Valve’s CS: GO Overwatch Are Always Watching

staring eye!
Did you wake up this morning feeling more judged than usual? Don’t panic. Just look around the room and see if you can spot the RPS Staring Eye of Judgement? Well done, you have chosen, or been chosen, to take part in our little experiment: the RPS Staring Eye of Judgement will, eventually, monitor your every move to ensure that when you sit down and make a comment on the site that you’ve made the correct decisions in life up to that point. I say “eventually”, because right now it’s just a helium balloon with an eye drawn on it. Could you just drag it around and pretend, maybe making whirring and clicking noises? Thanks! We got the idea from Valve, who’ve decided to hand the matter of CS: GO’s policing over to the community. The sinister sounding Overwatch will be community members who have the power to review cheating cases reported by players and ban those responsible.

Blimey. An interesting experiment from the Counter-Strike team.

Prospective Overwatch Investigators are presented with an active Overwatch button in the main menu, which indicates that there is a pending case for them to evaluate. The investigators can then choose to participate by watching a replay (eight rounds’ worth or roughly 10 minutes) and selecting a verdict.

If the investigators collectively agree that an offense has occurred, a ban will be issued. The ban duration will depend on the severity of the offense and the suspect’s history of convictions, if any.

They’re only out to snag cheaters: the system will only identify the player as “The Suspect”, and both chat and voice won’t be in the replay files, so people who use chat to air unfortunate political discourse or play obnoxious music are safe for now. The language here is pretty sinister. “Overwatch”? “The Suspect”? Aren’t you being a bit overly dramatic? That’s exactly the sort of language that makes people go a bit funny with power.

Via PCG.


  1. MadTinkerer says:

    This needs to be implemented for Half Life 2 Deathmatch. Not that most people care about cheating in HL2DM servers (last time I tried playing it, I couldn’t find a server that used vanilla settings or even an appropriately-themed map), but it’s far more appropriate for the backstory of The City 17 Resistance.

    (Anyone else realize it’s a little weird that HL2DM hasn’t been updated with any content since before the Orange Box? Timeline-wise, it’s supposed to be the “missing week” in Half Life 2, but a heck of a lot happened in the Episodes released since HL2DM. Pretty much all of the HL2DM map locations were vaporized at/by the end of Episode 2. I hope Valve release a CS:GO style updated version of HL2DM when Episode 3 is released.)

    • Jikid says:

      You really enjoy hope, I see. :D

    • hamburger_cheesedoodle says:

      HL2DM has a story? I played the heck out of that game, and this is news to me.

    • Kobest says:

      Oh man, HL2DM, those were the days :'(

    • SwENSkE says:

      Keep on hoping.
      The Duke finally made it, too. I just hope Episode 3 will be better……

  2. hamburger_cheesedoodle says:

    ‘Overwatch’ is also a military term for someone who is in a high/concealed position, protecting something. I think it is meant more like that and less like psychotic-transhuman-policeman-with-a-stun-baton.

  3. amateurviking says:

    Valve seem to spend most of their time these days working out ways to get their customers to their job for them :)

    • baby snot says:

      I think it’s part of Valve’s overall business model. As little distinction between employees and customers as possible by design. Everything Valve does should be seen as attempts to scale the company by non conventional means. Aside from making games that is.

    • Screamer says:

      Now if only they can get the community to develop HL3 for them. * scratches chin*

    • Ross Angus says:

      If only RPS would let us flag spam again…

  4. GameStunts says:

    Why select a random game? What good does it do? Surely a reported incident is a more appropriate selection?

    What do you mean? The quote from the CS:GO team says they’re asked to review a case where cheating has been reported, not a random game?

    • Craig Pearson says:

      I rewrote a quote and took the thing out that I was referring to: “Investigators are presented with a replay of a randomly selected eight-round segment from an accused player’s match, and their task is to determine whether or not that player has committed any offenses during that replay.”

      I’ve yoinked the statement, as I can’t determine either way what it means, but I’d imagine it’s selected from a match that someone reported.

