I can remember a very Very long time ago (pre Y2K) , reading an article by some "expert" that theorized that software communicating over the internet, especially video games could get mixed up with Military communications meaning that a user could inadvertently mobilize a B52 nuclear bomber with a click of his mouse on a video game.

This theory sounds ridiculous enough as it would be very unlikely that packets of data from different software going to different Hosts on different ports could get mixed up. PLus the recipient would have a hard time interpreting the data that may only be understandable by the application it was meant to be read by.

However I digress, does anyone know about this theory and if it had a name, and has anything like this happened in software.

Apologies if this seems like a Networking question but since its ultimately about Application data then I felt it would suit this Stack site.


Any other info would be welcome on similar theories or ideas.

  • 9
    I've never heard a serious theory like that, but that's essentially the plot to the movie War Games. :P Mar 14, 2013 at 17:02
  • In that movie though the guy hacks into a computer, he's not playing a game like Command and Conquer over the internet with his buddies when his packets accidentally get intercepted by a military command system :P Mar 14, 2013 at 17:03
  • 3
    I would hope that there is some kind of confirmation system for mobilizing B52. Such as: Communications guy: "We just got orders to send a B52 to bomb Iceland!" Commander: "Ok, I'll just confirm that with General Somethingorother" General: "No, I didn't send that message!" etc etc... Mar 14, 2013 at 17:09
  • 2
    There were lots of self-appointed 'experts' talking shite around Y2K. This sounds like another one.
    – Alan B
    Mar 14, 2013 at 17:26
  • 1
    @MikeBrown on second thought, make that Dr. Strangelove. Peace On Earth.
    – ott--
    Mar 14, 2013 at 21:52

4 Answers 4


This scenario of "nuclear apocalypse by inadvertence" would require some inordinate incompetence at some point. Namely, we can imagine a buggy router which mixes some packets together, and sends the wrong packet to the wrong destination. And then, inexplicably, the military system which receives the packet which, by a stroke of bad luck, contains what that system interprets as the coded order for "it's payback time, commies !" just goes ahead with it.

Normally (hopefully), command systems which can trigger the nuclear Ragnarok are supposed to resist wanton attacks from bad people with evil intentions. They use authentication and cryptography and dual control and many other things to prevent outsiders from launching missiles, even if they try real hard to craft packets which will mimic the normal launch packets. The scenario envisioned here is that a random bit flip in a router due to a cosmic ray just succeeded in mutating or redirecting an innocent game-related chunk of data into a one-way-ticket for fission inferno. In other words, the scenario assumes that all of the following hold true:

  • the security systems and failsafes on the controls for nuclear missiles are so awfully weak that a single misrouted packet is sufficient to trigger the launch;
  • none of the existing attackers, be they spies from other nations, nihilists who crave emptiness, or bored students who should know better, succeeded in doing a task so simple that a random bug achieved it (it's like being crushed at StarCraft, when your opponent is a cat who decided to take a nap on the keyboard; it is very embarrassing);
  • the nuclear strike force controls are connected to the Internet.

Any of these conditions requires a lot of people to be terrible at their own job, to a magnitude which is remarkable even if we consider that some of them are a public organization, and others are pre-pubescent geeks. The whole picture is then quite improbable.

It may happen, but a lot of improbable things are more plausible than that, and arguably worse, including the usual example of a ten-kilometer asteroid striking the Earth, something which has already happened in the past, will probably happen in the future, and totally dwarfs out our puny nuclear arsenals when it comes to destructive consequences. Such an event occurs on average about once every 40 millions of years, i.e. has probability 1 in 40 millions to occur every year. Anything which could happen with a lower probability is therefore negligible.

A generic name for such theories could be "misaligned danger assessment". It is worrying about the very improbable instead of the very probable. If you want to live longer, don't worry about nuclear missiles; instead, quite smoking, drink less alcohol, and eat more vegetables and less fat.

  • 1
    +1 for "nuclear Ragnarok"
    – bedwyr
    Mar 14, 2013 at 21:50
  • 1
    +1 for taking my answer and turning the volume up to 11 Mar 14, 2013 at 22:18
  • 2
    +1 It's like being crushed at StarCraft, when your opponent is a cat who decided to take a nap on the keyboard; it is very embarrassing Mar 14, 2013 at 22:23
  • In the real world the packet is going to need authentication codes and it's going to be encrypted. The authenticator alone is going to be long enough to preclude guessing and in addition we need a minimum of the weapons system, yield, target and time (although the latter will normally be ASAP.) If they can get all that in 30 characters I would be very surprised. Since it's encrypted you need to hit that with 30 random bytes. 8 bits/byte, there are 2^240 possible 30 character strings. By chance? Hah! Mar 15, 2013 at 0:00
  • Thanks for providing this comprehensive answer, I knew this theory would be balls however I wanted to try and actually find the source of it but I am guessing it's been lost in the void of the interwebs (probably a good thing!). Mar 15, 2013 at 13:57

So there are several things that need to happen in order for such an event to occur:

  1. A game has to have a sequence of packets that match the protocol for say launching a nuclear missile (or the packets the game sends gets corrupted somehow to match the protocol)
  2. The packets have to be routed incorrectly to the server responsible for launching the nuke
  3. The server's fail safes and verification mechanisms somehow fail.

The likelihood of this happening non-maliciously is slim to none. The possibility for an intentional hack is there but I'm assuming there are so many layers of security and double and triple checks that it would be nigh impossible to bypass them all.

An interesting sidenote, there was a predator drone that was intentionally hacked by sending it invalid GPS data causing it to crash in Iran. I don't know the details, but I don't think there is any encoding within the GPS system that a client can use to verify "yes this signal has come from a satellite hovering over the earth"

  • 1
    Do you have any sources for the crashed drone incident? I'd be very curious to read more about it.
    – MrFox
    Mar 14, 2013 at 18:04
  • @MrFox - google "GPS spoofing" and you'll be on your way. Reasonably well known and valid concern.
    – user53019
    Mar 14, 2013 at 19:03
  • 2
    Apparently it really is a well known incident! en.wikipedia.org/wiki/Iran-US_RQ-170_incident
    – MrFox
    Mar 14, 2013 at 19:46

It wouldn't be a stretch for the military to hire a game development company to work on something for them. I doubt they would start from scratch, so there would be similar code whether they knew it or not.

The game could be designed to host its own shared game over the Internet and even reach out to other games. Of course they would exclude that feature from the military's. Maybe not.


Does anyone know about this theory and if it had a name?

"Scare mongering", "B grade Hollywood movie plot", "Vested Interests" come to mind.

has anything like this happened in software.

The Ariane 5 rocket failure, and the AT&T long distance crash in 1990 were caused my something remotely similar to this - as far as both were caused by incorrect messages causing an otherwise working system to fail.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.