[sneaK]

As many of us are aware I'm sure, cheating (at least in CS:GO) has become a much bigger issue as time has gone along. It's become incredibly easy to buy a new account on the cheap, inject and go. Some of these accounts have hundreds of hours and have been cheating for months, yet no VAC or overwatch ban.

That being said, we have few options at the moment to combat cheaters in our community servers;

• SMAC - Quite effective against basic/poorly coded cheats but outdated, ineffective against most modern cheats (in fact, many have anti-smac features), not to mention the whole DMCA issue disallowing its distribution. I personally have made small modifications to it to make it more effective, yet there are still numerous cheats that bypass SMAC easily.
• VAC - Kind of a joke as of late, if players are detected, generally bans are dished out after the damage has been done.
• Manual review - Active admins are helpful, however is not possible 100% of the time.

I run a small community, with fairly popular servers. Daily we get 3-4 SMAC bans on average (with a slightly modified SMAC, mainly hard value tweaks), and generally 1-2 duplicate accounts of previously banned cheaters (thx SourceSleuth, but that even isn't very effective vs users with dynamic IP's or VPN's). The sad part is that I only run 2 servers that are even remotely competitive, the majority are movement-based servers (surf, bhop, etc).

Obviously there is only so much you can do with server-sided anticheats, but there definitely can be improvements made. Outside of what SMAC already does, there needs to be:

• Strafe hack detection
• Spread variation detection (for consistent/perfect sprays)
• Aim assist detection (vectoring, etc)
• Silent aim detection (with no hard set minimum angles)
• Trace detection (for players who lock/snap onto other players through or not through walls, without actually triggering player death or even hurt)
• Improved bhop cheat detection
• Anti-airstuck (it's still possible in csgo, even tho valve "fixed" it like 6-8 months ago)
• Blocking spectator viewing information (it's possible for cheaters to see who is spectating them so they can toggle off when getting spectated by an admin)
• ??? More

That being said, is there any interest in creating a collaboration to come up with a newer, more effective anticheat for SM?

[Franc1sco]

Originally Posted by blackhawk74 generally 1-2 duplicate accounts of previously banned cheaters (thx SourceSleuth, but that even isn't very effective vs users with dynamic IP's or VPN's).

https://forums.alliedmods.net/showth...26#post2485726

[sneaK]

Originally Posted by Franc1sco https://forums.alliedmods.net/showth...26#post2485726

I saw that actually, would definitely be an invaluable addition to sourcesleuth!

[Fragkiller]

Strafe Hacking is more than interesting to block.. I mean there are realy shit versions out there, which can be easily busted like from bTimes-bash or bTimes-cash ("anti-cheats" for blackys timer).

But there are also stuff like "Silent Strafers" which are pretty much not "sending" the keypress, so its kinda impossible to controll that.

Other than that, anti-airstuck is uhm again, you cant just ban people who randomly stuck in air maybe cause of a lag or something else. And I guess you must meet a realy dumb cheater if he uses it more than 5x during few minutes..

[OSWO]

Originally Posted by Fragkiller Strafe Hacking is more than interesting to block.. I mean there are realy shit versions out there, which can be easily busted like from bTimes-bash or bTimes-cash ("anti-cheats" for blackys timer). But there are also stuff like "Silent Strafers" which are pretty much not "sending" the keypress, so its kinda impossible to controll that. Other than that, anti-airstuck is uhm again, you cant just ban people who randomly stuck in air maybe cause of a lag or something else. And I guess you must meet a realy dumb cheater if he uses it more than 5x during few minutes..

PHP Code:

public Action OnPlayerRunCmd(int I_Client, int &I_Buttons, int &I_Impulse, float F_Vel[3], float F_Angles[3], int &I_Weapon, int &I_Subtype, int &I_CmdNum, int &I_TickCout, int &I_Seed, int I_Mouse[2]) {

    if (((I_Buttons & IN_FORWARD) || (I_Buttons & IN_BACK)) && (F_Vel[0] == 0.0)) {
        B_ButtonError = true;
    } else if (((I_Buttons & IN_MOVELEFT) || (I_Buttons & IN_MOVERIGHT)) && (F_Vel[1] == 0.0)) {
        B_ButtonError = true;
    }

    if (B_ButtonError) {
        for (int i = 0; i < 2; i++) {
            F_Vel[i] = 0.0;
        }
    }
    return Plugin_Changed;
} 


That one for sure cannot be patched. Sceptical on releasing other AC's since they're easily vulnerable to being patched.

