Is anything being done with rfactor2 to stop ram cheats?

+1

 

If you finish this i will buy it off you

 

I can see the merit in the proposed solutions, but it's still nearly impossible to rule out cheating completely. Car performance is dependent on a huge number of factors, especially cornering speed. I guess you could calculate downforce yourself and combine that with tyre temperature, load, camber gain and whatever else is involved to ensure the 'grip' achieved isn't out of the ordinary, though tyre temp itself could be tampered with so then you need to try and calculate that with loads, slip angles, track temp, and whatever else... then along comes rF2 with variable track temp around the track (sunlight affected), variable wetness, tyre condition in different parts of the tyre and, taking a guess, more than a single surface in contact with the tyre at a given moment, and things are starting to be pretty complex. Still, you can do your calculations and add a bit on top for leeway... but any 'good' cheat will only utilise that margin anyway. Cheats that give seconds per lap are usually apparent just from watching the replay.

Davie, that's the sort of approach I could imagine taking to rule out the 'common' cheats out there, only advice I would give is to do a lot of testing and don't always trust the figures. I've had some figures that ought to be absolutely consistent come up with some oddities at times.

Still, I'm not saying things shouldn't be done, because even if it's never 100% certain you can pretty much eliminate all cheating by waving a big enough stick. As for the rest... well... it's not life and death. Enjoy the racing and move on.

 

Who wants to join a server i will not tell anyone how to cheat but to show a number of people at the same time we use isi bmw 2008 and a isi track you get with rfactor we all meet on a server all do some laps and me also so you can see how fast i an normaly then i activate the cheats and pull 5sec on all of you and everyone can see how easy it is then mabe some rfactor people can figure a way to stop this

 

True, but who determines what is valid?

 

What you could do is have a client sending the name and some major values through to the host. Every client needs to have this running in the background. The client could either send an OK/Failure signal (vulnerable) or all major data to the host.

A point of reference is important though. You would have 2 options: read the game files (if they are not encoded) or step in every car model and save the typical car data. If a client plays with abnormalities from the reference, kick it. Game files should be correct to the host by the game checks (if I'm correct), so cheating by adjusting the point of reference wouldn't be possible (furthermore you can check the game files on both client and server side).

I could imagine you would want to check on brakes force, RAM's, weight, downforce, drag, engine boost, engine power, tyre grip, all temperatures, fuel, gearing, and possibly more. It would be a hell alot of work to be checking the game versus own software (if you don't have all the exact formula's implemented in some kind of program). For a community edition memory readouts isn't the biggest problem, as I've collected quite some figures for several personal data acquistion telemetry and dashboard projects. Just comparing it to the right reference, deciding when it's not valid any more, etc.

It seems to me ISI could implement these checks the best way. If only the min/max values are checked to the reference on the host every 5 minutes, it's good enough.

 

That's a non-question for the system I proposed, but I probably haven't made it clear enough.

In a nutshell:

1. clients record control input during the race
2. clients send control input file to a validation server after the race
3. validation server replays input through its instance of rfactor.

If the lap times on the validation server are different than those reported during the race, then the client is using different models. (can't determine *what* is different -- just that something is different)

There's no way to subvert this client-side. Mess with the input data all you want before sending it to the validation server -- it will only increase the discrepancy.

On the topic of corruption: make the input data files available on a web server. That allows anyone to do the validation themselves. If your validation results agree with other people but not the admin, well then it's time to find another league.

 

Any anti-cheat system that trusts the client is dead on arrival or very soon after. Sorry to rain on the effort - but people need to stop trying to invent unbreakable clients. It's not possible.

 

it is possible the simulator needs a program that changes the memory values every 10 sec and can not be changed just as ram cheat alows you to stop the game from from changing then it wont work

 

The memory values are mostly constantly changing - as a result of the valid calculations being done. Reference values (those that would be specified in files) don't change because that would obviously cause problems - but whether you validate them against figures somewhere else, or try to move them around 'randomly', the game has to know where they are and there will therefore be pointers that can be used to find them.

As for a program 'locking' a value, more than likely it just polls the address at intervals and changes it to the set value if it finds something different. If you could actually lock a value in memory there'd be no cheats/trainers for games (and there are heaps).

As openloop says, you can't be 100% sure by relying on client reporting. The input-checking idea has merit, but could be an absolute nightmare to make as reliable as it would need to be.

If it's made much more difficult than it currently is, great. But nothing is unbreakable. That's not pessimism, nor defeatism. It's just a fact.

 

Well, we need damage control more then anything. stop the flood of easy-cheats, make it difficult for people to cheat. if its realy easy more people will try it, its its not-so-easy we will see less people getting into it.
Making it more difficult might also mean its easyer to detect?

 

i am trying to do something but look i was messing about last night and had suspension failure http://youtu.be/Kk9iSsAMZ64?hd=1

 

Uploaded with ImageShack.us this is getting out of hand now

 

... and you can easily pick those up with a custom plugin.

 

Sorry, replying again, bad form I know...

Let me just be a little clearer on this; that screenshot is obviously from a tool someone's made, but your concern is misplaced. Not only are those cheats as easy to detect as they are to implement, the fact that they're known about means anyone wanting to try and detect cheating will have them on their detection list.

It's the ones no one else has heard about that are the problem.

My concern is that you appear (to me) to think rFactor is almost solely susceptible to this kind of thing, and that you think it's possible to completely stop it (and therefore want ISI to do it). Every game/program is susceptible to varying degrees, and you can't stop it completely. Yes, you can reduce it, and I'm sure rF2 will have more in place. But, any single client-side system is breakable - in the same way that I could release a plugin that picks up those values you've captured and either reports to the current server or writes to an encrypted local file (to be sent to the server admins on request), and that'll work fine for a week or two until someone decides to break it because half the league servers in existence are using the exact same system.

