[NUC] elusive1: Unleashed

Hahah people like you really don’t know what an aimbot is do you?

It's a hack which either aims, shoots, or otherwise makes sure that bullets hit a target. What are the odds for the hit at the 8 min mark without one? Not speaking of two such shots.

Answer: the screen resolution is on the order 1000x1000 = 1 million pixels, the target is a couple at most, so its perhaps 1 in 100.000

If you pay attention to his movement at 10:42-10:50, you will see his gun snapping around as he walks. This part of the video is very suspicious, to say the least.

You mean where he hits regardless of the concussion and flash grenade?

Ahh ok I hadn't seen the 8 minute mark one - have to admit - it made me laugh ! (In a sad way though; the target hasn't played since July so I guess he just got fed up).

Typically I'm not one to accuse people of anything, but this made me raise an eyebrow. Mad skill, Crazy luck... or...

He was simply shooting in the general direction of that doorway.

Those are very lucky shots but definitely not impossible if you take such chances all the time. And your odds calculation is completely wrong.
Typical resolution is 1920x1080=2 073 600, he's not simply randomly shooting at any part of the screen so that's not where the bullet can go so it doesn't matter anyways. To get the chance of hitting you would need accuracy stats of his weapon and distance to the target.

1&2: Collision detection uses the current position of the target. When lag happens and people go off floating, only the hits on the model are scored as hits. You can also check this with a jumppad: The game has trouble keeping enemy positions up to date and they will rubberband as they fly, best visible when you step on the pad alongside them. If you fire while flying, only direct hits on the model as it's rubberbanding will score a hit.

Then theres this:
at 2 seconds into the video, he dragshoots a guy, you can see the muzzleflash, recoil and the bulletcounter goes down at the same time as he passes the head, it's a headshot as expected. At 6 seconds into the video you see the same muzzleflash, recoil and ammo counter drop while he's off to the side of the player (protodip), yet it's still a headshot. At 1:50 if you use frame-by-frame you can see that he actually hits the shoulder (even accouting for movement) yet it still scores a headshot (I tested similar shots at various angles and got no headshots). Then at At 2:09 he shoots a Heavy as he's aiming for the belly, but you can see the muzzleflash and ammo counter drop again a frame before he starts dragging upwards, so it shouldn't have been a headshot, yet it is.

The video is absolutely filled with shots that are at odds. Considering that he even scores headshots where you can see the indicators of a shot after he passed the head (like with the shoulder shot) there is something wrong there.

How to explain it? Lag or graphical wizardry? Then it would be reproduceable and then it would still be consistent. If a shot straight for the head is a hit then a similar dragshot that is fired before he passes the head should be a miss, and a shot that happens after these indicators should definitely be a miss.
So a dragshot somehow "smears" a bullet across your screen for a larger to-hit area? The client renders bullets and has them travel so that can't be it, and it wouldn't explain hits with shots after you passed the head.
Framerate? We know that framerate has an effect on the ROF of your weapon, so it might influence single shots. But that would mean that this guy managed to learn how to "save" a shot by dragshooting and causing frame drops due to the sudden speed of the dragshot and getting the right frame with the bullet at the exact time he passes over the head. Which is more difficult than just timing your shot and still doesn't explain hits with shots after he passes a head.

There's only two explanations that I can think off that work:
1: His client somehow works differently than that of other players, hence it is impossible for me for example to reproduce as when I fire similar shots and look at them frame-by-frame then a miss remains a miss and a hit remains a hit.
2: He deliberately changed files or uses cheating software.

Occams razor would suggest number 2, as this guy has been doing this for a long time and through numerous updates. Also considering in this thread for example; https://forums.daybreakgames.com/ps...qc-sniper-montage-60-fps.208746/#post-3032846
You can see him do far more dubious things, like shots that are pretty far off-target that still register as headshots. This was ofcourse before it was common knowledge that you could change a file and edit the size of the head hitbox, and the day that they created a fix for that abuse was around the same time that this guy stopped posting video's. Convenient. So we know he has abused things in the past before and I see no reason why he wouldn't again if he thinks he can get away with it.

This guy himself tried to cover it up back then with things like "but you are missing frames!", but that still doesn't explain most of the kills I point out to him. And I'm tired of having to tiptoe around a bunch of worshippers that think that anyone who looks good has to be clean.

Concs aren't that super scary when you are already looking where you need to shoot. You can clearly see him strafing to align the shot due to the effects of mouse speed of the conc and once the flash hit he ended up in a situation he might hope fore the best and take the shot. Also, if you were doing a montage and let's say you pull off these lucky shots 1/10 times, would you use the clip where you made that shot or the 9 other clips where you missed and died?

