On Letterpress and Solvable Games

Letterpress is one of my favorite new games. I’m not very good at it, but part of its appeal is the combination of word building and territory acquisition. I haven’t played it enough to feel like I’ve come close to mastering the game’s strategy, but play any game enough and you come close.

Noted game designer Raph Koster, creative director of the Star Wars Galaxies MMO and author of the book A Theory of Fun for Game Design, wrote on Twitter that his early enthusiasm for the game was rapidly fading since he’d found “too many winning strategies.”

According to Koster, following a four-step process will ensure victory almost every time:

1. Play for territory, rather than trying to simply get big words.
2. Expand outward from a single point.
3. Lock down vowels and extenders (such as “er,” “ed” and “ly”) early.
4. Go for the kill by taking out the last remaining gray tiles once you’re already winning.

Other players have reported similar success by following certain strategies, with some saying that they’ve “solved” the game. But even if these strategies work nearly every time, does that make Letterpress a “solved” game?

Solved games are ones which will result in a win or draw by playing a specific set of “perfect” or “pre-ordained” moves. Tic-Tac-Toe, for example, is a solved game. As is Checkers and Connect Four. Chess, on the other hand, will never be solved because of the nearly infinite combination of moves. Letterpress isn’t a solved game, either, because of the inherent randomness of the letters doled out on the board.

Letterpress is an impeccable game, in my opinion. It’s beautifully designed, remarkably simple to understand, intellectually stimulating, and doesn’t require a huge investment in time. I still prefer SpellTower, if only slightly, over Letterpress, but that’s because I can play SpellTower by myself. If Letterpress let you play a computer and play a friend from a single iOS device then it would be a perfect game.

Comments on this entry are closed.