The question mark is in parentheses because I’m not sure if I want it in Delve.
One of the initial assumptions I latched onto and have maintained through the writing process has been the idea that the adventurers are each unique archetypes, fixed in their defined state. There is only one Swashbuckler, who has always been and always will be the Swashbuckler, with the specific combination of traits and specials listed in the back of the Adventurer’s Guide, and tweaking that at all makes the Swashbuckler into something else. (That seems to be shades of both Apocalypse World and Gauntlet.)
But a question occurs to me: the game gives methods and incentives for creating AP, and then spending them, but what about creating a tension in the economy by also instituting a competing incentive to hold onto AP rather than burning through them as quickly as they appear in the Pool? On the one hand, such an incentive already exists once players twig to the fact that their spent AP go right into the Hazard Pool to be used against them. But a hot streak on the dice and some reckless bravery on the part of players can undo that: hours spent carefully siphoning off AP to drop a big climactic dragon boss fight end sadly for the gamemaster who has to hand back all of those points as the monster sputters and falls to the adventurers’ onslaught.
Once again, we think about the trope of the adventurers in D&D-esque games making forays into the dungeon and periodically returning to town with their loot, “leveling up” to return to face the harder obstacles. This provides a useful conceptual model for a way to handle Adventure Points, though the exact implementation needs to be examined. Having adventurers spend their AP to raise traits would be tremendously broken, even with some kind of limiter (such as having to raise your lowest trait, and maintaining a cap of 6; a concerted player could have a stat block of straight 6s with just 15 AP, which would be easier and easier to obtain as the odds of success go up in direct proportion to the traits). Because the trope often involves liquidating the loot from the dungeon, we might also have to consider whether treasures can be converted back into AP, and the sudden influx to the point economy that would result. This does not seem to be a good idea on the surface, but given the proper framework, it might be allowable.
Buying specials seems at first to be an appropriate way to mark character advancement, provided that they cost a reasonable amount: off the top of my head, it seems that 10 AP would be sufficient, but that will require actual testing to verify. Because specials are a means of removing AP from characters, they shouldn’t be too expensive to obtain, but given their utility, they can’t be cheap, either, or adventurers will suddenly find themselves defeating enemies too easily and flooding the Pool with more points than they should have. It also threatens the aforementioned unique flavor that each adventurer ought to have: if the Berserker can just pay 10 AP and get Magic, the notion of defined character roles implodes.
Let’s look at what “level” implies: it’s a measure of general improvement in competence. A brainstorm occurs – adventurers may be able to buy up their “level” to use as a free-floating modifier to die rolls. This bonus could be split amongst multiple rolls, but doesn’t refresh until the party next breaks camp. That implies that adventurers start at “level 1,” and so have one point to add to a roll already from the beginning of the game, unless we want to arbitrarily define them as “level 0″ or the bonus as “level -1″, both of which are clunky. So we’ll keep it that way, and now adventurers have a new toy to play with. To keep it from being immediately abused, an adventurer can only gain one level at a time, and only by exiting the dungeon. (I always hated when characters suddenly level up in the middle of a dungeon.)