Some background

The gods who have watched over us for so long are clearly angry. They've sent us this terrible trial in the form of a giant dungeon and told us if we can't complete it they'll wipe out our continent and start over (our intel says the other continents have gotten much the same message with their own trials, but hell if we'll help them--this'll be hard enough for us alone). We're told it won't open for another 18 years and for Young Adult fiction reason #57, the only people who can enter are those under the age of 20.

After many discussions and much analysis, we (the "government") have decided that we're limiting the number of kids we'll let enter to 1,000 and we've made it known that in 15 years we'll begin searching for those chosen 1,000 (this'll give us a year to search and 2 years to train them before the dungeon opens).

Needless to say the whole schooling system is getting a major overhaul--every school wants the prestige of teaching one of the chosen thousand. Not to mention birth rates are at an all time high. Everyone wants to be able to say that their kid was one of the few that helped save us all. Or they may be interested in the rewards we're offering those chosen. Who knows.

Thanks to some recent breakthroughs from the Mage's Guild, we're able to generate a spell to put about 1,000 people in a compressed time space where time will move at 1/10'th the rate as outside. Because of this, we're able to realistically test in groups of 1,000 kids for 8 hours with 20 proctors.

Our problem

But here's the problem: how on earth do we test these kids to narrow down from the almost million kids in our age range? We want those who have strong physical attributes, so that they can go toe-to-toe physically with whatever they find in the dungeon. We need them to be quick rational thinkers and problem solvers--who knows what kind of puzzles the gods have chosen to employ. And we also need those with a strong aptitude for magic1.

We've discussed the idea of holding standardized written tests with our own proctors, but with how corrupt the nobility can be, we imagine that they'll bribe a test taker in a neighboring town (or maybe one even further) to get the test early.

The facts broken down...

  • We need to narrow down from almost a million kids (ages 14-16) to 1,000
  • We have the ability to test groups of about 1,000 kids at a time for 8 hours with 20 proctors
  • We're judging children by the following criteria:
    • Physical prowess
    • Ability to think and act logically/rationally under pressure
    • "Magic ability" (or out of universe: programmer/math ability)
    • Ability to work well in a team is a plus (but we're willing to send in kids who prefer going solo)
  • The search will begin in 15 years, so we have that long to prepare
  • The location of the testing can be changed to whatever is needed--from a classroom to a coliseum. Nothing will be spared to ensure we find the best of the best.
  • We've thought long and hard, but we have no other magic that will practically3 help us test the children besides the time spaces.
  • We are aware that the nobles will try to bribe and cheat their way to getting their children chosen and want to actively combat that4
  • It's safe to assume none of our proctors or magicians can be bribed or corrupted.

How can we effectively find the top 1,000 potential mages on the continent between the ages of 14-16?

1. Out of universe explanation: magic is similar to a mix of our world's programming and math. There is a congenital limit to how much energy one can hold at a time, but especially at the higher levels magic is more limited by your fluidity and skill with wielding it. As such, you can assume those who would be good programmers in our world would make excellent mages.

2. For example, if my 5 mages erect a compressed time space for 2 hours real world time, it will feel like 10 hours to them and they'll need to rest for 4 hours before erecting another compressed time space.

3. Teleportation magic is far too costly, and our emotion reading magics are both expensive energy-wise and far too inaccurate.

4. Note: cheating from the kids is perfectly fine (and may even be encouraged). A clever kid who can find holes in rules is something we'd love. What we don't want is for the system to be cheated by those outside the test not during the test.

  • 43
    $\begingroup$ How willing are you to kill off an entire generation of children Battle Royale style? $\endgroup$
    – Separatrix
    Commented Oct 16, 2018 at 14:01
  • 2
    $\begingroup$ @scohe001 Unfortunately that is a really, really vague description of magical ability. Does one just execute genocide.exe to win a war? If logic and intellectual power imply strong magic ability that does not really define how that works and what kind of magic we're talking about. What is possible with that limited amount of energy? mountain-sized fireballs or tiny sparks? I think you need to elaborate more on the magic in your world for your question to be answerable. $\endgroup$ Commented Oct 16, 2018 at 14:17
  • 5
    $\begingroup$ You mean you’re making fantasy Ender’s Game? $\endgroup$
    – Telastyn
    Commented Oct 16, 2018 at 22:48
  • 2
    $\begingroup$ @Telastyn that...is actually startlingly on-point. I'm surprised I didn't realize that as I was writing the question. The kids are of course older, the trials they'll be put through completely different and most of the attributes I'm looking for are different, but there is a definite similarity. Good catch! $\endgroup$
    – scohe001
    Commented Oct 17, 2018 at 1:25
  • 2
    $\begingroup$ @scohe001 I was wondering where the incentive to send your kids into the dungeon is coming from. 18 years is a long way out and many of your rich and powerful families would probably much rather have their children live while they let someone else handle the problem. The world ending isn't enough of an incentive (global warming for example). I think you need an very large incentive one that is worth the risk so rich and powerful people don't purposely flunk. $\endgroup$
    – Shadowzee
    Commented Oct 17, 2018 at 5:01

