The concept of niche protection is one that every design and GM has to think about in their tabletop roleplaying games. This is the concept that there is a number of specific functions or nifty abilities that work best when they are rationed out so no more than one character in a group has them at a time as they either can become overpowering, or have a finite limited use and become moot if everyone has them, or they loose their shine and niftiness if they are duplicated.
In games of fantasy or super powers, this is abilities like Flight, Teleportation, Clairvoyance or being able to go desolid and walk thru a wall. If everyone can do it then it's not nifty, or the GM will be planning foils for it so it's not as useful, or it can crash stories or lower their value and impact ("so we all teleport away when the counter reaches a digital readout of :02 if bob fails to defuse the bomb").
In other games with less 'power orientation' it's specific skills or skill sets. Thief Skills (Stealth, expertise at locks and security systems, pocket picking, etc.), Vehicle Skills, Martial Arts, Scientific Expertise, Medical Skills etc. If everyone is a thief and has the same skill set, or everyone is a chemist or everyone has some form of martial arts or everyone is a combat vehicle expert or everyone is a master surgeon then it devalues everyone in the process.
Some games try to fix this need by using a class based system, but there is little to stop everyone deciding to play a Thief or a Priest or whatever, unless the GM specifically limits the classes. And in Systems like D&D/D20 what happens is this becomes obscured by the collection of overlaps at the higher levels via feats and spells that duplicate what others do. (Thieves always had the problem that they could be outclassed by a mage with a wide enough number of spells to duplicate or even excell over what they could do).
There is also the problem that you need some overlap, because sometimes not all the players (and their characters) will make a game session. Or their character may be disabled for some sessions (I remember playing in many a game where the healer/doctor would get wounded in the first combat and the group would have to carry around their unconscious or immobile body thru the rest of the storyline).
Other games use things like complex race/tribe deliniation to develop uniqueness of characters, such as in the White Wolf games, and though these often achieve their goal, again they require some plot acrobatics to explain why specific characters would be gathered together (Why would a Corax, a Japanese Kisune, a Western Lupus, a Mage and and a Tremere Vampire ever cooperate as a group on a weekly basis?) Others put race in but, except for appearance and minor things like maximum physical attribute ranges or one or two minor abilities, they are just people in 'latex masks' (ex D&D Humans, Elves & Half-Elves; or the majority of the uplifted animal species that make up the world of Ironclaw and Jadeclaw).
My most recent design decision, used in Zamani and the onging 5th Age Worlds projects, is the Secured Trait.
In these games the Traits (favorable and unfavorable) are a larger part of the definition of the personality, past life and possible future personal storylines of a character then Just Attributes or Skills chosen. These Traits include species, childhood experiences, early formative training experience, dangling personal plot lines and stories, and unusual physical, mental, scoial or spiritual abilities or flaws.
The rules on Secured Traits is that only one character in a group can have each one. In design of the traits for a setting/world, I have to make sure that there are enough such secured traits that it is possible for each player to be able to take at least one of these in the design of their character, and thus ensure the chance to be unique.
This also means that the magic system for the setting/system has to support the uniqueness concept of the trait, so that it does not allow someone with magic to duplicate and be better or equal to a character with that trait. (And since magic itself requires specific trait, to ensure that no two mages in a group will be duplicating each other in knowledge or direction, so that all magic itself also remains unique in it's own way).
The flaw, of course, is that in end, if two players want to take the same secured trait, the GM has to make the final judgement of who gets it.
This could be decision by random selection (ex: rolling off for high die, cut for high card), or one based on who the GM feels could play the ability better (from past experience) or on how thw GM feels it meshes in with the rest of the character. Or it may be by arguing a trade system/barter system or bid system for secured traits (like they do for some abilities in the Amber RPG).
Finally, there is a loophole where the GM can over-ride the rule, if they truly feel that there is good reason for both characters to have the trait (ex: a pair of siblings, or a pair of characters in a Master/Apprentice relationship).
In playtest this seemed to work well in Zamani. No two characters had the same backstory, same childhood, same viewpoint on the world. No two mages perfectly matched in ability and knowledge (though we did end up with the start of an Apprentice/Master relationship between two characters near the very end of the test runs). Everyone seemed to think that the trait system gave them a really good feel for their character, and had provideed them with enough hooks to stir their imaginations for a character concept and possible personal stories for months of exploration of the character's lives.
Now I am interested in who else has approached this in design. I know I have a fairly hefty library of rolaplying material (I actually have more rpg books on my shelves then most rpg stores have at any one time), but probably have not seen every solution that's been taken out there. So if the designers, players and GMs have encountered this idea otherwise (I know about how Amber does this) or done in a more elegant manner, I'd be curious to know about where as I am always looking for ideas to mine from other games that can incorporate well into my designs and playing style.
The same with any GM related niche protection efforts you've encountered.... and well, any horror stories of failed niche protection that had crashed a game you were in or made it a lot less fun for you as a player/gm.
Curiousity... and way too many people not posting/commenting regularly over the summer makes me think of these things.