Normally, I'm not a big fan of deduction games. It reminds me of those old standardized test questions they used to ask (maybe they still do) to make sure you came from the right side of town: "Alex lives next door to Fred. Fred is friends with the girl in the red cloak. The person three doors down from the guy with the brown coat blah blah blah ... Who lives in each house on the block?". The key to answering these questions, it seemed to me, is not being able to reason, but having been told how to set up the grid into which you can plug the data that will allow the answer to fall out. Or that's what it took for me, anyway. A lot of deduction games seem to fall into this category for me: if I could just figure out where I could find the correct spreadsheet and note-taking technique on the internet, I could be as good as anyone else.
Of course, this simplifies the situation in many cases, possibly to ease the fact that I'm not very good at these games. Still, usually if there is a game in there, it's about picking the right question to ask, not the process of deduction itself. The few deduction games I like usually have something else going on. I like Coda, because there is a little bit of bluffing in with the deduction, and because the pain is over quickly. I also like Schwartzarbeit; since you can't take notes, the memory aspect is almost as important as the deduction, and you have to make some choices about what to try to remember and what to let slide. By contrast, I really disliked Code 777, which seemed to me a classic spreadsheet game designed simply to make your brain explode. Maybe entertaining as a puzzle, but not much of a game.
But now I can add a new game to the list of deduction games I like, Black Vienna.
The basic idea is there are 27 spies, A-Z plus Ö. These are distributed amongst the players, with the remainder left in a pile off to the side. You are, of course, trying to figure out who is in the leftover pile. The questions are a series of cards with three letters on them, and you have a couple of these cards available from which you must choose one to give to another player. That player then must tell you how many of those letters he or she has in their hand. If the answer is one or more, he or she marks the card with the corresponding number of chips, and the card is now fixed. Otherwise, it can subsequently be re-asked to another player.
It's still a game that is largely about effective note-taking, but there is enough flexibility and enough decision-making challenge in figuring out what the right question is and who to ask it that I enjoyed it. The deduction is also not straightforward, and it can take quite a few questions before you get even one solid piece of information. It's a once-or-twice-a-year type game; not a great classic, but pretty fun for what it is, and while not a simple game it's not so much of a brain-burner as to turn off people who aren't seriously into this sort of thing.
The only real flaw in the game is that it really cannot handle mistakes. If at any point someone says 1 when they meant 2, this can cause the game to explode catastrophically, as backtracking on all the implications of even one erroneous response can be difficult enough that it might be easier to toss the game and start again. I mention this only so that, if you ever play, you make sure that everyone at the table is clear that they need to be really, really sure when they respond.
In our game, I actually got wiped out by an error very late in the game ... it only took our player a minute or so to realize he had screwed up, but it was a critical question and by the time the error was discovered I had already made so many cascading deductions that backing out at that late date was more work than I wanted to put in. But at that point I had realized I wasn't going to win anyway, so I wasn't unduly disturbed, and still enjoyed the game. One I would definitely play again.
Black Vienna is long out of print, but you can find the rules here, and a template for all the required cards is available on BoardGameGeek.