9 Answers 9


Make dungeoneering a sport

Professional sports is an existing, effective model for finding the best of the best and making them even better. Start the National Mages League, which is a televised competition of dungeoneering skills, with a fixed schedule of matches, high pay for mage all-stars, and local teams with fan bases. 18 years is plenty of time to build a culture around the NML, with all young mages dreaming of making it to the big leagues. The best performers get to the highest levels in an organic, distributed way, without the need for intensive testing over a short period of time.

As a bonus, children who grow up aspiring to the NML will practice from a young age. There's no need to wait for 15 years before starting the selection or training process. It makes no sense to do nothing about this issue until 2 years before the deadline! Professional athletes pick up their sports at a young age, I imagine it's uncommon to start a sport at 16 and make it as a professional after only a couple of years.

Furthermore, the NML will build an upper eschelon of mages now, who can pass on their learnings for years to come. I expect that on average, modern sports teams would be able to beat sports teams from 100 years ago, due to improvements in training, diet, equipment, strategy, etc. You should start finding the best mages now, so that the next crop of mages are even better.

  • $\begingroup$ This is definitely a fun solution that would allow me to build a lot more awesome characters ("Johnny always looked up to Hector, his NML hero and the holder of the #1 spot for 3 years in a row now. He used to rush to get the paper the day after Hector would play to study how he'd cunningly won this time."). Not to mention this will help ease tensions between the generations just before those that are able to enter the dungeon by letting them play some part in saving everyone as well. That being said, televising may be difficult in the industrial age ;) $\endgroup$
    – scohe001
    Commented Oct 16, 2018 at 15:28
  • 1
    $\begingroup$ Reminds me of "Planet of the Damned" by Harry Harrison. Good read. $\endgroup$
    – Wildcard
    Commented Oct 17, 2018 at 1:33
  • 1
    $\begingroup$ would this require the building of dungeons? $\endgroup$
    – WendyG
    Commented Oct 17, 2018 at 11:06
  • 2
    $\begingroup$ The other nice thing that it does is that it gives you someplace to put the 999,000 increasingly-highly-combat-trained mages who /aren't/ going to the death dungeon. $\endgroup$
    – notovny
    Commented Oct 17, 2018 at 14:51
  • 1
    $\begingroup$ So basically set up battle highschools, PvP tournaments for ranks are common in that genre even when the goal of learning the combat skills is PvE. $\endgroup$ Commented Oct 17, 2018 at 16:21

You start now, not in 15 years. You set up schools that have only one thing in mind. After 3 years kids have first test, then second after another 3 years and so on. The kids with best results are send to better schools. You don't narrow it by much. 1 million to 1k in 5 stages is just 25% at each step. Each school level is harder in training those kids. And after 12 years when you have narrowed those few millions to 4 thousands you send them to super Uber BEst training camp there is.

You know, just like we do now with all those who are send to Olympics.

And rich people bribing their kids? You ever seen a rich parent bribing their kid way up to Olympics? And the kids will die in this dungeon. Why risk the possibility of losing heir when somebody else can do that job? Donnie didn't don bone spurs because they would look good on him. He did that to not die and have fun.