He's just doing the tried and true tactic of rapidly zig zagging to make it harder to hit because client side detection. If you've ever wondered why someone seem to run zig zag really quickly that's what's up. A fairly common escape tactic in PS2, especially with vets (not just from PS2). As you can see in the vid he is trying to get away to avoid a death and chug some medkits.

The times you see smoother movements tend to be on PS4 due to being played on a controller as you can't really do the rapid movements as you can with mouse+keyboard.

Just so you know, 5 or 6 players I know have aim nearly as good as elusive! I’ve seen clips of them but they are unlisted since the guys don’t want people to see them.(can sense your reply”COS THEY HACKING MEMEMMEMEMEME). Sure I admit shooting through concs and flashes is kinda suspicious and TBH I thought he was hacking st that point, then I saw it says HIGHLIGHTS and I thought “ohhhhh .”

Personally I don’t think he is hacking but showing some really good /sometimes lucky clips. I think someone should do this frame by frame thing on his stream and see if he consistently hits these wacky shots.

Anecdotal circumstantial hearsay. Now that's some strong evidence! And to think there was ever any doubt.

If you look you see I'm not questioning his skill. In my teens I did similar things as in thes video's, although in other games. But the difference is that I actually hit the opponents.

I question the validity of the actual shots you see in the video. They are at odds with how "normal" gameplay would happen. How can a shot that was off-target become a headshot, if a shot that is on-target remains a headshot? How can a shot that happens after it passed the head still be a headshot? Why would you ignore those grievances and come with a story of "but I know people of skill so everyone of similar skill HAS to be clean"?

It could be lag/issues with the recording software, he does also claim to livestream/use the inbuilt one. Perhaps the HUD/weapon animations are out of sync with his movement and aim. I still think it's cheating though, yet a possible explanation for the video anomalies.

All these other Rubes never looked into it... but those of us who used to Program stuff like that aren't gonna be fooled. Only total noobs just leave that Process loop running the entire time they're in a fight. Everyone else who still hasn't been banned yet knows enough to look for Toggle features and "Hold Key to XX" whether that XX is either subtle Angle-clamped Aim assist (to prevent blatant "Snapping"), or just a Trigger Bot that's frankly a lot more fun and engaging than just letting a Script do EVERYTHING for you. Infact I used to run one WHILE I was an official Admin (in some source games) banning other more blatant Cheaters and only 1 other person on any of the servers over a span of MONTHS ever questioned me about it

IOW: If this staff was actually experienced enough to figure out what's really going on, all those subtle people woulda been gone years ago back when Smedley was on the warpath, but they aren't so whatya worried about?

Looks like he ''took aim'' as usual if you watch in slowmo. No one knows how the ballistics / hit detection thing works for sure, not even iridar. But using pixel values is a gracious measure, the real odds are even less, and then its about 1 in 100k, definitively more than 1 in 10k all things considered. Which means he would have done the same routine (probably) 1000s of times, and would have way worse accuracy stats than his character stat-page says.

If that were true, shots in each segment of his video should have the same results. Yet we see kills seconds apart with one on-target and the other off-target still get the same result. Also a problem with the recording software wouldnt render a muzzleflash for example later than elusive would have shot, since the recording software records the litteral frame that was send to his monitor and not the events that caused it.

That's why I mentioned that it could be the inbuilt recorder, which might not record literal frames but split the HUD and other parts into different channels or something. Furthermore it could also be out of sync on his screen as well. There are many glitches with the weapon animations

That would be a grossly inefficient system, and it wouldn't explain why the muzzle animation, recoil animation, bullet animation and ammo counter animation all line up perfectly with eachother but not the model. Or that one second they do and the next second they don't even though we don't see any framerate drops? It would be a highly specific bug in the recording software.

Also PS2 is already very CPU intensive, do you really think they would almost double that work by recording each instance seperately and recombining it twice, once for the monitor and once for the recording?

And this still doesn't explain hits where the target never ever even crossed the bullets path because OP was aiming multiple frames off-target before firing but still gets headshots.

I don't know, it was just a plausible-sounding explanation for the dragshots and not the ones where the bullet goes where he has not aimed. I do think however, that the HUD/weapon animations are processed separately from the world model. Reason being that the former only depends on your actions while the latter is modified depending on what other players do.

The inbuilt recorder system could have some hidden structures and features beneficial for the devs or moderators, such as recording the raw view without interference from the HUD or weapon animations. In principle it can be less CPU intensive to output such channels directly to file and recombine afterwards instead of doing conversion and compression on the fly. But again that's just a possibility.

Also, what I think is the real explanation is that the hack IMHO he uses processes aim assist snapping differently from the modified projectile trajectories.

Ahh true I concede, I dunno maybe he’s just showing the BEST parts of his gameplay?