      • GameStunts says:

        Ah, well that does make a bit more sense. And in that context I agree with you, surely you should watch the match that people are complaining about rather than a bunch of random ones.

    • SillyWizard says:

      Matches are played to 16 (best out of 30). Overwatch cases will grab a chunk of 8 rounds from a match where cheating has been reported.

      Typically if someone is cheating, they’re not going to do it for only a round or two (and if they did, it’s unlikely anyone would bother reporting them). So having a group of rounds will generally show if there’s any funny business going on. (And I believe it’s 8 rounds in a row, not a random smattering of 8 rounds.)

  5. Archipelagos says:

    Valve + Overwatch in the same sentence equals my morning brain screaming OMG VALVE IS WORKING ON AN XCOM TITLE YESSSS. Dammit brain, reality is never that good.

  6. Senthir says:

    League of Legends does something very similar with their Tribunal, and the Tribunal has been a resounding success for them.

    With the Tribunal it’s almost exclusively people being jerkbutts and not cheaters, but the same concept applies; players handle punishing players. It worked extremely well and they have all kinds of ridiculous math and graphs and charts to back it up.

    I don’t play CS:GO but this seems like a solid move.

    • GameStunts says:

      I agree. Companies with a staff of people cannot possibly police X million players complaining about another player. This should help alleviate some pressure on the devs while allowing players to actually weed out problem players in the community.

    • Vorphalack says:

      I don’t agree. The last time I was anywhere near LoL, the unwritten rule of the tribunal was ”everyone is guilty.” That was a logical approach from the users who wanted to maximize IP gain without putting in any effort. What this means for the players, is that if you get reported to the tribunal, you will probably get a penalty regardless of weather or not you have done something wrong. As there was no penalty for false reporting, the tribunal just turned into a sledgehammer aimed at everyone.

      The same sort of thing is happening in DotA 2 at present. The player report system has been altered to hand out chat bans to people who are reported for chat abuse. The direct result of putting this sort of power in players hands is that it is being abused regularly. There is no penalty for false reporting as no one actually checks the system, so rage reporting is causing problems. The effect on the game is ridiculous, as now people are really cautious about saying anything in case they attract the attention of a report troll. In a game that requires some communication, this is a serious handicap.

      Conclusion: pier reviewed moderation does not work unless the moderators are being watched and evaluated. If that is the case, you might as well just pay for some admin staff to do the job properly.

      • Gnoupi says:

        LoL tribunal doesn’t reward with influence points anymore, so that decreased this effect.

        There is now a page showing you the judgments for the cases you voted on (link to, and from my perspective it seems rather close to a good judgement for most of them. It’s even more forgiving than I would be, at times.

      • darkChozo says:

        As far as I can tell, there’s never been much evidence behind claims of Tribunal abuse. It seems to be either be a theoretical thing, or an opinion held by those with a higher tolerance for abuse than the general community (I’ve seen things along the lines of “trash talking is a part of games” for why a ban is unjustified). Well, also as a convenient excuse for why you got banned; there are an impressive number of people who are convinced that they were victimized by the system when their Tribunal report shows some rather dickish behavior.

        For what it’s worth, Riot’s put out a bunch of statistics that suggest the opposite; specifically, among other things they’ve said that the average Tribunal user is actually less likely to punish a given case than a Riot employee would be. Not exactly a third party audit or anything, but, well, it’s something.

        • cyrenic says:

          ^ Basically this.

          The reports of “abuse” generally come from very vocal forum posters that are complaining because they were correctly warned or banned because of the system. LoL has a lot of negative behavior in it so naturally you’re going to see a lot of people complaining when they get marked by the system.

          And a large % of the people that get just a warning never show back up in the tribunal again.

          • Vorphalack says:

            I hate posting anecdotal evidence, but it has happened to me once in LoL. I got a bad couple of days games where I got match made with several stacks who trolled the game and reported me. In DotA 2 i’ve been sent to the low priority que 5 times without doing anything to deserve it, one game was actually after we won a 4v5 against a pre-made and got report abuse from the enemy team. Again, it’s mostly pre-mades abusing the system and venting on the random. That’s where most of the system abuse comes from; pre-mades who have enough people to trigger an automatic flag against one player. It certainly used to happen in LoL (can’t say if it still does) and it was happening in DotA 2, although it’s gotten better since the solo que has been introduced. Dismissing all accounts of unfair judgements as ”the vocal minority who actually deserved it” is simply wrong.