  • 14
    $\begingroup$ I imagine you'd get an interesting social dynamic where rich and powerful families will want their kids in the top 4000 for the prestige, but not the top 1000 because then they might die. $\endgroup$
    – Tacroy
    Commented Oct 16, 2018 at 21:23
  • 8
    $\begingroup$ I believe the word is "don," not "done." $\endgroup$
    – jpmc26
    Commented Oct 16, 2018 at 22:54
  • 5
    $\begingroup$ This might work for the Olympics but it falls apart when there are multiple metrics which need to be tested and combined. Think of it more like University than Olympics. You have your sports starts and your scholarly geniuses. You want both. The Olympics is a good example. You want the best athletes, but what happens if you just want the best overall athlete? Does a 100m sprinter out weight the marathon runner who beats the figure skater? That combined with finding intellectual geniuses and social people introduces flaws in a system that heavily weighs one attribute. $\endgroup$
    – Shadowzee
    Commented Oct 17, 2018 at 5:10
  • 2
    $\begingroup$ @Shadowzee The Olympics is just to show how we make selection on best candidates form early age. And it's clear we are looking for as many Dolphs Lundgrens as we can find. $\endgroup$ Commented Oct 17, 2018 at 7:38
  • 10
    $\begingroup$ @Shadowzee Why expect one person to be great at everything? Don't train up 1,000 dungeon-clearing jack-of-all-trades - train 200 5-unit squads where everyone has a primary and secondary proficiency that they're focussed on, even if they only have the minimum necessary skill elsewhere. Some people get in on the "combat" track, some on the "agility" track, yet others on the "magic" or "problem solving" tracks $\endgroup$ Commented Oct 17, 2018 at 14:01

There is a reason "best" is in quotes.

You continent sets in motion an elaborate system of training and vetting these young ubermages. 1000 of the best and brightest are groomed and ready to go when the dungeon opens.
4 hours before things are to start, one of the other continents (that you have pissed off over and over) attacks. Their dungeon opened before yours - things have not gone well and it is not looking good for their continent. There is war. The opening salvo is a nuclear strike on your dungeon - leaving it unscathed but wiping out your 1000, their handlers and nearly all of their supplies. Transportation is down and there is no prospect of getting the backup 500 to the site in time.

The "best" candidates are best because they are who is available. They are a random bunch of 8 kids from the area. Only 3 can do magic and only 1 well; she is 9 years old. One of the kids has Down's syndrome and one kid got burned in the nuclear strike. But these 8 are close enough to get in there before it closes. The surviving two handlers distribute the tech / weaponry that made it through the blast, wish the kids luck and send them in.

  • 7
    $\begingroup$ This is a good, if overused, plot hook, but not really a piece of worldbuilding. $\endgroup$
    – M i ech
    Commented Oct 17, 2018 at 10:28

This may sound a bit odd but... Crossword Puzzles?

This is actually how Bletchly Park, the british intelligence base that broke the enigma code looked for people with the correct sort of problem solving skills.

They put a load of Crosswords in the national papers with an advert saying soemthing similar to "Can you beat this in 5 minutes? if so write to us at [insert address] to hear about an exciting job offer" then once all those had written in they shipped them to a meeting and were sat down to do another one against the clock to identify the cheaters, and those that really were that good.


You could easily sort out some form of trial that tests the basic abilities you are looking for and then whittle them down with another test then another, these days this could all be done online and realtively easily linking from facebook!

  • 2
    $\begingroup$ Or (maybe) Cicada-3301 if you believe it's a recruitment tool for a Modern Day puzzle. $\endgroup$
    – hszmv
    Commented Oct 16, 2018 at 15:10
  • $\begingroup$ I know it's splitting hairs but the Enigma code was broken by the Poles before the outbreak of WWII and the information handed over to Britain immediately before the German invasion. This isn't to belittle the extraordinary work at Bletchly but to put it into context. $\endgroup$ Commented Oct 18, 2018 at 8:01
  • $\begingroup$ I completely agree without the Poles, it would have been exponentially more difficult, however the Germans also knew about the Poles and had made the enigma machine more complex as a result. $\endgroup$ Commented Oct 18, 2018 at 8:12

Stop looking for jack of all trades. Look for specialists.

You're sending one thousand people. That's practically a small army. You don't need a thousand people who are decent at everything, you need specialized squads that do their main job well, supporting each other and bringing everyone to greater heights.

You want frontliners. Ranged units. Support units. Logistics. Repairmen and craftsmen, medics and trapmasters and survivalists. All are key to surviving in a dungeon, and you'd do better to have your selection process focus on finding gems on each of these categories, and polishing them.

Of course, you still want everyone to have basic training in everything so they're not completely useless out of their one strength. But to train everyone in anything beyond the basics for all these categories would be a futile waste of time - that kid blessed with mana is better off getting taught by your nation's best mages than hitting the track field, and your best champions are better off focusing on training the kid talented in swordplay.

How do we train them?

First off, I believe SZCZERZO KŁY already did a fantastic job explaining this. You don't let kids fuck around and select them later, you train them from birth. The level of quality you can achieve with the latter method far eclipses the former, especially when you have the full force of a nation backing you.