Wallhack can be easily fixed with SDKHook_SetTransmit, but will require a fair amount of computation to traceray clients.

"AirStuck"

PHP Code:

((GetEntityFlags(I_Client) & FL_ONGROUND) > 0) && GetEntiityMoveType(I_Client) != MOVETYPE_NOCLIP)

if their last pos is the same as their current pos for 10 frames or w/e 


[sneaK]

Originally Posted by Fragkiller Other than that, anti-airstuck is uhm again, you cant just ban people who randomly stuck in air maybe cause of a lag or something else. And I guess you must meet a realy dumb cheater if he uses it more than 5x during few minutes..

You're thinking a little too narrowly. As OSWO has posted right below you, instead of blanket banning possible airstuck users, just block the ability to do it. Logging is also a very helpful tool, with after-the-fact bans.

Originally Posted by OSWO Sceptical on releasing other AC's since they're easily vulnerable to being patched.

Originally Posted by oaaron99 The problem with trying to make an anticheat on a site like am is that you need to release the code to the public. At this point any half decent cheat coder can figure out ways to bypass the detection methods.

That's the trouble with anti-cheats, it's always a game of cat-and-mouse with cheat developers. If you were to make an anticheat publicly accessible, with closed source, it wouldn't really matter. Sourcemod plugins are easily decompiled anyways. This would be in violation of the SM license as well, obviously. So it comes down to making the anticheat public or not. Even if you shared the anticheat with "trusted communities" the amount of hands it passes through makes it easily leakable.

Originally Posted by MrDanne It would be a dream having a cheat free CS GO, but I think it is not very likely until Valve do anything about it. Looks pretty clear that they do not want to stop cheats, it is a good business for them when people must buy new game several times. I think we need some kind of client side application to have a chance against this. Punkbuster was a good idea but of course you could find a way around that one too. But regarding SMAC, it is sad kigen and the SMAC developers couldn't come to an agreement so the development could officially get going. Guess that won't happen. blackhawk74, are you willing to share your smac version maybe? Could PM me or so

I can confidently say we'll never have a cheat free CS:GO, but I think my original goal would be catching as many cheaters, as effectively as possible.

A client-side anticheat isn't currently something I'm thinking about, I'm thinking practically for community servers, such as SMAC. Unfortunately there's only so much a server-side anticheat can do, obviously.

It truly is a shame. As I've state above, the tweaks I've done really aren't too significant - minor hard value tweaks, for example aimbot minimum deviation, number of autotrigger detections, added default cvars to check for, etc. If you open up the source you'll see what hard values to tinker with.

Originally Posted by Benoist3012 Scrapping smac, and starting from something new would be better, I've been working on my own anti-cheat, and I won't share it to public as well, because public source anti-cheats kills the whole point of it.

I spoke with GoD-Tony regarding this, and he agreed with your sentiment. Essentially an entire rewrite would be needed to create a new, effective and up-to-date anticheat.

This is the trouble though, now each community is forced to develop their own anticheat, when a much faster, and effective remedy to cheaters would be several communities & devs working together. The side effect would be more access to the code, and cheats getting around it, however any sort of newly discovered bypass could be discussed and squashed quickly.

[headline]

Originally Posted by blackhawk74 -snip-

A public anti-cheat which is moderately effective is still a good baseline for plenty of server operators to rely on. No one creates an anticheat without the understanding that people will get around it. As new methods of cheating arise, having a baseline platform in which everyone can contribute to and expand on will be the most effective. Those who want their private anticheats will still create their own if needed.

[Pelipoika]

Most cheats don't populate the mouse x and y params when snapping to targets so a validity check of those would detect aimbots

[MrDanne]

It would be a dream having a cheat free CS GO, but I think it is not very likely until Valve do anything about it.
Looks pretty clear that they do not want to stop cheats, it is a good business for them when people must buy new game several times.

I think we need some kind of client side application to have a chance against this. Punkbuster was a good idea but of course you could find a way around that one too.

But regarding SMAC, it is sad kigen and the SMAC developers couldn't come to an agreement so the development could officially get going. Guess that won't happen.

blackhawk74, are you willing to share your smac version maybe? Could PM me or so

[Benoist3012]

Originally Posted by MrDanne But regarding SMAC, it is sad kigen and the SMAC developers couldn't come to an agreement so the development could officially get going. Guess that won't happen.

Scrapping smac, and starting from something new would be better, I've been working on my own anti-cheat, and I won't share it to public as well, because public source anti-cheats kills the whole point of it.