            Anyway, as I mentioned above these player driven report systems impact the game in ways that are impossible to record statistically. Playing DotA 2 since that chat mute was introduced, I have observed that people are just more reluctant to speak. That is not healthy for the game, I don’t think a system that makes people afraid to talk in order to cut down on trolling is serving its purpose well.

          • darkChozo says:

            I’m less familiar with DOTA 2’s systems, but in LoL, getting reported by itself doesn’t result in any action besides getting yourself in Tribunal, so report abuse tends to be moderately self-limiting (though there’s definitely something of a report culture, which I think Riot should be working to combat). And, no offense intended, anecdotal evidence of that sort is really quite useless without a reform card; if you browse any LoL-centric forum, you’re bound to find a number of stories similar to your own, and 99% of the time the ban was justifiable despite the premade abuse (ie. one game is premade abuse, but 3 of the remaining four have punishable behavior).

            The system’s not perfect – there’s a self-positive rate, and while that’s unavoidable, getting punished unfairly sucks regardless. But overall, I think it works at least reasonably well, and it’s one of the few systems that actually make an effort to correct behavior of those who don’t realize they’re being negative – rather necessary, given the low barrier to entry in a F2P game.

            As for people not talking out of fear, I personally don’t see much of that, but it’s hard to say. I think that a MOBA where people only called MIA and such would probably actually be better off than a MOBA with lots of toxicity (extreme examples in either case), but, well, that’s hardly a concrete fact.

          • Vorphalack says:

            ”But overall, I think it works at least reasonably well, and it’s one of the few systems that actually make an effort to correct behavior of those who don’t realize they’re being negative”

            I think almost everyone who causes trouble in games is aware of what they are doing. But anyway, the point isn’t that pier moderation is trying to help, but that it could be done better. I’ve already said that the best way to moderate a game is to hire admin staff to do it properly. Failing that, Valve and Riot are both quite wealthy, they could introduce a system where by a random sample of reports are independently verified and anyone found to be abusing the report feature would lose the ability to make reports. As i’ve already pointed out the biggest issue is complete lack of penalty for report abuse, something like that to help cut down on report trolling would go a long way to improving the system.

      • InternetBatman says:

        Not only do I disagree, I think your allegations are completely unproven and actively harmful. Most people don’t spam reports like you claim, and since you have to have multiple reports over multiple games, it’s not nearly as easy as you make out to get on the tribunal. I actively did the tribunal back in the day, and I found most cases to be particularly egregious. Most of the borderline cases were held not-guilty or whatever.

        Also, the judgments are on a time limit, so you can’t automatically click guilty for a bunch of cases in a row. And like people said above, now they don’t even grant IP.

        link to

        • Vorphalack says:

          ”actively harmful”

          You have to be unbelievably naive to think that player driven moderation has no problems. Calling the discussion ”actively harmful” is ridiculously hyperbolic, the sort of comment that tries to shut down criticism you don’t personally agree with. I have been on the wrong end of the report system, it does happen, and yes it is extremely difficult to prove. I would need to record all my games and provide video evidence of a penalty coming in after a clean game. A futile exercise because a) I don’t have that much hard drive space, b) a penalty can be applied hours or days after the bad reports, and c) one mans complaints wouldn’t change the system.

  7. ZHsquad says:

    My job is basically a social media consultant. I have to sometimes (based upon my customers wishes) allow their regular and most highly active users to become volunteer moderators. Now it may not come as a surprise when I say this usually ends up pretty ugly. At first the users treat everyone equally and wield their power responsibly. But after a few weeks they always end up abusing it. Always. Giving people power over others when they themselves are not tethered to something valuable (say money or a job) will not end well. The internet brings out some dark things in people when it comes to justice and judgement.

    • joe balls says:

      Your experience is not indicative of human nature as a whole. Look to forum moderators for evidence.

