Some new thoughts on dealmaking: We can't have a general-purpose dealmaking verb or function. Instead, each deal combination must be specified in advance by the storybuilder. This is made necessary by the tinkertoy text considerations: there's no way to write a general-purpose tinkertoy text script that will address a variety of deals.
So, we're stuck with individual deal-verbs. They will take the form OfferXForY or DemandXForY, with arguements placed in the secondary objects. There remains the issue of how a character/player will prepare such a deal. Any such scheme must operate within conventional procedures, otherwise the player would not be able to make deals.
The starting point of any deal must be a decision to seek something; the character must say, "I want that; how can I get it?" But what is this a reaction to? Our system is fundamentally reactive; how can a deal-offer be reactive?
We could have a deal as a response to a rejection: "OK, if you won't kiss me, what if I give you this bauble?" Pretty weak.
We could initiate a deal as a kind of division of labor approach. In other words, a character sets out to accomplish a task, but then digresses to ask if the task might not be better handled through a deal. Gal sets out to get revenge on Guy, but realizes that it would be easier to recruit Guy #2 to do the job for her. The problem is, such processing is a long digressive excursion. The gal must explore all the possible combinations of co-conspirators, what she could offer them, and their likelihood of accepting the deal, before she can make a decision as to pursue the deal. If the terms look poor, then she was to un-digress back to the earlier point in the decision tree. What a mess!
Look, we don't need general-purpose dealmaking. This is not simulation, this is drama. The only purpose for deals is to provide an opportunity for some joint skulduggery and for deal-betrayals as a means of increasing the dramatic potential. Perhaps, then, I should dispense with the notion of generality in dealmaking and think of deals in more narrow terms. That is, each deal is custom-crafted to a specific situation.
Here's an example: guy is trying to seduce gal, not too successfully. After one of her brush-offs, his options run from giving up to making another try to offering a deal. The deal has only one variable in it: what he offers her. This still raises an issue of look-ahead: should there be any lookahead? No, I think I have to dispense with lookahead. It's too complicated for the current system. Ouch. That means that the deal options had better be pretty limited.
But how then could any deal be made for some amount of money, if there is no lookahead? How is a price determined? Will we have to resort to haggling?
I think so. Without lookahead there's no other way.
Thus, a deal is always proposed as a way out of a conflict. The storybuilder sets up a conflict, and then provides a deal as a way out. The deal has two components: the conflict-resolving component, and the price component. That price component will then be haggled. So here's a model deal-making situation:
1. Eutrapelus: Let's do it my way.
2. Fabulla: No, let's do it my way.
3. Eutrapelus: Grr! Conflict!
4. Fabulla: Here's a way out: if you do it my way, I'll give you this.
5. Eutrapelus: That's not good enough.
6. Fabulla: OK, I'll sweeten the deal this much.
7. Eutrapelus: Agreed!
8. Eutrapelus: I am now doing it your way.
9. Fabulla: I am now giving you this.
Note that there's a skip in the seesaw at step 8. Note also that this arrangement skips over some of the possibilities. For example, steps 5 and 6 were optional; Eutrapelus could have accepted the deal immediately. Alternatively, he might have rejected the deal outright, taking them back to the earlier conflict. A third possibility: Eutrapelus could have refused to carry out his side of the bargain, in which case Fabulla would need to remind him, when are you going to carry out your side of the bargain? Finally, Fabulla could renege on her side of the deal after step 8.
There's still a problem with lookahead: how does Fabulla know what to offer Eutrapelus? If the storybuilder hardcodes the tree, then there will only be a couple of possibilities. Of course, they could be organized along lines of desire: "I'll give you 5 bucks; I'll give you kiss; I'll give you my bauble." Then Eutrapelus picks the avenue he likes best and perhaps raises the ante.
This involves a lot of technique (and work) on the part of the storybuilder, but I don't see much way that I can automate much of the process.
There's also the problem that this thing can get overly specific.
What if a deal were always a response to a demand? In other words, the interaction is organized around an initial demand from one character to another, and the dealmaker offers something in return for the previously-stated demand. Now the interaction looks like this:
1. Eutrapelus: Conflict
2. Fabulla: I want this from you.
3. Eutrapelus: This is what I want in return.
4. Fabulla: That's too much.
5. Eutrapelus: OK, I lower my price this much.
6. Fabulla: Agreed.
7. Eutrapelus: Here's my side of the deal.
8. Fabulla: Here's my side.
The idea here is that each character concentrates only on his side of the deal. This allows tinkertoy combination of deal-components. However, it also requires a means for evaluating the value of a deal-component. When in Event #2, Fabulla demands something from Eutrapelus, he needs to evaluate the cost of that demand before he can set his price. How to do so? My best approach so far is to have him anticipate the emotional cost of doing so, with some sort of lookahead virtual reaction. But this is not material covered in the adjust formulae.
I'm just spinning my wheels. It's time to call it quits for a while.