I know having ISI stay on top of known cheats sounds great because it's easy for everyone else, but for a small team I really don't think it's feasible... and still can't guarantee you're not racing someone who's cheating. If you play any game online you have to accept it's possible. (driving, FPS, RTS... it doesn't matter)

I can only assume from your apparent shock at the sort of tools you're finding that you've never looked for similar tools for other games.

 

Nobody thinks it will be eliminated altogether, we just wanna have a basic system in place from anybody, ISI or community that will stop 90% of the wankers. It's too easy atm.

All the leagues I run in have nothing in place to detect cheats.

Except one... running Tyka. And the suspected cheater suddenly got slow and left, either through intimidation or the anti-cheat actually worked in that case. I think he was too scared to use it when we went to Tyka. He was the 3rd cheat in a medium sized league we've caught this F1 season, and we weren't even a few races in at that point. When nearly 20% of the field is cheating, you know that its become too easy.

Hardly any Leagues are using Tyka, at least that I can see in their lobby. Why is that? Is it flawed?
I'd like some opinions.

So far michael1 has done a great job of showing those of us who don't go around looking for cheats what they look like and how easy it is.

 

Very true. It is a bit like the dreaded debates on gear shifting and workarounds that exist to optimise shifts be it using hardware or software.

It then "explains" to people just why they are getting beaten, after all if I am driving down a straight and someone breezes past and we are both running flat out then only upshifts can be the difference, surely? Casting aside the guy behind might have carried 10mph more speed at the corner exit. It just becomes an obsession and becomes the responsibility of the devs to fix. Yet we want the same devs to allow us to run all sorts of different hardware and offer ultimate customisability of controls.

It becomes an epidemic. Fact is a guy with thousands of laps in a given car at Mills is going to have an advantage to someone doing a "realistic" race weekend with practice sessions that would mirror a real world competition. We have to accept some variability but once the dreaded cheat rumours start then any variability is under scrutiny.

Tough nut to crack. I agree with the suggestions that something must be done to stop it being too easy and some deterrent is in place but it can only go so far and the community should do whatever it can to instill a culture of fairness.

 

guy i have found something that works i tested it with the ram cheat also with the super cheat tool you see in the pic you can all test it is made by fsone all drivers must run it befor rfactor is then suspect of cheating the can prove they are not by sending the admin the log files but a proper hacker you can never ban from a server trust me you press ban as superadmin and it tells you player not found

hear rf scan tested works http://www.megaupload.com/?d=D372EXZ1 problem solved

 

rF-Scan - Anti Cheat for rFactor



Uploaded with ImageShack.us

A cooperation of "Simsync Pro", "International PRO Modding" and Sim-Racing.info

Today the new tool "rf-Scan" was published. rF-Scan checks whether a legal version of rFactor is used, monitors the RAM for wrongful changes and indicates any plug-ins, services and processes which are being used during an rFactor session.



Heute wurde das neue Tool "rF-Scan veröffentlicht. rF-Scan überprüft, ob eine legale Version von rFactor benutzt wird, überwacht den RAM auf unrechtmäßige Veränderungen und zeigt alle Plugins, Dienste und Prozesse an, die während einer rFactor-Session benutzt werden.

Alle Details zur Benutzung und zur Arbeitsweise sind im Downloadbereich von Sim-Racing.info zu finden.

The tool works in rFactor the following way:

- the rF-scan can be obtained via the SimSync and is placed in a separate sub-folder by the sync

- rF-scan can be moved/copied by the user to any other rF-installation and can be run either after having been placed in the rF-root folder or in a sub-folder

- creating a desktop shortcut is recommended

- starting the rf_scan.exe (or using the shortcut) will start up the tool's dialogue box

- choose the driver profile you wish to use

- to begin with, the rFactor.exe will be checked for validity. Should the file not pass the check it has to be replaced:

- start rFactor via the tool's dialogue box

- after having closed rFactor a folder named "rf_scan" will have been created which contains a logfile

- the logfile is an encrypted file and needs to be sent to the Race Control

- the name of the logfile is always a combination of the driver name and the respective date

- No data is transmitted online!

- No personal data is stored!


The following parametres and all plugins are supervised:

Folgende Parameter und alle Plugins werden überwacht:

Fuel, Water, Oil

Tire Wear Front Left

Tire Wear Front Right

Tire Wear Rear Left

Tire Wear Rear Right

Tire Grip Front Left

Tire Grip Front Right

Tire Grip Rear Left

Tire Grip Rear Right

Veh Mass

Launch Effect

Fwing Drag

Fwing Lift

Rwing Drag

Rwing Lift

Body Drag

Radiator Drag

Radiator Lift

Brake Duct Drag

Brake Duct Lift

Brake Torque Front

Brake Torque Rear

Engine Wear boost

Fuel Consm boost

Max Boost boost

FL Lat Tires Grip

FR Lat Tires Grip

FL Long Tires Grip

FR Long Tires Grip

RL Lat Tires Grip

RR Lat Tires Grip

RL Long Tires Grip

RR Long Tires Grip

Tires 20% wear FL

Tires 20% wear FR

Tires 20% wear RL

Tires 20% wear RR

Tires 40% wear FL

Tires 40% wear FR

Tires 40% wear RL

Tires 40% wear RR

Front Tires Compound

Rear Tires Compound

 

I use a 4gb patch, this picked that up as a cheat.