      • DrScuttles says:

        Or going to extremes, the Stanford Prison Experiment. There’s lots of ways people can abuse power, be it behind the veil of internet anonymity or the role playing of prison guards.
        Of course, there’s plenty of ways people responsibly handle power; they’re just not as interesting to read about.

      • AngoraFish says:

        My experience, as a regular poster on a wide range of different forums for nearly 20 years, is that the vast majority of volunteer forum moderators are a bunch of petty, hypocritical, self important, inconsistent, holier-than-thou, overzealous, grudge-holding, rent-seeking jerkoffs.

        Similarly, as a Wikipedia editor with well over 1000 substantive edits, my experience is that Wikipedia is even worse, free as it is from any hint of appropriately remunerated oversight whatsoever.

        Full time employees are, at least, much more hands off; intervening in only the most egregious cases, a much more sensible policy that is significantly less prone to abuse.

    • Deadly Sinner says:

      Except in this case, no one person has the ability to ban anyone they like. The anonymous, random Overwatch members must come to a consensus on the accused cheater. They cannot Overwatch someone they have a grudge against, because their matchup is random. With no voice or chat, there is nothing to hate or get attached to. They are simply looking at footage of gameplay and determining whether someone is cheating. It’s about as mechanical as you can get.

    • derbefrier says:

      Moderators don’t have it so easy though. Theres a reason they all start out nice and end up assholes by the end. Its because people are bastard coated bastards with a bastard filling.. People make shit up about them, lie, insult and even threaten moderators who at first want nothing more than to help make a good community but naturally after taking so much abuse they become vindictive and get that “fuck everyone” mentality. Dont act like they just turn into monsters for no reason. we are the company we keep after all and having to put up with so much hate every day will turn anyone into an asshole eventually.

      • Brun says:

        This is hardly unique to volunteer moderators, it can happen even to paid staff – in the “early” (2005-2007) days of WoW there were several well-publicized cases of Blizzard’s CMs lashing out at the community (and obviously losing their jobs over it).

    • Widthwood says:

      (sorry, wrong reply)

  8. FuriKuri says:

    This’ll end well.

    I can’t count the number of times I’ve been accused and/or banned from servers for “hacking” by skill-less tards or butthurt admins. Despite, of course, never having used anything of the sort. While immensely frustrating it was always offset by the fact that, hey, there were always more servers in the sea.

    So now I have to worry about being locked out the game entirely and, who knows, maybe even others too (since a VAC ban affects more than just the game you were being naughty with?).

    And the funny thing is, I’ve never considered myself to be *that* good.

    • Mccy_McFlinn says:

      I hope the system is used to get rid of the most obvious hackers. I played CS: GO for the first time during the weekend and the first server I joined, there was an absolutely blatant aimbot user. Ruined the experience for me and made me remember why I stopped play CS: S.

      The next two servers I joined were much better – I still had my arse handed to me but at least you could see it was through skill rather than cheating. I did see several people slamming the top three leaders for cheating and that’s where the problem lies. It doesn’t take much of an education to determine who’s cheating and who’s just plain good – so I’m hoping any cases submitted will be judged without prejudice (which is never how an accusation comes into existence).

      When it was fair, I found CS: GO far more fun that I ever found CS: S. The addition of Deathmatch and the random weapons mode (which may well be in CS: S by now but I’ve not played in many a year), made each new spawn a mini adventure; but will all games of this ilk, it always seems more fun when you don’t expect to get many kills – as soon as you start gaining any skill, frustration pours when a round isn’t going your way.

    • dE says:

      Absolutely. I’m in the same boat, I probably have a ban on an estimated quarter of CS1.6 servers for supposed cheating. Only thing I did, was to use my bloody headphones and listen to footsteps. Folks are really quite triggerhappy and I somehow can’t see the Overwatch not turning out trigger happy as well. Same with Forum moderators, where most start out great and quickly turn into petty tyrants.

  9. honuk says:

    great, now we have even more people at Valve doing absolutely nothing

    • Waleebe says:

      Maybe they’re all working on something, something big…

      Just kidding.

