How to conduct organized in-world tests for attributes and stats?

[Peter Knutsen]

Any suggestions for this, particularly for a medieval world? Or a largely iron age-type culture in a medieval world, with a burning need to be able to sort people out?

The Magocrats of Bretagne, in my Ärth setting, test their own children, and kill those who don't show sufficient aptitude to learn magic (except a few get selected for involuntary vampirehood instead), and can afford to be fairly thorough with these tests.

But due to the way genetics work, if they just breed among themselves (with only the very rare liaison with a commoner) their numbers would be greatly reduced over the centuries. So they also subject relatively many commoners to similar tests, in order to find ones that can be recruited into the magocracy and indoctrinated to preserve the social order. Those commoners who fail these tests obviously aren't killed (although those who pass, but then resist indoctrination, are), so they're less nervous during testing than the magocrat children are, but due to not having infinite resources, the tests given to commoners tend to be much less thorough.

Or maybe there's a first-tier test for commoners, which is sloppy and very coarse-grained, and those who fail get to live, but then there are increasingly thorough 2nd, 3rd and 4th tier tests, and those who fail the 3rd or 4th tier, maybe the 2nd too, are killed because otherwise they might learn magic in secret and become a threat to the social order. Killing too many commoners could incite rebellion, though, which is why those who fail the 2nd tier tests may be more likely to be allowed to live, but covertly kept track of.

The main criteron for having an easy time learning magic is Psyche, mainly the Attribute for Arcane magical aptitude. In GURPS terms it's much like 10+Magery, although in Sagatafl it also affects creativity and intuition, and along with Perception is the basis for "Quasi-Psionics" such as Danger Sense and Empathy.

Intelligence and Will are also important for learning magic, but less so. Perception isn't important to become a magocrat, but some of the magocrat-children who fail on Psyche and Will may get to become Vampires if they score very high on Intelligence and Perception. So they test the children for Perception as well.

The tests don't need to be perfect, and in fact if they were then that would be bad worldbuilding. They also need to reflect the more primitive understanding of the Human mind that medieval and iron age people had. Positing that a bunch of severely paranoid 2nd or 3rd century ex-Druids can create something that is as good as 20th century psychometry would be ridiculous.

Commoner tests are given to people in their mid and late teens, and as stated above have to be somewhat fast and fairly easy to do, at least the 1st tier tests. Magocrat tests are given to children in their late preteens or early teens, and can be much more elaborate. Magocrat parents don't want their own children dying, and children of prestigious individuals can get a 2nd chance, or get a variant test as a sort of mild cheat or because it is accepted as fact that a particular trait, such as unusually high or unusually low Will, runs in a given (high-prestige) family.

But how to test? What do the tests look like?

I think that the Chinese, far to the east and ruled by an even more pranoid Lich Emperor (as suggestd in GURPS Who's Who), use tangram tiles as one of many bureaucracy entrance tests. That tests the Spatial Intelligence sub-Attribute (which, incidentally, also makes you better able to learn skills like Strategy and Tactics). The magocrats may have heard of that, even stolen the idea...

Thinking Speed can also be tested for. That goes back to Francis Galton of the Victorian period, but would probably make a lot of sense to medievals and iron age people as well. Quick Mind = Bright.

Will seems harder to test for. Frank Herbert posits the "gom jabbar" test, of intense but harmless pain.

Pre-literate peoples (and the Kelts, who came to disdain "Roman" things such as writing) probably taught all their children primitive mnemotechniques as a matter of routine, and the learning of that seems mostly based on discipline, thus Will. So that can be looked at as well, perhaps.

Thoughts and suggestions?

[johndallman]

Well, you need something that depends on Psyche, but I don't know what that might be that's easy to measure.

For an example test under GURPS, one can get a rough idea of IQ+Magery by presenting the subjects with many small sets of test objects and asking them to pick out the ones that are "different". The significant difference is the ones with an enchantment on them, which doesn't need to be hugely expensive, just enchanted. You'll rapidly learn who has no magery at all, and can get some idea of IQ+magery by the proportion of the enchanted objects that are picked out. Then you can go to actual detection magic, such as the Aura spell, to get a better idea.

Almost any world where magic is a regular and practical activity should have magic for detecting the magically talented, if such talent isn't universal. The fact that many rpg magic systems don't provide it is their failure.

[jeff_wilson]

Originally Posted by Peter Knutsen

Any suggestions for this, particularly for a medieval world? Or a largely iron age-type culture in a medieval world, with a burning need to be able to sort people out?

The sorts of tests you are proposing can't be done comprehensibly with the concepts available to such cultures, and I expect Bill will be along in a bit to go into further detail. If you want to have it, you can use the same hand wave that lets "medieval" worlds have modern coinage being used to hire dwarves to build traditional castles that are useless against the spells of the magicians and flying mounts of the knights whose loot is being taxed by the lord of said castle to gain the coins in the first place.

