In parallel, we settled upon another big abstraction. Originally, the game contained a "clue table", and each card gave you a unique clue. The problem was, we never came up with a great solution for how a player could know what he was going to get a clue about -- if you know the temple is in Asia but really want to find someone who knows what city it's in, how do you get that information specifically? We experimented with a few things but never found a clean solution. In the end, I suggested a different approach that we eventually adopted: Each theme card has several symbols representing "subjects about which this card has some knowledge", eg "the temple location" or "the challenges you'll face in the temple". When you track down a given theme card, instead of that card giving you a specific piece of information, what it actually gives you is a "check mark" on your player notepad, in each category about which the card has knowledge. Then, when you go to look up a clue in that category, you count how many check marks you have (1, 2, or 3) and look at a clue of the equivalent level. So, for example, a level 1 clue might tell you "the lost temple is NOT in Europe", and a level 2 clue might tell you "the lost temple is NOT in Europe but IS in a major city". This is of course an abstraction, but it represents the idea that consulting multiple sources on a subject is going to provide you with better information than consulting a single source.
The last core ingredient of the design is the Nazis, or as we call them, "the enemy". These are represented in two systems; first, there were "enemy cubes" that were placed on the various cities, and that set the difficulty level for the challenge that you face when you arrive in that city. This worked well enough, but didn't really convey the idea that players are racing against the enemy to be the first one to find the grail. Originally, this didn't bother me too much -- although the game doesn't assign roles, you can think of the other players as your rivals, and the race between players seems to capture this well enough. The problem with this system is that if the "groupthink" is such that no one rushes into the temple, and all players linger outside the temple to get perfect information, then the temple phase falls somewhat flat, and the overall feeling of the game is not as tense and suspenseful as the theme calls for. So, we added an "enemy track" that abstractly represents how close the enemy is to finding the temple or the lost relic, and once it reaches 0, the game ends. What's more, the rate at which the marker on the track progresses accelerates as the game goes on, which gives a heightened sense that "time is running out."
With these ideas in hand, Steve built a lovely prototype and we began testing the game extensively at "Spielbany" (http://www.spielbany.com/)", our quarterly playtest gathering. Our friend Seth Jaffee was also kind enough to print the game and play it quite a few times with various groups. The feedback we received helped us tremendously to sharpen the game's focus, so that the excitement of being in an IJ movie could come through more clearly. During this process, one realization we made was that the clue categories ("where is the temple?", "what feature contains the grail?", etc) were each independent of the other categories, meaning that instead of a single clue/solution table for each adventure, we could create cards that would have clues on the front, solution elements on the back, and have a separate deck for each clue category. Each time you play the game, you shuffle each clue category's deck, take the top card, and begin playing. So we have a system that provides unlimited replayability, which we found very exciting.