uMod - Custom anti-cheat (Rust, paid $$$) - Plugin Requests

Posted 7 years ago

Last updated 6 years ago

Icefuse Networks is currently looking for an experienced individual to make an anti-cheat for our servers.

The goal of this anti-cheat is obviously to catch people using the following
- Aimbot
- No recoil
- Suspicious player pvp behavior.

NOTE: This system is not designed to catch the user using said systems directly but rather to collect a set list of data and compare it to what normal players can achieve vs someone cheating would have. For example, people with specific parameters would be banned off of the server for x amount of time using enhanced ban system commands "supports timed bans" and would directly write a message to our discord using webhooks for suspected cheaters.

Budget: [300 USD] This is the placeholder for this project but can and will be changed depending on the complexity of the stated task.

Contact:
- Steam https://steamcommunity.com/id/icefuse_networks "Leave a comment before adding"
- Here on Umod

Posted 7 years ago

Last updated 6 years ago

NoRecoil calculation can be correct if it will be based on triangulation.

https://en.wikipedia.org/wiki/Triangulation

Calculating evenry 3 shots is problemathic in perfomance case. (200 players are shooting every second 5-10 shots)

P.S. best anticheat - teach players to press F7

Posted 7 years ago

Last updated 6 years ago

pretty sure that's not what he requested to tell him it's going to cause him lag and poor performance keep it on subject.

on a side note Icefuse Networks has enough resources at their disposal and some of the top-of-the-line best money can buy dedicated machines I'm pretty sure they don't have to worry about performance.

Posted 7 years ago

Last updated 6 years ago

It will cause lags any way. Rust architecture not so good to take all of resources that available

Posted 7 years ago

Last updated 6 years ago

yes but can be done there is a few big networks right now that use custom made ones and only suffer a little bit of lag

Posted 7 years ago

Last updated 6 years ago

That's not true, it will only lag if you don't know what you're doing

Posted 7 years ago

Last updated 6 years ago

I'm currently not looking for it to directly pick up on aim-angles, simply DEFINE a set list of parameters to follow if a player has "suspicious" stats.

4 months after

Posted 6 years ago

Last updated 6 years ago

#19

In response to Orange ():
NoRecoil calculation can be correct if it will be based on triangulation.

https://en.wiki...

first of all i would like to pay my respect to all developers , thanks to you guys rust is better game with you.

sorry for bringing the subject back alive but there is a question that i desperately looking for answer ; is that really hard to catch exactly same mouse movements ? i mean ;

Macro/script does corrections accorrding to how weapon recoils, it knows what weapon type and how it recoils, so when the time is correct it does corrections. ( >x:+20 y: +10 , x: +5 y:+12 bla bla )

no person can do that in a row, if we could , we would be Vulcan XD. you cannot swing your mause exactly twice, i know there is a pattern not single mouse swing, and there is alot of people to monitor can cause performance issues. but if we put performance issues aside, is that so hard to catch exactly same mouse movements ?

and basicly cant we make just that observation and send it to a another computer to analyze ?
is monitoring and storing "that data" cause performance issues ?

we could save some of our processor for that issue, its a real problem that makes people quit rust.

Posted 6 years ago

Last updated 6 years ago

#20

In response to kuduz ():
first of all i would like to pay my respect to all developers , thanks to you guys rust is better ga...

I used to use an early Anticheat plugin on an old Rust server and picked up a few cheaters.

Very keen to get something like that again for a current server.
We used to be able to run the plugin on only the suspected players for X Minutes so the drain on the server resources was limited.

I have found a Russian site where there is a paid plugin that seems to offer the solution, but would require an English translation.
I would be happy to pay for any AntiHack put out that will compare bullet trajectories or mouse inputs.
Also reporting which attachments (If any) were on the gun
If the acuracy is higher then X% then make a decision based on what is really possible

I am mainly interested in catching norecoilers and aimbots.
I have noticed most of these players that are suspect do not use any attachments on their guns but seem to have none or hardly any recoil.
It feels like about 80% of players use them these days, I might be just getting more paranoid as I get older.

Here is what the Russian site claims their plugin can do:

AntiCheat plugin - private anti -cheat for your server.
Reports on the work of such cheats as FlyHack, SpeedHack, NoClip, and many others.

In one plugin:

- AntiCheat
- OfflineBan

Features:

- Saves all types of violations in oxide / data / AntiCheat / DetectLogs
- Convenient visual logs in chat for filtering out false detections.
- Detailed information in the log about the player when it is detected.
- Sending logs to administrators in the chat.
- Preservation of admins data
- Save data on the players (for future use)

Detects:

- Flayhak
- SpidHak
- AimHak
- SilentAIM
- KillVallHak (DPS through obstacles such as walls of the house)
- Macro (norecoil)
- AimLok
- WallHack - testing

Chat command:
/ ac - Enable \ Disable admin debug (Enable / Disable administrator check.

Console commands:

aim.server - view average Aim (head-to-hit ratio) on
aim.check server STEAMID - check Aim player
check.server - view kdr and Aim players who are currently on

Posted 6 years ago

Last updated 6 years ago

#21

In response to kuduz ():
first of all i would like to pay my respect to all developers , thanks to you guys rust is better ga...

Soon, (a week or two) I am going to publish a plugin on umod, which can detect and show low recoil shots (in a standing position and in motion) and also detects and show projectiles, the trajectories of which may have been changed using a cheat. It remains to clean the code a bit, translate messages (from Russian) and write manual with correct English grammar. And I also need to do final testing with cheat and macros. For this, I am looking for players who have them.

Due to limitations of the ddraw function, only server admins will be able to use the plugin.