[Polydamas]

Originally Posted by jeff_wilson

The sorts of tests you are proposing can't be done comprehensibly with the concepts available to such cultures, and I expect Bill will be along in a bit to go into further detail. If you want to have it, you can use the same hand wave that lets "medieval" worlds have modern coinage being used to hire dwarves to build traditional castles that are useless against the spells of the magicians and flying mounts of the knights whose loot is being taxed by the lord of said castle to gain the coins in the first place.

Yes, measuring aptitude is very hard in real life (even before sociology interferes with the search for ideal candidates: privileged people tend to want to give their children privileges, and to assume that their successors should be people like them rather than people with some ineffable quality, while testers tend to chose easy metrics over effective ones and candidates try to game the system).

I have heard that the Israeli military is well satisfied by its system of aptitude tests for specific military jobs, but they have statistics, scientific psychology, and a lot of data from 70 years of mass conscription. So they would not be a good model for the 10th century world of Aerth.

The Chinese (and later Imperial British) solution for choosing officials was to make them study literature and picking the candidates who did best on a standardized test, on the principle that if they could learn languages and literature and write well they could learn whatever else they needed, and that literature was a morally beneficial thing to study. It worked reasonably well, but in GURPS terms it tested for Writing, Literature, and Language (Classical Chinese or Classical Greek) rather than Intelligence.

[Anthony]

A lot of it depends on just what traits they're trying to test for; breeding for particular traits, as long as those traits are identifiable, has plenty of history (I'd look at the history of horse breeding). The problem with breeding humans for particular traits is a combination of very long generations, human resistance to being bred in an organized manner, difficulty separating genetic influences from environmental influences, resistance to culling those with improper traits, and traits that aren't controlled by a simple set of genes.

On your specific example, it's really going to be dependent on the trait you're testing for. It's pretty easy to test for things like red hair or height, much harder to test for poorly defined traits like intelligence. Also, don't ignore the possibility that their testing doesn't actually work.

[Polydamas]

Originally Posted by Anthony

A lot of it depends on just what traits they're trying to test for; breeding for particular traits, as long as those traits are identifiable, has plenty of history (I'd look at the history of horse breeding). The problem with breeding humans for particular traits is a combination of very long generations, human resistance to being bred in an organized manner, difficulty separating genetic influences from environmental influences, resistance to culling those with improper traits, and traits that aren't controlled by a simple set of genes.

On your specific example, it's really going to be dependent on the trait you're testing for. It's pretty easy to test for things like red hair or height, much harder to test for poorly defined traits like intelligence. Also, don't ignore the possibility that their testing doesn't actually work.

Physical tests are historical, since I know of several low-tech societies which used them, but Peter seems mainly interested in mental traits. (The Romans had a height requirement for soldiers; some Chinese emperors made officer candidates perform with strong bows and heavy polearms; Thomas Lord Cochrane made volunteers march a certain number of miles on the double with a hundredweight of pewter). I don't think that anyone before the Victorians was optimistic enough to think that you can objectively measure mental traits.

I think that focusing on education is more plausible than focusing on selection. That is, all prospective mages would do things as children which were meant to make them excellent mages, with the failures being removed. Some mages would have the right to chose likely common children for an education amongst the children of the mages. I am not convinced that their judgement would be any worse than most “scientific” aptitude tests today.

[Anthony]

Originally Posted by Polydamas

Physical tests are historical, since I know of several low-tech societies which used them, but Peter seems mainly interested in mental traits.

Actually, he seems interested in aptitude for magic, which might be a spiritual trait, and may well be something that can be tested for with magic. In GURPS terms, testing for IQ is going to be very unreliable, testing for Magery might be quite reliable (in GURPS Magic, it's just "cast aura").

[whswhs]

Originally Posted by jeff_wilson

The sorts of tests you are proposing can't be done comprehensibly with the concepts available to such cultures, and I expect Bill will be along in a bit to go into further detail.

I've got nothing sophisticated to say. I just don't think it works in general.

Bill Stoddard

[LemmingLord]

"Detect Stats" Spell. To make it more story friendly, have it have the effect of "death vision" where the spell target gets to answer cool philosophical questions and such.

[Polydamas]

Originally Posted by Anthony

Actually, he seems interested in aptitude for magic, which might be a spiritual trait, and may well be something that can be tested for with magic. In GURPS terms, testing for IQ is going to be very unreliable, testing for Magery might be quite reliable (in GURPS Magic, it's just "cast aura").

He says that in Ärth, magic depends on four attributes, one of which mainly governs magic and three which mainly govern perception and mundane skills.

Of course, if magic can detect magical potential, it may be able to detect vague mental traits too! (Either LemmingLord's Detect Stats spell, or divining whether someone will be fit to join the mage-lords after their education).