Saw an article on Kotaku about a new kind of aimbot, which they did not link to nor would I link to it even if they did. This one apparently works on PC or console, and is utterly undetectable on any game (they claim), though the video I saw was running Warzone.
Did anyone else see the article? And is it even possible to code a hack that’s truly undetectable? I would imagine that sooner or later everything is detectable, but since I’ve never been interested in being a dirty cheater, I don’t know if being “undetectable” is how all aimbots are advertised, or if that’s something that’s even possible.
God knows how long this cheat has already been out there for PC players, how many matches it has already ruined.
The way it works is actually pretty interesting–but I’m not quite seeing how it would differentiate friendly from enemy in some situations where player skins are very similar. I suspect there will be varying degrees of success if each user has to train their own model.
I’m also curious about any issues/viability with latency. Streaming the game to a secondary machine (presumably with a GPU requirement?), running the classifier, and then sending input back to the first machine sounds like it might slow.
Most detection here might have to rely on detecting non-user mouse and key input.
It’s a hardware device that you put between a KBM or controller and the console or PC. It’s undetectable because there is no software running, no driver required. It just intercepts the input, reads the screen, and uses AI to respond with an aimbot and a reactive script that auto-activates abilities when needed.
It isn’t new. It’s been out for a long time, but there’s a new version of it in preorder right now, and it’s easier to use and has less lag.
These hacks have been around a long time. This is news right now because the company selling this aimbot looks legit, not sketchy like usual aimbot sites. And it isn’t selling it strictly as an aimbot, it’s just “scriptable hardware” or some other marketingspeak. You can script it to auto-build in Fortnite, or automatically crouch after you shoot in Overwatch… or you can make it read the screen and aim for you. For that matter autohotkey isn’t an aimbot either, but it’s cheating when you use it.
How many matches has it ruined? Every match I’ve played this week in comp, and about half the matches in QP and Arcade. People cheat everywhere for all sorts of reasons, not just to climb the comp ladder. I just went 6-6-2 in comp today, not even one match was clean.
This is something I’d like to drill down on, with people who know more about it than I do, because I only know about hacks what I read on these forums, which falls into two camps: 1) There is no cheating, or such a fractionally small percentage of active players that it’s not anything to worry about, or 2) Cheating is rampant, it’s everywhere, in almost every match.
I would imagine that reality is somewhere in between these two positions.
But if you had to guess (not just the person I’m replying to, but everyone), just based on your own anecdotal evidence, about what percentage of active competitive OW players would you guess are using this type of hack in the game without being detected? And at what percentage does it become totally unrecoverable? Are we already past that?
I suggest that it’s got to be thousands of people, because Blizzard bans tens of thousands of accounts at a time. They’re not unique, so it’s a far lower number than the total number of accounts banned, that’s still thousands of unique people. And when each unique player gets banned, they don’t just go away. They start a new account and resume, right?
And last, Overwatch 2. Matter of time before cheats like this are available after OW2 releases. Do you think OW2 will have a few months where the number of cheaters are minimal (until hacks are made specifically for OW2), or do you think those hacks will be ready and in use by a significant portion of OW2’s starting player base on launch day?
Because if the side that believes these hacks are prevalent in the majority of comp matches are right, then Overwatch is cooked, as far as I’m concerned. If a player is likely to find the majority of their matches also have another player with an undetectable hack (really, any more than about 5% of matches would be unacceptable), then this game has literally become a cheat-fest, and the only way to compete would be to join them. If they’re right, a cheat/hack-free Overwatch as it was intended to be played will never actually exist again.
I can’t say from my experience because I have seen other people play several matches in a row with no cheating. The moment I’m grouped with them, they’re blown away by the constant 3-stacks of low-level accounts with throwaway names. If you play really well, your MMR goes up, and you get matched against people of the same SR range who also have a high MMR. That’s the Cheat Zone. If I soft throw, I’ll lower my MMR (and SR, unfortunately), and get into normal matches.
So there are some people who genuinely don’t see cheaters often because they aren’t above average skill.
Some people claim that cheating doesn’t exist because they are shilling for cheat resellers. Obviously they all want to claim that their hacks are “undetected,” and disinformation on public forums is the perfect tool for spreading that lie. Even if Warden doesn’t detect a cheat, a manual review will, and that is triggered by a certain number of player reports.