    • SkittleDiddler says:

      Whadya mean? The people at Valve are always working on important projects. Like trading cards. Or sweat detectors. Or cross-marketed hats. Or cleaning Gaben’s buttcrack.

      • PoLLeNSKi says:

        You are so correct:

        link to

        Only 19 games in the last decade (ok a fair few sequels/episodes/DM versions of singleplayer games in there, but we’re still looking at roughly one EXCELLENT game in each year.)

        …that and messing about with steamboxes, VR and maintaining the largest digital distribution network (for games ofc) in the world…

        Damn slackers deserve all the original fat jokes the internet can think up.

        • SkittleDiddler says:

          I didn’t make a fat joke. And we’re not talking about what Valve’s done in the past.

  10. bstard says:

    Sounds like a good trade-off to me. In practice, hackers are always a step ahead. This social control will catch the excessive cheaters before the slow and limited software catches up. I also dont understand the policy we the people versus the elite freedom of being an asshole tone of the article: almost every single FPS is ruined by hacks. Any attempt to counter this is very welcome.

    • P.Funk says:

      [quote]almost every single FPS is ruined by hacks. Any attempt to counter this is very welcome.[/quote]

      Correction: almost every single FPS is ruined by hacks on servers where there are no admins.

      Its always been my experience that you don’t float around games most of the time if you want a good experience, you find dedicated servers run by closed communities with a good reputation. The best servers always involve people who actively admin most of the time and nab the hackers quickly.

      Its this console mentality of the global server hosted by the publisher that befuddles me. As someone who’s played TF2 since launch it just seems like a pretty straight forward fact – Valve Server = Hacker Haven. Don’t like that? Wonderful, go find a nice community, seed their server, use a mic, and enjoy the benefits of community. Maybe even throw a few bucks their way and get a reserved slot. Very civilized.

      Communities like that are like direct democracy or something. Or free market capitalism or something. I dunno, pick one based on your ideological life’s story. The point is that the power of a community’s health is based on the integrity of its admins and the recurring donations of core members. Most people enjoy this free of charge but maybe don’t get in when the server is full. Its what defines PC gaming over consoles (aside from a mouse and keyboard, and good games).

      Here’s an idea. How do you lodge a complaint against the guys who judged your case? Can you? Who arbitrates that? WHO WATCHES THE WATCHERS!?!?!?!?!?!?

      Ugh, I want 1999 CS back.

      • Monchberter says:

        @ Pfunk.

        Agree with you 100%. A decent community is what keeps games alive. Find one you like and stick with it.

      • Hahaha says:

        It’s sad you even have to do that. dicks just need to be perma banned with a variety of methods so they will have to learn how to do some technical things before they can play again….will keep most of them out.

  11. Didden says:

    Have they actually used the internet? This won’t end well.

  12. MariaStepp46 says:

    until I saw the bank draft of $6203, I accept …that…my best friend was like they say realy receiving money in their spare time at there computar.. there sisters roommate haz done this for only about ten months and resantly repayed the depts on there home and bought a great new Lotus Elan. go to,

  13. Eater Of Cheese says:

    Dear CS:GO,

    Fuck off.

  14. Joc says:

    CS did used to have that ‘vote’ command, by which if enough people in a game reported a certain player, the alleged offender would be kicked. Of course, that’s not as severe as banning someone, but it’s not as though this move doesn’t have some precedent.

  15. Monchberter says:

    Oh this brings back fond, and not so fond memories of running a decent Team Fortress 2 server and community – namely the trouble in finding decent admins among your player base.

    While you always had a handful of people you knew would handle their responsibilities well and be fair, it always was a constant struggle with inconsistent admins who swung wildly between benevolence, fairness and petty protection of those they wanted to remain friends with, letting them get away with all manner of crap behaviour. Giving someone you know you can trust 80% of the time a hefty bollocking for permitting bullying of other players or homophobic abuse is not a pleasant place to be in, particularly when they’ve contributed so much to making the community what it is.

