of the optics. Shot was stopped by the armor. No damage. (8) 20: Ricochet off the shot trap below the gunners optics. Ricocheted again off the drivers hatch and was stopped by underside of the turret slightly in front of turret ring. Hull damaged, Vehicle exploded. (9) Lessons learned: This was a surprising test. At the distance, I had trouble with the first set of shots to not hit any parts that would immediately destroy the tank in one shot. Kuma stores ammunition on the loaders side in the turret bustle, which explains why the shots that penetrated and perforated that side of the turret resultet in one shot kills. I also ricochetted twice off the turret roof, and in both cases the main gun was unuseable after that. No clue why, the shot wasn't even close to the gun. Maybe hitting the gunners seat counts as hitting the weapon? On the commanders side of the turret, the tank is surprisingly resilient. It obviously would have been unuseable as a fighting vehicle because the main gun was down immediately, but overall, hitting the turret front on the commanders side without impacting the hull allowed it to survive 8 hits, 3 of which did not penetrate at all. The 9th probably killed it because it made contact with the hull (bouncing off the drivers hatch) and going into the bottom of the turret. I suspect it was not hitpoint depletion because then it would've burst into flames immediately. The distance was 2126 meters. At that distance, only the small, vertical parts of the turret cheeks actually were effective in stopping incoming rounds from the Slammer. I would imagine that the bigger gun on the T-100 could probably do better than the Slammer. Hitting the shot trap at the lower parts of the cheeks results in a one shot kill most of the time. Ammunition is ricochet happy. I am not sure that this is realistic, because modern APFSDS is specifically designed to not bounce. There's 60 years of engineering behind these, if anything they should nose into the target unless the angle is so acute that the shot would just plow through armor all the way anyways, and not perforate the armour. That's another reason why sloped armor isn't a thing anymore, unless it's ERA. Angled armour sandwhiches with air gaps are used because of this tendency of the rounds to nose into the target, with alternating layers causing the projectile to fragment as it passes from layer to layer. 3 was okay, but 5 would probably have plowed into the armour and gotten stopped by it as it tried to nose into it from its very acute angle. Shot 8 was interesting because the round ricochetted off the turret cheek, went through the roof of the hull and had enough force left to go back out through the bottom of the tank. I hope this is at least somewhat interesting. I'll see if I can do more of these, maybe with more detail. I'll also try different angles and shooting at different parts. Edit: I reviewed this and now the damage to the gun on ricochets explains itself. All there in the pudding. This is really something that needs to go: if your gun can damage the vehicle, it will always do damage no matter if it actually penetrates the armor. Killing a tank by ricochets only is completely feasible because they often disable weapons. Disabling the tracks and/or engine also makes the crew (if AI) bail, so killing tanks is -way- easier than it should be.
  11. Hey, I have a question. Being completely inept at scripting, I've not even the faintest how to do this, but what I am basically looking for is a working shot timer. Ie a script/device that upon start, within 5 seconds gives a random start time, and then logs each time it registers the players gun firing. This is how shot timers work in real life, and its quite different from how the timer on bi's time trials works. If anybody has any Idea how to do this so a potato like me can do it, I'd be really happy. thanks in advance for your help, Cheers, Insta
  12. Tanks - Fire-control system

    Holy dingle. This looks awesome The AI is extremely good at spotting things inside their arc of vision, and blind outside it. They have to get shot at to spot something outside of their arc of vision. They also don't scan. There's the "scan horizon" command, but that's not the same. For example, in Wargame:Air Land Battle you can see your little tanks scanning their assigned arcs of vision, turrets rotating back and forth. This addon with the lasers is great for players, but the vision and lack of scanning problem for the AI isn't really alleviated by these advanced gun laying systems. If the AI spots you, it is super deadly, and if it doesn't spot you, its cannonfodder.
  13. Was the tank AI always this crazy?

    Saying that they are completely terrible is too much, I think, they are just not good. As for the racing around, they are doing the same thing as the infantry, getting sent ahead by their commanders piecemeal. For the players platoon, usually mine get stuck on terrain, flip over from the bad physX or forget how to brake. The "always move fast" rule for combat mode is also terrible. I won't expect AI to ever learn to do tank combat correctly. Current AI tech barely manages infantry being decent. Arma is the only game right now to do this (on the scale it does) anyway, so we should count our blessings. Most other games would rely on making the AI even more basic, and then giving them artificial boosts or make them appear in massive numbers to make them compete. Although maybe BI finds a way to make them cheat convincingly, sort of like the AI in FEAR. That would require a complete rework, it seems... which isn't a solution after 16 years of work going into the ofp/arma AI. The only other game I know that runs on an approximately similar level of complexity is SABOW, but that one is designed around tanks specifically, and tanks only. So, the tanks are super, but everything else isn't accessible to the player and is running much more simple animation sets and behaviours compared to Arma. The one thing SABOW AI Infantry does way better than Arma is that they pick up dropped AT weapons, though.
  14. Has anybody experienced this issue: https://feedback.bistudio.com/T122467 Tanks not switching to main gun on enemy vehicle targets unless they are directly looking at it. I tested it with multiple vehicles, with mods and without, and all tanks experienced this issue. Commander gets stuck in a "target that vehicle" loop while gunner has MG selected, then the commander orders his tank to run away. Only if the driver points the hull towards the target does he switch to main gun. Would be nice to hear what other people find.