I don't know if this warrants its own blog post. But just wanted to capture a quick idea for posterity.
There are two candidate versions of the encounters/challenges system that we've discussed as a replacement for the approach of v7.
The first was tried in v8: you go to a city, flip over an encounter card appropriate to that city, and it shows you the category (luck, wits, etc) and degree of difficulty of the challenge, as well as the reward you'll receive if you pass the challenge. To attempt to pass the challenge, you add up your stat in that category, any adventure cards you wish to contribute, and the result of a d6. The advantages of this system are that it manages the location-specific distribution of challenges automatically, and that it's very simple: every card has a clearly-specified reward and difficulty.
The second was tried in v9: the encounter card shows only a picture, and you choose a response (category and degree) based on what you think will "beat" the challenge -- eg "Luck 3" or "Wits 5", then slide the card into the Interrogator to see what the outcome is. The features of this approach include the possibility of different outcomes for each challenge, and that it feels thematically appropriate as well as "real" -- you're forced to guess at a "solution" rather than being told by the card how to beat it.
The point of this post is to say that there may be a couple of in-between options that have the simplicity and ease-to-prototype of the v8 approach while still retaining some of the features of v9.
First, I started a discussion on this subject at BGG, and one suggestion that emerged could be adapted in this way. Start with the v9 approach -- you are shown the scene, and you pick your response (category and degree) -- then, you slide the card into the Interrogator, and look up the /category/ you're attempting. And instead of an "outcome", it shows you the die that you roll to find the challenge's difficulty.
For example: You face an Arab Swordsman -- you decide to respond with Fight, and commit, through your stats and cards, to a "degree" of 5. You look up the result for "Fight", and it shows "d6". So you roll a d6, and the result is 3 -- that represents the degree of difficulty of the challenge. Since it's less than the 5 that you committed to the challenge, you pass. If instead you had chosen "Wits", the result may have been "d10", or "2d12", indicating that it's more difficult to fight a swordsman with your wits than with strength.
This approach is more like v8 in that the reward for passing the card would always be the same, but it would still have the possibility of different solutions like in v9. And part of the appeal of this is that you'd often be choosing between an obviously "better" solution (eg fighting the swordsman) and a solution that is probably harder, but that plays better to your strengths (eg debating with the swordsman).
Speaking of which, I wonder if we should contemplate reworking the challenge categories. For example, "fight" could be generalized to "use force" -- you can use force to beat up a Nazi thug, OR to try to push through a locked door.
A second way that some of the v9 concepts could seep into a v8 approach could be through the idea that different visual aspects of the card affect the challenge or the resolution of the card. For example, every player has a weakness/phobia, and if other players see your phobia on a card you're facing, they can announce this and force you to face the challenge at a -2 penalty, or whatever. This kind of thing could encourage players to be attentive during other players' turns, which might make the game feel more interactive. Maybe this scheme also fits in with Steve's idea of some encounters giving you an auto-clue -- it's there on the card, but you have to spot the trigger to be able to get it.