# How to defeat a villain who can predict the next day? [closed]

Let me clarify the exact powers and limits of said villain:

He can, with extreme accuracy, predict everything that might happen in the next 24 hours. He does not know 100% for sure what will happen, say, a few seconds after now, but his predictions are very accurate to the point where the probability that he is wrong about some situation is almost zero for all practical purposes.

He can predict everything from a range of one second from now, to 24 hours from now, in increments of one second. In other words, the villain cannot predict what is going to happen, for example, half a second from now (since it's both less than a second and a non-integer value), or 24 hours and 10 seconds from now (since it's more than 24 hours), etc.

Every event that the villain sees causes his mind to alter his predictions, and almost always, this causes the predictions to become more and more accurate. If an event has occurred that he does not know about, then he cannot factor it into his prediction calculations until he gains sufficient information about said event.

His predictions are powered by an advanced AI datacenter, where complex calculations are performed. The datacenter is powered 100% by renewable energy, is located in a secret location only the villain knows about, is protected by disaster-proof walls, and connects to the chip in the villain's mind by means of a heavily encrypted connection on a custom frequency, not through traditional means such as cell towers, etc.

This villain may be able to be defeated (eg: see similar questions on here), however, I'd like to know, is there a good way of explaining this that is relevant to the story? Something that isn't too confusing but isn't too simple.

"Defeated" here refers to making the villain stop his villainous acts, and putting him in prison/jail where he belongs, not murdering or killing him.

• Hello and welcome. It seems you are not building a world here, just writing a story set in a world. This would make your question off topic here. – Mołot Jan 29 '17 at 16:34
• @Mołot, we've had a lot of questions on these lines and most of them are considered acceptable. Usually the villain gets a time stop rather than prediction, but it comes to the same thing. – Separatrix Jan 29 '17 at 16:40
• @Separatrix most of them earned close votes all right, and "story based" is explicitly a close reason. – Mołot Jan 29 '17 at 16:42
• Story about one character? If he lives in our world, or a world you already built, is not world building. – Mołot Jan 29 '17 at 16:43
• @SRM I agree and I'm voting to leave it open but I feel like there are about ten near-duplicate questions along the lines of "villain has unstoppable power, how do you stop them" that overlap substantially. – Zxyrra Jan 29 '17 at 18:54

The biggest weakness seems, to me, to be that he has to observe something to predict it.

So, if his food is poisoned while he is not watching then he would have no idea what would happen, right? So I suggest a non-lethal poison.

Honestly, there are so many ways to kill him with that limitation.

You could just go to him with a Tazer (an electrical shock device) and shock him as soon as he sees you. The moment he can predict this is when he can see you... but most people can predict what is going to happen when someone with a Tazer comes through the door.

You could throw(/drop/hide) a concussion grenade at him... when he sees it coming, it's too late.

Edit: I think I know now what you mean. To make this a good story, I would focus on the process of his opposition slowly finding out what exactly his powers are and testing the powers' limits. That way, the trap is only the last step and not the most important part of the story.

The way you defeat him is exactly the same as how you defeat the many other similar villains which we have seen in the past on this forum. You create a trap which takes more than 24 hours to spring (and probably much longer), is unstoppable once it's sprung, and doesn't give a clear hint that it's a trap until at least 24 hours later. An individual with that level of predictive capacity is going to be swamped dealing with all of the self-interactions he's going to experience during the next 24 hours, so unless your trap is obvious, he probably wont see it.

I'll leave the particular trap up to you. After all, if this is a central theme to your story, you should have some creative say in what the trap actually is.

• Comments are not for extended discussion; this conversation has been moved to chat. – HDE 226868 Jan 30 '17 at 14:43

Send a huge army to arrest him. If your force has a lot (I mean A LOT of men) he won't be able to escape. He might come to a situation that he will know that he will be arrested no matter what move he does. It will be like a check mate in chess. He will know what will happen but he will not be able to escape it.

Another idea is to capture him using time travel. Send somebody back in time (more than 24 hours ago so that the villain will not be able to predict the journey in time) and catch him off guard.

The technical term for what you need, I think, is a xanatos gambit

What you need is a very long plan, with lots of ways for you to win, and as few ways out for him as you can possibly make. A xanatos gambit essentially has all paths leading to victory, including paths that superficially appear to be failures. Since he has 24hrs lead time, you will likely need several "apparent" failures in a row first, to get beyond his window of prediction and to bait him into actions which look like victories to him at first, but actually leave him in increasingly worse positions when other branches of your plan are revealed.

