I’ve been working on a thousand tasks arising from other parts of the project, but now that I have dealt with the most pressing issues, I am free to return to what is really the central design task of the game: the verb list. I approach this task with trepidation; it is the most difficult task in the entire project. I enjoy the programming work; it’s simple and straightforward. The occasional frustrations I experience with programming are minor inconveniences compared to the the immense complexities I face in designing the verb set. This is the Achilles Heel of the Storytron technology, yet it is also the fundamental challenge of any interactive storytelling technology. The verbs are the content. Getting them right is the crux of the design problem.
I had written up extensive notes on the next round of verbs, and now cannot find them. Just as well; it’s always good to start fresh. My task is to flesh out the verb list. Here it is as of now:
greet
what’s new?
no news
gossip
I see
request deal
what is your goal?
agree
reject
tell auragon
threaten
submit
defy
dream-attack
defend
There are a number of other verbs that Fate uses to govern the activity, but these are all the verbs available to an actor. My task is to increase the amount of interactive opportunity to the actors. These are some of the verbs I have been contemplating:
Beads
I have set up Siboot-beads as a means for increasing an actor’s perception of other actors’ truthfulness, but there’s an ugly gotcha: the difference between what the player perceives and what other actors perceive. For the other actors, we can simply program in the change in their trust, but for the player, we are relying on their ability to perceive changes in the animation of the face. For some players, beads will be of enormous value; for others, not so much. How can a player reasonably evaluate the value of beads under such circumstances?
There’s another problem of lesser import: beads lose value as the game progresses. A bead early in the game will be useful for many interactions, but will lose value as the game progresses and fewer interactions lie in the future.
Thus, I have no way to properly assess the value of beads. I suppose, however, that I can rely on playtesting to establish that value. The way to do this, I believe, will be to assign each bead a small increment of perceptiveness and vary that value during testing.
How many beads should be in play? Hmm… How about everybody starting off with three beads, a gift from Siboot IV?
Should it be possible to talk about bead transfers? Does it help somebody to know that one person has more beads and another person has fewer beads? I suppose so, but since there are only two people who originally know about a bead transfer, why would either one of them want to give away either the information that they have fewer beads or that they have more beads? No, I don’t see it making any sense.
Promise not to attack
This is a barter option, just as with beads. One actor promises not to attack another. I wonder, should this be done symmetrically? That is, do players agree to a mutual no-attack policy, or can a promise not to attack be traded for information? If the latter, can such a deal ever be balanced? It seems to me that a promise not to attack is worth a lot more than a piece of information, especially given the uncertainties in all information. This argues for this to be a symmetric arrangement.
Again, I don’t see this as a subject for gossip; neither party could benefit by releasing this information. However, it would definitely be important if somebody reneged on the deal. So I need a verb expressing the idea that somebody promised not to attack and then broke the deal.
Unite against a third party
This would require both players to attack a third party. The pact lasts for only a single night. It also assumes that the two parties will share all they know about the third party. This is strong stuff, probably best reserved for late in the game when one actor gets close to victory.
This would definitely benefit from reports of violations of the deal.
Algorithms for these verbs
There are two levels at which the algorithms for these verbs could operate. At the simpler level, there is no deception: each party evaluates such a deal solely on its direct merits. This, however, would deprive the game of its all-important challenge of understanding the motivations of the other actors.
The deeper level would require the actors to engage in deliberate deception. There’s no deception involved in bead trades, but the latter two verbs could be used deceptively with great effect. A person could promise not to attack and then attack; a person could unite against a third party and then tell them of that fact, suggesting a reverse uniting against the victim of the deception. This could get extremely complicated; do I really want to dig into algorithms of such complexity?
I shall begin with the bead trades, then the promise not to attack, and lastly the uniting verb.