Some people see cheating often enough, but they have strange ideas about what cheating actually is, so they think the problem is “smurfs.” There are no smurfs, and if there are, by the time they’re at level 25 they’re already in the high-MMR Cheat Zone. Secondly, legitimately good players don’t have magic aim. Go look at OWL player stats. They can’t cheat during matches, so that’s the only known-good data to compare to. You’re not going to see an amazingly high accuracy or crit rate above average OW competitive data ala Overbuff. Some like to claim that it’s harder to have good accuracy at higher levels because players know how to move and position, but there is no data to support these claims. Regardless of context, it’s moving your mouse to an animated head on the screen accurately and quickly, and you adjust for movement patterns. A true high-level player doesn’t stand out in the open snapping out perfect no-look headshots. They tend to use cover and don’t yolo apart from the team repeatedly. Cheaters are utter, utter noobs – truly bad players who often can’t win even when they cheat.
If you want a ballpark figure, take a look at other games in this space that report all cheat ban numbers. Last I checked (I actually made a post about this a long time ago) that was anywhere between 12% to 20% of the playerbase of any modern game with active anti-cheat measures. At this point I’m sure Overwatch is higher because Blizzard just isn’t good at heuristic detection, the cheats are much more advanced now, and every time there’s a sale the cheaters stock up on accounts. Keep in mind, though: this does not mean that 20% of the players that any given person will see in matches are cheating. For some it will be very high, for others it will be very low.
Cheating is also heavily region-dependent (less cheating in EU) and time dependent (late at night and during normal working hours are cheat-heavy).
Lastly, cheaters have many accounts, and you can’t really tell when they switch. I know I’ve seen 2-stacks of identical players with several sets of accounts, and there are certain silly behaviors, skins, icons, grudges, and patterns of bad play that I see often enough that I’m sure the number of actual cheaters is lower than what it seems – it’s a certain number of dedicated cheaters (very likely boosting accounts) with a ton of throwaway accounts that are in the queue for a large portion of the day.
It all depends on how Blizzard handles player accounts, and whether Windows 11 with TPM enabled will be a requirement. If OW2 ushers in a new era of bank-grade KYC account verification, then there will be no safety in having many accounts – or at least it will be much more difficult and expensive to have as many accounts as cheatos currently have. The TPM / Win11 requirement would allow Overwatch 2 to be protected, at the lowest level of your PC hardware, from tampering by cheat programs. Not sure what this will do for the hardware hacks, but I’m sure there’s something. Lastly, Blizzard could get smarter about data-oriented cheat detection. If I can look at a killcam and ID a cheater, then someone smarter than me can create some machine learning rules for automatic account flagging. An AI system could be fed thousands of VODs of known cheaters, and develop a model that would accurately detect almost all cheating – and it would continue to evolve over time. Eventually this is what will defeat cheating for good and all, but it’s expensive and requires Google-level talent to implement. Blizzard wastes all its money on its CEO’s salary, and all the talent has fled the scene.
If none of that happens, then there will be OW2 hacks from beta onward, and it might get so horrible that the game never takes off. Video game ranks are a major source of identity for an unwanted generation of young middle-class men. They will pay a lot of money to uphold their false public Gamer image via boosting and/or cheating.
Years ago, I read an article of an interview with a former game hacker. I think he was mostly a hacker (or scripter) in call of duty. In that article, he said that in every game lobby, there was about a 70% chance that someone was using some sort of hack or cheat.
It is still sending inputs back to the user’s M&B. If the input device/software [I’m thinking something similar to AutoIt] cannot itself be detected, it can still be detected through pattern recognition via regression models. Even if the input pattern classifier is relatively weak, covariates might include: player reporting, in-game statistics looking at percentile deviation, trust level (similar to CSGO), previous violations with hardware ID/IP (although it can be spoofed), reaction time averages significantly below human ability etc.
Although it’s not a surefire solution.
Blizzard indicated in the May banwave–which caught 10,000 hackers–that they also ban around 1,000 accounts per week for hacking.
The season 24 banwave resulted in 65 accounts from Europe Top 500 getting banned, 42 from NA, and 75 from Asia. 182 accounts in Top 500 across 3 regions (1500 account pool) means at least 12% (182/1500) of Overwatch accounts in Top 500 were cheating. That’s also only looking at DPS roles, but there may be overlap with other roles; and that’s only from the cheats that were detected…
[Tinfoil hat]
-If 1% of Overwatch accounts are using hacks and you do not hack, you will encounter a hacker in ~11% of your games. When there is a hacker in the game, 54% of the time they will be on the enemy team.