You would do well to first gather your nation's best teachers. Gather your best champions for teaching combat, and your best trainers to improve the kids' bodies. Get the best teachers to sharpen their minds, archmages to nurture their magical capabilities, doctors to make sure they grow up healthy and strong. Then get various specialists to teach them in various skills. Priests (I assume they'll be the ones in charge of healing magic). Blacksmiths. Craftsmen. Locksmiths. Trap disarmers. Cartographers. Get all your nation's best and get them ready to raise a generation of masters, you need the best to raise the best.

Next, do notice that even in the same "category" (i.e. combat, or magic) there will be different branches of specializations. One could be a gifted conjurer, while the other is a healer or illusionist. One kid could be skilled with the sword, while the other never misses a mark with the bow.

As such, a standardized test might not be effective. No matter what, there will be talents that are less useful for that specific test, and some others that are especially useful. You'd be better off creating a really general scenario - several tests perhaps. I'd suggest something along these lines:

  • A test of combat: Various dummies scattered about, where you are allowed to do use any weapon or ability at your disposal to disable or dispose of them.
  • A test of knowledge: A test with open-ended answers, to determine one's tactical aptitude or the breadth or their knowledge. Magic tests could be included here too.
  • A test of miscellaneous skills: Their ability to survive, to mend wounds, to repair broken items, pathfinding, noticing traps, etc.

These tests should be graded subjectively by your masters, taking note of the standout performances.

Another point to consider is to test in groups rather than individuals. By letting it be team tests rather than solo, you also train them to work together, and get to discover which ones are the natural leaders.

What next?

Once you find out what each kid is good at, you focus on polishing their specialized skills, sending them to specialized schools or trainers you've chosen previously. Every few years all the children are gathered once more, mixed up and grouped to teams again for another round of testing.

This will, of course, take much more effort and time compared to a standardized test taken by everyone. But that's where SZCZERZO KŁY's answer comes in - When you train them from birth, you have plenty of time to perform these tests over regular intervals. You keep track of how each kid performs at each assessment, and in the end of the 18-year-period, have a trusted group judge each kid and choose the elite 1000 to represent your nation. I'd say this would be much better compared to one ultimate test that everyone has to take at the same time.


First test your system, not the people

As has already been mentioned, you start now, but the testing you're starting now isn't for magical ability. You're not looking to find the best possible magic users, you're looking to create them. Since your eventual protagonists possibly haven't yet been born or have only just been born you have some time to play with.

What you're testing you're testing right now is the output of the current education system and how effective it is at creating competent casters.

You're also testing your testing system, to make sure you're actually looking for the right basic skillset before the magic manifests.

Then upgrade the system

Take the best aspects of the most effective schools and spread those to the rest. Test again every year for effectiveness of the system. With 18 years to play with you should be able to create an idealised education system capable of putting out youth with the exact skill set you need.

The tag implies a highly stratified society with some highly wealthy getting the best education and the vast majority being poor, undernourished, and undereducated. Whether you want to work on the second group is up to you, but some generic trickle down to the basic education the poor get would be appropriate.

Finally test the people

How you open up the final trial to get the standard issue protagonists (one orphan male from underprivileged background who is a natural highly effective magic user and one overprivileged female who has had to fight everyone in her life (apart from her overindulgent father, mother died young) to get to that position) is up to you, but your entire society is now geared up to complete whatever challenge exists in that dungeon, so this final challenge should really just be a formality.


I think the best approach is to test only their Magic Ability. SZCZERZO KŁY Made an interesting point with the Olympics as a comparison, however you don't just need the best athletes. You need a Usain Bolt combined with a Steven Hawkins. A genius and athlete. The only issue is, its much easier to create an athlete than it is to create a genius. Sure you can teach someone maths and programming, but not everyone gets it... anyone can be forced to exercise and be fit.

So rather than testing multiple attributes, I propose you test and only train their Magic Ability.

Much like our school system, children need to be systematically taught the basics and magic and perform live tests (actually perform the magic feat, not just memorize it). Practical ability is more important that theoretical knowledge. Tests are based on a weighted average of execution speed, power and complexity.

Once a child reaches a certain standard, they are moved onto the next grade. Not based on years or time spent. They go to the next grade once they have achieved a bench mark score. There also needs to be a minimum for each attribute, because a super powerful spell is useless if it takes too long to cast. For example, little timmy here can produce a grade 1 fireball of 10 power in 1 minute. He can move into the next grade. Tom here can produce a grade 1 fireball of 20 power, in 10 minutes, but 10 minutes is too long, so Tom doesn't pass until he lowers the time. (Of course, you will have exceptions and geniouses, like maybe Little Timmy here can cast a grade 1 fireball with 1000 power in 1 minutes, but like all systems, those exceptions will need to be specially catered for and identified.)