Unfortunately, the developers (support staff Holmzy), refused to consider the proposal to add changes to the game code that would help make the calculations more accurate. After the April update, the trajectories of the projectiles have a random floating trajectory deviation, which is calculated on the client side. But no data is transmitted to the server by which it is possible to accurately recalculate such trajectory. Also, for accurate detection of low recoil, I would like to use the position and direction of the player’s gaze, but unfortunately, this data is transmitted to the server with a delay. So I was forced to use the position and direction of the shot itself, to which the random sway value has already been added.

3 weeks after

Posted 6 years ago

Last updated 6 years ago

#22

In response to Antidote ():
Soon, (a week or two) I am going to publish a plugin on umod, which can detect and show low recoil s...

I am keen to help you test it. We definitely have people shooting acurately with AKs with no attachments very quickly at 250-300 metres.
I class that as impossible, they claim they just move their mouse in a pattern. I call BS. LOL
You should start it as a beta test then move it to the paid site so you can get support for it.
I will donate either way.

PS. I was away for a few weeks with work, hence the late reply.

Posted 6 years ago

Last updated 6 years ago

#23

In response to AusCopter ():
I am keen to help you test it. We definitely have people shooting acurately with AKs with no attachm...

While I was testing the plugin, I found two issues that gave occasional false detection of violations. I need a little more time to find a solution on how to reduce the number of false detections.

Posted 6 years ago

Last updated 6 years ago

#24

In response to Antidote ():
While I was testing the plugin, I found two issues that gave occasional false detection of violation...

I previously tried the plugin NoRecoil (https://umod.org/plugins/no-recoil)
The main issue I found with it was it would detect slow shooting as no recoil.
Eg. A shot fired once every second could be detected as No Recoil.

To me that didn't make sense as anyone abusing the games recoil would be just holding down the fire button and getting shots fired at or very close to the same trajectory or target location.

A few suggestions I have to help design a "No Recoil" detection would be: (If possible to do)

1. Allow the plugin to be run time limited (Eg. X Mintutes from a config) per SteamID or PlayerName from RCon or in-game Admin/Mod
The idea being it can take the load away from running it constantly on all players if desired.

Possibly have an option to run it on new players for 5 minutes but queue up new players so it only runs on a limited number of players at a time, then an option to randomly run it on existing players already checked.

2. Only check for X number of bullets shot in X seconds, Eg. 5 bullets shot in 1 second or quicker

3. I am not sure what you have available to use on checking, if available either or both target hits (Eg. part of body) and bullet trajectory, (Being the same, very close or different path of bullets) possibly a combination of Trajectory and hit location.

4. Take into account if attachments are used and what difference that makes to recoil.

5. only pick on certain guns like the AK and M249 where no attachments on the gun would make it easier to detect No Recoil.

Just a few things I have been thinking about, hope they help.

Also the Russian Rust site has a paid AntiCheat, maybe they would be willing to do an English port for us. I would buy it.

Like any AntiCheat, it won't be perfect, but if the same style of player keeps getting caught you know you are on a winner.

On my server I see clans of say 2-4 players that shoot exactly the same all the time, there is no deviation and they seem to have identical shooting patterns and sometimes other surprising abillities that don't make sense.
They never use attachments and normally join the server and within seconds ask where the PVP is.
Those people will take down multiple targets from multiple directions with no hesitation as to where they are which also makes me suspect some sort of aim assist.

I see a lot of good players and know the difference between a good player and someone who is using assists.
Have been running Rust servers for many years and hate banning without solid proof which is hard to get.
Having a benchmark we can set for what is possible without assists would be a nice start.

EDIT: 1 other suggestion I had was even if we get the information late about what has happened, as long as all the timelines meet up it won't matter. Eg. If there is a way to historically time-match up things that happened a delay on data received by the server may not matter. (If it is possible)

Posted 6 years ago

Last updated 6 years ago

#25

In response to AusCopter ():
I previously tried the plugin NoRecoil (https://umod.org/plugins/no-recoil)
The main issue I fo...

In my case, the plugin collects the data of all the player's shots until the time from the last shot is more than 4 seconds. Then begins the analysis of the collected shooting data of the player. From the collected data, groups of shots are distinguished, the time interval between which is not more than a specific value (for example, for AK this value is 0.15625f). Groups of shots in which the number of shots of 5 or more are analyzed for No Recoil. After 9 seconds from the player’s last shot, the data on the collected shots is erased. The projectile lifetime in RUST is 8 seconds, there is no need to store information about them longer. I attach a link to the video, which shows how my plugin displays a violation of No Recoil. For testing the plugin, I do not use macros. I simulate a macro by simply increasing the interval between shots.

When analyzing shots on the AIM hack, I was faced with the problem of correctly calculating the trajectories due to the random floating trajectory of the projectile introduced in the April update of the game. The second video shows how the random floating trajectory of the projectile affects the new calculation of the trajectory.

Paid anti-cheat on the Russian site, in fact is a collection of plugins that are freely available. Includes NoRecoil plugin

Posted 6 years ago

Last updated 6 years ago

#26

In response to Antidote ():
In my case, the plugin collects the data of all the player's shots until the time from the last shot...

Hi Antidote,
Love your work. This takes a lot of the guess work out about where a bullet left and ended up.
I am happy to be your 1st customer as a paid subscriber.

As always with this sort of thing, it is evidence that can be used to help determine a possible cheater and nothing is expected to be perfect.
At worst it would be good to be alerted on RCON when a possible Aim/NoRecoil cheater is on the server, then know who to watch.

With the old Rust antihacks when they were working well, we nearly had 100% of all hackers caught just from the antihack. The hacks are a bit more sophisticated these days, they purposely don't shoot perfectly now so it makes it harder to detect them.

1 other piece of information that would be good to track would be attachments.