Turing Me Up

Clue is one of the few "board games people have heard of" that I think is actually not all bad. (Though like Monopoly, it's a game probably not often played according to the actual rules.) I played Clue a lot growing up, and that's probably part of why I enjoy deduction games so much today. Even though it's is better than, say, Monopoly or Risk, there's still plenty of room for improvement -- and plenty of board games have done just that.

Unfortunately for me, the enthusiasm for deduction games within my regular play group is considerably lower than my own. So it isn't often that I get to play them. Like Sam Beckett hoping that his next leap will be the leap home, I always hope that the next deduction game I try might be the one the rest of my group enjoys. Or tolerates enough to play every now and then.

Deep down, I suspected that Turing Machine wasn't going to be that game. But I had to buy it anyway, because it just seemed so damn clever and cool. At core, it's both quite simple and quite similar to other deduction games that have come before. In each game, there's a mystery 3-digit code. Each digit is a number from 1 to 5. Up to four players are competing to find the code first.

It's how you go about it that captured my imagination. During game setup, at the center of the table, you create an "analog computer," made up of 4 to 6 "verifiers" that each know one key piece of information about the code. Verifier A, for example, can tell you whether the blue first digit is higher, lower, or equal to the last purple digit. Verifier B might know how many instances of the number 3 are in the code: zero, one, two, or three. Verifier C might tell you whether the sum of all digits in the code is an even or odd number.

In each round, each player secretly forms a theory about what the 3-digit code might be, and is then able to test that hypothesis with up to three of the verifiers. Say I test the code 3-3-2 against the three example verifiers above. Verifier A tells me "False," so I know that because I picked a first digit higher than the last digit in my proposal, the first digit is not higher than the last. Verifier B also says "False," so I know there are not two 3s in the code. Verifier C tells me "True"; I picked digits that add up to 8, so I know that whatever the code is, the numbers in it do add up to an even number.

How you conduct these tests is the piece of the game that simply amazes me. The game comes with a thick stack of numbered square cards with seemingly random grids of checkmarks and X marks. Setup tells you exactly which cards to place with exactly which verifiers (there's also a stack of dozens and dozens of those). To propose a code, you grab three punch cards representing your three numbers, each with different patterns of notches cut into them, just like old-fashioned computer data entry punch cards. When you line any three of the punch cards up together, they'll leave exactly one small square exposed... and when you line that up against a verifier's grid, it will expose exactly one checkmark or X on that card, giving you your true or false answer.

I cannot even begin to wrap my head around how these mechanics were actually achieved. Game designers Fabien Gridel and Yoann Levet somehow devised a system where all the info on all these punch cards returns exactly the information to solve exactly the logic puzzle put in front of players for each game. And it's extendable; the game comes with a handful of puzzles listed in the rulebook, but it also provides a website link where literally thousand of puzzles can be served to you at random!

I was, frankly, a little worried that my interest as a designer in this technical feat was clouding my expectations of the game itself. It was likely this would be more clever as a design than it would be good as an actual game, right? But it would be worth owning even if just to pore over it and satisfy my curiosity for how it was made, right?

I have played Turing Machine a handful of times now, and was pleased to discover the game is more than its cool mechanisms -- it was fun to play too. It's incredibly fast-paced; the box's claim of a 20-minute play time is indeed correct. (Not counting any explanation, of course. Which does take a bit of time; understanding how verifiers work can be tricky.)

Deduction games can sometimes be a bit too driven by luck, when a player coincidentally asks just the right question early on. But 1) Turing Machine actually offers a mix of puzzles more dependent on luck with others requiring more skill; and 2) you need luck to be a fairly healthy element in the game, or you will never get deduction game haters to play with deduction game lovers.

And happily -- the deduction game haters seemed ok with it! The alchemy of this game's speed, the way to approach the puzzle, and the information constraints seems to have done the trick. The simultaneous play surely helps too -- because you're asking the verifiers for information and not the other players, multiple players can all formulate their guesses at the same time.

This probably won't be a regular new game in my group. I don't want to push my luck and burn everyone out on a deduction game they actually like. (Or tolerate, if they're just humoring me.) But in light of that development, I no longer want to dig deep into the layouts of all those grids and punch cards to see if I can figure out how the designers actually made it work. I'll just enjoy the marvel.

If I'm rating Turing Machine purely by how much I've enjoyed playing it so far, I'd probably give it a B+. But the fact that I will get to play it more (unlike other deduction games), combined with just how damn clever it is, makes me want to boost that mark. So I'll call Turing Machine an A-. I'm glad to have it in my collection.