[More tinfoil hat]
I started playing OW in 2017 and took a break in the fall of 2019 after climbing to within 100 SR of my Open Queue rank [support main]. Came back in the spring of 2020 and felt that games were significantly more difficult than I remembered. I initially chalked it up to hero changes, new metas, and player experience that I was struggling to adapt to. But, it just seemed like people–especially DPS–were just better than I remembered. In the past few months I’ve been wondering if my intuition about better players is due to an increased number of cheaters.
Yeah, heuristic anti-cheat is the eventual (and final) solution, but it requires more AI than Blizzard is capable of right now. When I say it’s difficult to detect a hardware hack, I mean by current process ID / signature standards – what Blizzard currently does.
It should still be possible to detect a hardware cheat, but I’m not sure that could be done at the level of permissions that a game typically has. It’s possible that Windows could detect it, and driver developers could potentially do something to query the keyboard and mouse to see if there’s unusual latency or a bad identifier built into the hardware. Console controllers each have a unique hardware ID, so the console knows that you’re using legitimate input devices and which users are using them. The reason why the “new” hardware hack is so dangerous is that it’s actually a rebuilt controller, so when the console detects it, it says “yep, that’s a real controller, no problem here.” I’ve also seen a proof-of-concept hardware hack that was built into a PC keyboard.
Hardware hacks cannot escape latency issues, so if there’s some way for Windows or a console to sort of “ping” the controller or mouse, then intermediate hardware hacks should interfere with that to the point that they’re detectable. The latency would be well out of bounds.
Let’s keep in mind that most cheaters are not cheating on their main accounts, and they often have many extra accounts to cheat with. I’ve seen cheaters whine about as many as 20 of their accounts being caught in a single ban wave. If 10k accounts are banned for cheating, the actual number of cheaters is some amount less than that. I’ll take a wild guess and say that represents 2500 people, and I’ll take an even wilder guess and say that of those 2500, 2000 are professional account boosters. For them getting banned is expected, it’s part of their overhead. They will not stop, they will not go away, they will just use another one of the hundreds of extra accounts they bought during a sale.
Even with ring0 permissions? Although, that still has limitations and can be bypassed.
That sounds pretty neat–what kind of hack are we talking about? A pixel bot with macros or something?
Yeah, I suppose in this case, with the need for a secondary PC to stream, convert, classify, and send input back to the gaming PC–I was thinking there would be predictable minimum latency responses.
Yeah, that’s why I tried to say “accounts” rather than “players.” I wish Blizzard would be willing to release more data on the banned accounts like the average # per user that could be identified.
Something like that. It wasn’t as fancy as I made it sound. You know how a lot of keyboards have extra USB ports? Well, he just hardwired a USB thumbdrive to the port internally. Same as plugging one in, except you couldn’t see it. It was just a clever way of plugging in a thumb drive with (probably a pixel aimbot) software hacks on it. Still 100% detectable by current anti-cheat. If I remember correctly, the CS:GO professional gamer who got caught cheating in a tournament a couple of years ago was only caught because the officials saw him plug in his thumb drive.
If we’re thinking of the same hardware hack, the new one that everyone’s worried about right now doesn’t require a PC. The older model does.
I think Blizzard should give us all aimbots and wall hacks for one season, just to even the playing field.
Interesting, I’m not sure how one could remove the secondary PC and remain undetectable via anti-cheat software. The hack that this thread is about–detailed in the Kotaku article and shown on GamerDoc’s Anti-Cheat Police Department Twitter–requires a secondary PC.
Edit:
Maybe quantum computing will offer surefire anti-cheat in the future.
They’re not, though. Watch the replays. They spend a minute or two shooting the ground or the walls, obviously intentionally missing shots – to lower their overall accuracy stats so they can avoid heuristic cheat detection. Then Hanzo or Widow will stand out in the open, away from the team, snapping no-look headshots every time an enemy player is visible. Closer to the fight, you’ll see a McCree snapping to headshots when two shields are between him and his targets. These are not good players. There’s no teamwork, no use of cover, no coordination, no sense of strategy at all, and they only ever switch when they die to a hard counter. Good teamwork still defeats them. Unfortunately that’s hard to come by in metal ranks, or in QP or Arcade. (Yeah, people cheat in Arcade).
Watching replays of really good players is truly a pleasure. It’s offensive to watch these aimbotting clowns.
Yeah, I’ve definitely seen ground shooting to skew accuracy and manipulate heuristic detection.
Movement, cooldown management, and decision making also tell a lot.
I often look at the replays and think, “How the hell did my team just get curb stomped by this person? They play like a confused and lost puppy but have astounding aim.”
