Anti-Cheat vs. Content Development

Many people in multiplayer games want devs to reassign developers to focus on the anti-cheat, but that's not easy. Let's go through some fallacies in this chain of thought.

In the world of game-development there's a very common misconception that I've seen even from other game developers.

This misconception is summed up as: "Why are you doing X, you need to spend more time on Y instead."

A more clear example is: "Why are you adding a bunch of new weapons/skins/maps, when what "we" need is a better anti-cheat."

Picking one aspect of that: To make a weapon you potentially will need at least: Artists, Game designer, animator, game programmer/backend developer.

Different roles

Now, there are some people who might work on weapons who also potentially COULD work with an anti-cheat, but what is true especially in larger studios is that you have specialists. People who, for example have spent years on making guns work and feel great.

A producer could, of course, reassign one of these specialists to work on the anti-cheat instead. But the likelyhood of that person being able to contribute to that work from day 1 is quite unlikely.

What is more true is that everyone who are already familiar with cheating and development of the anti-cheat are already tasked to work on its improvement. Regardless and independently of what else is being added to a game.

Just do what Valorant did!

Another notion I see is that "more anti-cheats should do what Valorant did, and use a kernel-level anti-cheat".

The problem with that reasoning is that Valorant's AC, while being very good, isn't primarily that good because it's a kernel-level AC. It's that good because it was developed for, and specially custom-made and tailored for Valorant. There is a team of AC-specialists who update this anti-cheat on a regular basis based on new cheats that pop up. To keep the AC up to par, they have to be constantly vigilant. A process that is both extremely time- and cost-demanding.

Members of such teams have often spent years on this. Some of them are potentially "rehabilitated" old cheat-devs who got an offer they couldn't refuse.

Many AC's today run on kernel level. Unfortunately, it's not as fantastic as some people seem to want it to be. On its own it essentially doesn't do anything. It essentially determines WHEN something is loaded. Lots of anti-cheats are able to detect kernel-level cheats even though they're run after them, and the other way around too. Lots of cheats that are run after a kernel-level anti-cheat work just fine.

A really skilled cheat-developer can bypass any kernel-level anti-cheat out there. The bottom line is to detect the method used, detect, and then take action on it (banning).

Weapon specialists are best left working with weapons

Again, here's where the specialists come in. Do you REALLY want someone who doesn't have years and years working with cheat detection jumping in to "help" with developing cheat detection?

No. No sane producer would assign someone to something like that, unless they already had lots of experience with it. If a beginner at anti-cheat starts working against cheaters, that anti-cheat is likely to fail miserably.

And herein lies the entire flaw behind the common but flawed idea of just reassigning people inside a studio to prioritize something. You simply need to have the right people working with the right things.

But... just hire more AC-specialists, then!
I had a feeling that you would come up with that argument. Here's the crux though. These people do not grow on trees, and as you might have already figured out... the entire multiplayer games industry have them in high demand. Many of these clever people are of course highly aware of this demand and thus form their own companies where they aren't locked down into one project, but can charge multiple different companies for their services, rather than working on one game at a time.

This benefits the companies as well, as they can outsource the incredibly arduous and continuous work that is required. With this, gaming companies can spend less time worrying about anti-cheats, and more time working on the development of the game. But this solution comes with a few flaws... having one company working on several games is likely less effective than having that team working on one game. Some studios are successfully able to compensate for this by making extra internal efforts to work together with an external anti-cheat. But some don't, for their players' detriment.

There are no easy solutions

The bottom line when it comes to this whole debacle is that there are no easy solutions. The suggestions you see all over Twitter/reddit/game forums when it comes to this stuff does not work. The likelyhood of a random gamer on Internet coming up with an anti-cheat idea that the anti-cheat devs haven't thought of is tiny.

And reassigning developers is not the solution either. The fact of the matter is that as a consequence of the profitability of cheat development, we are likely to see more games struggle significantly because the people buying these cheats make the cheating business very lucrative.

The "light in the tunnel" that some see in the form of AI taking over anti-cheats is also not the all-encompassing solution that some hope for either. Like with all tools, AI can also be utilized by cheat developers. So at the end of the day, the tit for tat will continue, until perhaps more countries start legislating to combat this. I think that might just be necessary in the end.


- Tobias Solem Posted on: 2024-03-10 21:54:53

clicktothetop
clicktothetop

All rights reserved - © Copyright 2024, Tobias Solem .