Gaming
We speak to the man who wants to help eradicate cheating from tournaments with a hardware fix.
Every sport has cheats. It's unfortunate, but in a world where everyone wants to win, that extra edge over the competition is enough to lure people over to the dark side. It's happening in the world of eSports too, though as we've said before, it can often mark a sport's growing maturity, as those in the industry are forced to address it.
Just mere months ago, the Counter-Strike: Global Offensive community was rocked by a cheating scandal, and pro gamers have come under intense scrutiny to prove their skills are legit. And you only have to log into a public session for a few rounds to start seeing cheating accusations hurled around.
That sort of stress is not ideal for pros when there's potentially millions of dollars on the line – but there's help at hand, beyond Valve's (in)famous Anti-Cheat system (VAC). Beyond software entirely, in fact. David Titarenco, a 28-year-old coder, former Counter-Strike pro and anti-cheat software contributor has come up with a clever piece of hardware based on the Arduino platform that could help get rid of cheaters in pro gaming.
It's called Game:ref, and unlike previous anti-cheat software, this tiny gizmo is going to be a lot more difficult to circumvent because it's based on a simple principle – you have to plug your mouse into it, and it'll compare your physical movements with those happening on-screen. To find out a bit more about the project, we spoke with Titarenco ahead of him launching the Game:ref on crowdfunding site Kickstarter.
Where did the basic idea for the Game:Ref come from? Did you think a hardware solution would be the next step in stopping cheaters?
The idea for Game:ref actually came out of my frustration with CS:GO's default matchmaking. Every now and then, I'd like to unwind, have some fun, and play a CS match. Unfortunately, I'd waste my time more frequently than I'd have fun. Playing cheaters sucks and I want to fix that. Since I knew what limitations software has, I thought about how consoles beat cheating. And the answer is simple: hardware. Consoles are almost completely devoid of cheaters because they provide anti-cheat solutions baked-in their hardware. We need to do that with PCs as well.
Have the recent CS:GO hacking scandals made you want to push work on the Game:Ref?
Absolutely. What was first just a project to pass the time turned into a full-fledged idea that I thought could not only benefit games that obviously are in trouble, like CS:GO, WarZ, DayZ, and the ArmA series, but also games like Dota 2, League of Legends, and Blizzard's upcoming FPS, Overwatch.
You’ve played Counter-Strike professionally in the past; has your background experience in pro eSports given you the kind of knowledge of what players and leagues would want?
I'd like to think so. Having played Counter-Strike professionally, I know that cheaters have constantly been an issue. For the past 15 years, every time you make a new player pick-up or play online, you need to be vigilant. Even though companies like Valve and the ESL and even ESEA try their best, it's obvious that the current solutions aren't up to par.
So how does it work, in layman’s terms?
The device simply compares input that the player is making with their mouse with what the game says is happening. If the two don't match up, the player is cheating. An easy way to understand it is by looking at how a triggerbot works. The idea of a triggerbot is that the software cheat 'simulates' a mouse press when someone is in your crosshairs. But the user doesn't actually click! So with Game:ref, we can see that.
The Game:ref looks like it’s mostly suited to FPS games, but do you think it would be feasible to use in MOBA and RTS titles too?
Game:ref is absolutely suited for MOBAs and RTS titles. In Dota 2, for example, we already have illegal auto-last-hit macros and auto-skill-use hacks that have been downloaded by thousands. These are essentially all 'triggerbots' that work with the keyboard instead of the mouse. They are very easily detectable.
What are the Game:ref's limitations?
For the time being, Game:ref only targets software-based input cheats. I have some ideas that could target cheats like wallhacks, ESP, and hardware hacks, but for now I'm focusing on input-based cheats. After all, Rome wasn't built in a day.
What’s wrong with current software anti-cheat implementations? Are hackers and cheaters always finding ways around them?
I think this question can easily be answered by anyone playing more than three matches in Valve's CS:GO matchmaking, or by checking the Dota 2 subreddit. Cheaters are out there and they are killing the experience. In some cases, they are killing indie games. WarZ, for example, was a crowd favorite, and as was Rust, before they were both gutted by cheaters. If indie developers could focus on making awesome games and not on banning annoying cheaters, multiplayer indie games might be more of a thing.
What response have you heard from pro gamers? Do you think it’s a positive step in helping take eSports mainstream?
I've had some people from Complexity, SK Gaming, and a few high-profile streamers reach out. I would say everyone seems onboard with making online PC gaming a more enjoyable experience. After all, most cheating on consoles has been eradicated, why should PC be so far behind?
When do you think you’ll be going live on Kickstarter with it? Are there more bugs and kinks to iron out?
The Kickstarter will be launched later today. I'm currently working full-time polishing the prototype and setting up the campaign. If I do get funded, the next step is getting game developers and LAN organizers on board. Obviously, more money raised will mean a bigger impact on gaming as a whole. My fingers are definitely crossed! Get the best pro gaming stories delivered straight to your inbox with the Red Bull eSports newsletter.