So, when a character or characters enter a scene, the referee often either has a particular challenge in mind for the protagonists, or sets one based on what the characters do as the scene unfolds. Part of this dynamic, but one that goes unremarked, is the idea that this is a sort of gambling on the part of the players: if you enter a scene with a character strong on social interaction but weak on physical combat, then you’re betting on the fact that the challenge in the scene will in fact be social rather than physical. Character generation starts this gamble by having the player choose (or in some systems, randomly determine) what will be a “payoff” for the character. The above character’s player, in essence, put most or all of their metaphorical chips on the challenges in the game relying on interpersonal interaction rather than fighting.
Many games prescribe that the referee tailor scenes to the character’s strengths: the general advice is that if your group of players creates a bunch of hardened soldiers, they want fights, so the referee is supposed to give them fights. Then again, those interested in what might be called (to appropriate a clumsy and slightly inaccurate term with lots of baggage) “simulation” could argue that one doesn’t always get to choose what challenges come up in the course of a day: sometimes the hardened fighter has to navigate a fraught social event, and indeed some very interesting and memorable narrative conflicts arise from tropes like “the fish out of water” or “compensating for a deficiency”.
Where these two intersect, the system for Project Atlantis seems to suggest an answer. I’ve talked previously about the ability to invest dice in the traits a player wants to emphasize in a scene, so that a character with a relevant social trait can charge up that trait with a die from their pool, bringing it to the forefront of play at the cost of some of their ability to take more flexible actions with the remaining die pool. I think this provides an intriguing solution to the dichotomy described above: a player has to choose what the character’s strengths are, but can choose from scene to scene whether to invest dice to bring those strengths to bear, or instead use the slightly less reliable die pool for whatever needs to be accomplished. That flexibility is tempting, too; players may want to never invest their dice for a known and fixed benefit if they can just roll at their full pool for everything.
To emphasize the tactical nature of the decision, I would make the act of investing dice at the beginning of a scene an actual bidding/betting process, with a tangible reward for those who opt to rely on their chosen traits and thus put both mechanical and narrative weight behind them. When the scene opens, characters invest dice in what they think will be necessary, with the referee secretly designating a “primary challenge” in the scene that serves as the source of experience (for lack of a more elegant term). Players then get a tantalizing choice: keeping their dice free improves their chances of short-term success in the scene itself, but investing dice risks inconvenience for a more lasting reward (if they, say, go into an aristocratic conference prepared for a swordfight to break out, then they are ill-footed up until the point that the hypothetical melee does indeed ignite, in which case they look extraordinarily prepared!).
This becomes more of a bet as traits increase in value: while I haven’t discussed this too deeply, keyword traits can have numerical ratings, which is to say they can be taken multiple times to improve their value and potency. (The Iconoclastic school of magic is the primary example of this, which replaces the standard magical elements with custom selections, but others exist: devoting multiple slots to being a Noble, for instance, should – and does – bring commensurate prestige and reward.) It’s tempting to make the reward for investment the ability to raise the relevant bid trait, but that’s too pat, and it doesn’t create an equilibrium between the level of risk and the prize at stake. A high-value trait will succeed more often, after all, and still only “costs” one die to activate, which means the possible detriment to the character doesn’t increase. Making the experience point reward equal to the value of the trait bid sounds okay at first, but that makes the experience progression of a character highly variable, because it increases the allure of bidding the highest trait in every single scene, but also asks that the character put all the metaphorical eggs in that one basket and miss out on experience more often than not.
A side-effect of bidding high traits, though, is that players will be encouraged to act in ways according with their traits, so the very social character attempts to socially manipulate situations, the master fighter will resort to drawing swords at the first chance, etc. in hopes that the powerful trait in which they’ve specialized will be the one that resolves the scene. Meanwhile, the referee is forced to create a “one true way” to resolve a scene, and removes player agency. In computer games, the term “pixelbitching” was coined to describe something similar, in which the tolerance level for getting the “right” answer to an onscreen dilemma is so rigid and unreasonable that the game becomes frustrating. This goes double if the referee has chosen a trait that none of the characters have, or have at a high level. Such decisions come across as punitive rather than empowering.
More thoughts on this later.