To put it more concretely - there is a plan which has a trap closing at 24hrs. Obviously, he will see the trap, and figure out a way around it - in fact, there's a nice, easy escape route in only five hours...unfortunately, it was planned, and is leading him straight into the 30hr trap! Maybe he can faint and dodge around that trap, but he has less time than he might have because he was focused on the earlier trap (and it's still active, so it limits his moves anyway). And if he dodges around that trap, well, the 36hr trap is really clever, it will take a lot of time to get around that one - especially if avoiding the 30hr trap lead him to commit a chunk of time he can't exit in the middle (say, a long flight). He has considerably less than 24hrs at that point, to dodge the 36hr trap, and there's a easy-to-predict way out leading straight into the 40-hr trap. If he gets past that, the 41hr, 42hr, and 44hr traps are all lined up to catch him, if he starts counting on having more time between traps. And in the meantime, you should have enough time where he's dodging, distracted, and generally paying more attention to what he's doing rather than what he's predicting to make the 48-hour trap absolutely inescapable, or at worst the 50-some hour traps all lined up next. That's why you need a trap longer than 24 hours, not because he can't see further as it comes closer to time, but so that you can shape his actions in the beginning hours in ways he can't yet see the end results of until he's already committed to a course of action - the dodge that would actually let him out of the 48-hour trap is lost because he can only see, and feel arrogant about dodging, the 24-hour trap.

And of course, not only does dodging out of earlier traps set him up for later ones, it can also (in proper xantos gambit style) achieve other objectives for you along the way - delay or destroy his own plans, or encourage him to be hasty with them (and thus easier to thwart), prevent him from creating or executing new plans, distract him from your various events and plans (to make his predictions less and less accurate), separate him from allies, show those allies he will turn on/abandon them (so they won't be future allies), enlighten oblivious pawns to his true plans so they won't help or may turn against him, give yourself good press, give him bad press, get him to underestimate you so he falls into a trap, get him to overestimate you so he ends up with paranoia or decision paralysis, triangulate in on his datacenter (by noting movement, his need to get access and his methods of getting it, or following various breadcrumb trails) in order to to destroy it or feed it false information or protect your own plans from it, deny him access to said datacenter because he's too busy running to log on and see the latest updates, deny him access to other resources, get him to spend resources in a non-optimal-for-him manner, and, well, anything else you can think of.

And if he manages to get out of that? Just keep going. If you don't have him in hand by the end of 48hrs, then start planning a new set of traps for the next 48hrs, since he's already exhausted and distracted from the last set. Don't give him time to stop and reflect, don't give him time to predict, and make plans based on his predictions (that plays to his strength, not yours), just hit and keep hitting until he's down. Don't pre-plan the traps exactly, in case he finds a genuine gap out or can see them further in advance because he knows the predictive events better, you should have enough flexibility built in to change things around depending on what he does, alter your plans (so maybe a bit of xantos speed-chess[previous warning still applies]), keep a couple people who are brilliant at improvising plans, or flying without plans altogether, so he finds it harder and harder to predict ways out of them, and continually scrap what isn't working and elaborate on what is.

Remember, there's only so much a person can genuinely concentrate on at once. if he's running about plotting and planning, and then thinking about how to get out of each trap as he predicts it, he is less able to concentrate on predicting what will happen with the next trap. And, he needs to know enough about events to add them to his prediction - so fast, flexible decisions and last minute alterations and adaptions are hard for him to see anyway. Between the two - hit him hard, keep your plans moving, and keep the attacks coming (including traps, outright fights, unofficial tricks, official maneuvers, betrayals, reversals, anything) and at some point he will be busy, distracted, and annoyed enough to start making mistakes. He will end up overdriving his predictive window, so to speak - trapping himself in situations where he can't see a way out, or even if he does see a way out, he might not be able to reach it in time with the resources he has because he spent those resources dodging previous plans.

worth noting: im reasonably sure this is off topic, but i'll give an answer anyway

you either: set a trap for this villain, if they know everything that happens in the next 24 hours it is reasonable to assume they can't process all of it and may miss something

or: you think about doing it, then rapidly act and assassinate them, probably using a sniper or a small explosive

you could just have the hero do something that the AI would not predict. like he could call his biggest enemy to help him kill the guy.

or you could take him to a different universe where there is no time. almost how Captain Britain defeated Mad Jim Jaspers.

I do not find any extraordinary ability in him, except that he has a profound ability to process data and coordinate it with other bits of data he has.

Hence secrecy is the key. Simply setting up a sniper ambush should be enough to finish him off, provided that the whole setup is pulled very quietly without leaking the least bit of information about what you (the ambushing party) are doing.

Or you could just plant a remote controlled explosive on his vehicle. Or poison his food. Or just ... whatever. Really, it is just a high security assassination case and nothing more. When you are up to assassinating some very high profile target (like the president of USA, China or Russia), you face the same problem: any information leak means your plan will be sabotaged and you will be arrested before you can even set everything up properly.

Provided that the protagonists do everything secretly, the villain is sure to fall.

Shoot him to death with a gun. When he sees himself getting shot. There is nothing he can do. If he sees himself getting shot to death, then that is exactly what will happen. That is the entire premise of the question. If he avoids getting shot, then that means his vision of the future was wrong, which goes against the whole premise to begin with. Just shoot him, and let him see himself getting shot.

EDIT: Just reread your post. You don't want him die. Hit him with Taser, cuff him, read his rights, and put him jail. Then follow due process.

• thats recursive though, because if he doesnt see himself getting shot... then he wont get shot – Alex Robinson Jan 31 '17 at 10:01