… and by that, I mean that they should be inspiring good, old-fashioned awe with the things that they do.
This is something I first talked about in D&D: Calibrating Your Expectations and I developed the theme in E(X): The Many Games Inside the World’s Most Popular Roleplaying Game.
Recently I was speaking with someone who was unhappy with the “crazy scaling” of Perception checks he was seeing in the arms race of high level Pathfinder games: The players crank up their Perception modifiers. The GM responds by simply cranking up the DC necessary to accomplish previously much easier tasks. The result is “silly” and “ridiculous”.
This is what I said:
FIRST, CONTEXTUALIZE THE NUMBERS. Instead of just blindly cranking up the numbers, think about what those larger numbers really mean. If a DC 30 check reveals a “well-hidden secret door”, then what does a DC 40 check really mean? Well, it means something that no one in the real world without special tools would ever be able to detect. So maybe that means that the door has been phase-shifted onto the Ethereal Plane; or painted with the illusion-infused blood of a demon; or covered with the alchemically-treated hide of an animal that has evolved the ability to make people psychically ignore its presence.
In other words, embrace the fact that the PCs are doing awesome things and really emphasize how awesome it is.
SECOND, EMPHASIZE NOT CHANGING THE NUMBERS. Instead of trying to keep the same old tasks challenging, focus on the paradigm shift that’s occurred.
Yup. They’re really, really good at finding hidden things. Similarly, they’re really, really good at killing 1st level goblins. Instead of resisting that change by leveling up all the goblins in the universe to match their new abilities, focus instead on exploring how their interaction with the world shifts.
A mechanical option along these same lines would be to include guidelines for improving the quality or speed of a check by accepting a penalty on the check. For example, I have a generic set of guidelines for “quick checks” that drop the time required for the check by one category in exchange for accepting a -10 penalty to the check. (High level characters, for example, can make successful Gather Information checks in 1d4+1 minutes instead of 1d4+1 hours by accepting a -20 penalty on their check.) For Perception checks, you might apply a -10 penalty to allow characters to notice hidden doors/objects/etc. without actively searching for them. (They just walk into a room and spot the hidden door in the corner.)
An extreme example of this sort of thing Doctor Who: The Doctor can open the door of the TARDIS, sniff the air, and instantly identify the atmospheric content, the planet he’s standing on, and the time period. (I like to imagine that this is based on complex spectrographic analysis compared to charts which Time Lords study in school much like we study multiplication tables.)
ALTERNATIVELY, PUT A CAP ON THE AWESOME. If you don’t want to embrace the awesome, on the other hand, you’ll be much happier simply stepping out of the arms race. Cap their advancement before they become “too awesome”, either drawing the campaign to a close or finding other ways of advancing their characters. (This is where E(X): The Many Games Inside the World’s Most Popular Roleplaying Game comes back into the picture.)