    I know servers, as communities can tend to have their own particular (inconsistent) rules (‘yes we accept griefing and spawn camping’) but I do hope to some extent that this system can play nicely with such ‘local laws’ if it’s rolled out beyond CS:GO. Hopefully it will work alongside the VAC system to give some credibility and trustworthiness to those servers who would ‘opt in’ to the Overwatch.

    • Widthwood says:

      The difference is that steambans/overwatch system does not allow admins to ban you right away in-game.

      To ban a player they have to upload a demo of your play to a central server, where it will be publicly reviewable (in steambans, not sure if demos will be public in overwatch), and other admins will give their verdict. If demo does not contain any cheating, and they still banned you – this is easily provable since everyone can check their decision.

  16. Minksta says:

    I don’t understand why people are bashing this system. Valve has always used a delayed ban approach with VAC updates so that they can catch as many hackers as possible without updating VAC (which in turn the hacks gets modified to get by VAC). It is a constant game of cat and mouse and this system will be perfect for that 2-3 week window where hackers can run free in the game. The higher ELO ranked matchmaking games are plagued with hackers unfortunately so this is the best approach to getting them out. There is an anonymous panel of people who can vote three different ways, basically “definitely hacking”, “possibly hacking” or “inconclusive”. Its also important to know that not one persons vote will result in a ban and that the people that are chosen to be part of overwatch have hundreds if not thousands of hours in the game and have a high amount of MM wins and high ELO. I for one am very happy with this system and I think it is the best way to weed out the hackers.

    • Eater Of Cheese says:

      This is a travesty. Just make a good game. Formalising a Stasi-style network of informants to police an FPS?

      If this were a kind of oblique commentary on the inherent hypocrisy of the modern consumer-surveillance state, I’d laugh, but I suspect it’s not.

      I’m all for measures to protect against naughty people, but this goes too far. Lost a customer, and a fan.

      Sublime’s classic ‘Get Ready’ sums up what I think is an appropriate response. link to

      What’s next? Game developers and publishers employing information ‘managers’ to patrol forums and comment threads to control the conversation? Get lost.

      • Widthwood says:

        Imagine playing your favorite online FPS and some people from other team just instakilling enemies, spoiling fun for everybody. What you are saying other players HAVE to be powerless in that situation, otherwise the game becomes police-state commentary?? In real life you can just try to stop criminal yourself, or call the police – both actions don’t apply to CS.

        Calling admins works only against complete idiots even if they come in time, since anyone smart just stops cheating when someone is watching in spectator mode. And on most servers there is a command to check if admins are on server.

        And ordinary instant voteban system does not work well for CS. With CS being very technical even minor advantage can lead to major difference in capabilities, and there is a huge difference of skill between players. Because of that newbies keep votebanning higher skill players and higher skill players usually can’t gather enough votes to ban a smart cheater – after being banned by others time and again without any reason they give much higher benefit of doubt to others.

        “Record now – ban later” system works because it separates skills necessary to record and to detect cheats, and lowers amount of effort needed to stop a cheater. And it leaves hard evidence with ability to appeal in case of wrong ban. By the way it wasn’t even invented by Valve – they only improved it with better protection against admin abuse, the system with similar idea was developed by members of community years ago and is used to this day.

  17. Widthwood says:

    What they want to do is to take already existing and working system of / , improve it and extend it to all servers. Those saying generic pieces of “wisdom” like power corrupts and all that – really don’t know what they are talking about and probably haven’t even read the article.

    The system works by someone recording and uploading a demo of a suspected cheater, and later a GROUP of experienced admins will review it and vote on a verdict. These are not your average angry children screaming CHEATERR!!!111 at everyone even moderately skilled, they actually know what cheats people are using, how these cheats can be detected and usually have a separate account where they test latest versions of cheats themselves. Couple of years ago playing CSS on steambans member servers meant almost complete freedom from cheaters. There were some crazy ones that kept buying new CSS boxes after being banned, but these kinds of persistent cheaters are obviously rare.

  18. AlienMind says:

    So, Gabe’s outsourcing frenzy has now brought back witch burning. Happy burning, everyone! *Walks away because he has no Steam installed*