This means that children will proceed based on merit. A 5 year old might be in the same class as an 18 year old if he is talented enough. Sure a rich parent can bribe a child through a couple grades, but at higher grades you publicize the test to the nation. That way after a certain grade, children are no longer able to progress since bribes won't help when the entire nation is watching.

So depending on your magic system, you create a grade system. At higher grades, they are taught more complex magics which put greater strain on their bodies and mental capabilities. At lower grades, they are taught simple magics. For example, a Fireball might be a tier 1 spell, and time pausing is tier 10. Once your children have reached the final tier, they start to help develop new magics or spend time improving their magic capabilities (sort of like graduating)

Once the time comes, you simply take the children in the highest grades and start to train their survival skills. Do it military style. You will get some drop outs, people who can't cope with the pressure or break under too much pressure, but you just drill them every single day for a 2-3 years and your going to end up with the cream of the crop.

  • 1
    $\begingroup$ I'd argue that there is a huge difference between someone who is "just" physically fit and a professional athlete. Strength and stamina are certainly prerequisites, but running a marathon every day will not allow you to beat, say, a world class martial artist, football player or rock climber. $\endgroup$ Commented Oct 17, 2018 at 15:01
  • $\begingroup$ @RutherRendommeleigh You would still be closer to a professional athlete than if you took an olympian and taught them how to code when compared to an innovator in the technology sector. Its simply easier to change someones body than it is to change their mind. $\endgroup$
    – Shadowzee
    Commented Oct 17, 2018 at 23:09

You need to make some key decisions about how much the world population & the trainee pool knows about what is coming, and just what these kids are being trained towards.

To my mind, the critical issue is whether you commit to having candidates trained "in the dark", with potential conflict around the trainees discovering the endgoal/deception, or take a character-exploration approach, where you can examine different candidate's reasoning for pursuing a grueling training leading into a probable suicide mission (or perhaps the glory that awaits victors).

While Ender's Game & Battle Royale (both mentioned earlier) come immediately to mind, anime has been exploring this "train kids to fight for the end of the world" genre for decades. The original Gundam TV series, and its post-modern rebuttal, Neon Genesis Evangelion, are both from a gritty realist "war is hell" perspective, while Mai-HiMe and its post-modern remake, Puella Magi Madoka Magika, take the idea of battle-school-within-a-school to the magical girl genre.

All are highly recommended, or you can just look up some analyses to get an idea of their respective approaches.


You want to foster creativity and different solutions, so you'll want create an environment where different strategies naturally arise. This means you'll want a decentralized approach; a government-sponsored test will tend towards a monoculture. So you want to isolate regions from each other enough that they will develop their own techniques, but you want some cross-breeding and feedback on the effectiveness of the technique. So maybe within each region there is a set of competitions, with the winners competing, kind of like in a Robot Wars kind of challenge. But you want to make sure they can handle the unexpected, so maybe you set up a secret, Second Foundation kind of school somewhere which develops independently. This would provide you with an unexpected underdog element if you wanted.

You also want passing the dungeon to be highly desirable, so success should result in money/respect/etc., whatever it is that people want. No one wants their kids to die, so it has to be good. It should probably also benefit the parents/family, since you'll need their help and encouragement.

On a larger scale, though, you need to address the issue of the gods. What do they want from society and why didn't society do it? If the gods are real, then you can bet your last gold piece that there are people cultivating their favor. If the gods are manipulatable, these guys will figure out how (humanity has been trying to manipulate the gods our whole history); if they are not maniuplatable at least some people will serve them for their character and will have some idea of what they want. Do the gods love humanity, hate humanity, or are indifferent? If the gods are indeed rulers, what kind of ruler doesn't make their desires known? And your schooling/testing situation doesn't address fixing the fundamental problem at all. You might pass the dungeon, but you still won't have changed how you relate to the gods. And if you change how you relate, maybe the gods will relent? For instance, when Ninevah repented after hearing Jonah, God relented, despite the city being the capital of a horrible empire. At the very least, one strategy for passing the dungeon would be the political/corruption strategy: some combatants are going to try to get in good with the gods and cheat.

  • $\begingroup$ Welcome to Worldbuilding.SE! We're glad you could join us! When you have a moment, please click here to learn more about our culture and take our tour. Thanks! $\endgroup$
    – JBH
    Commented Oct 17, 2018 at 22:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .