Status: I have an okay high score.
Most Intriguing Idea: Exponential scoring
Best Design Decision: The tile preview
Worst Design Decision: The scoring period.
Threes is a sharply-made puzzle game, where tiles marked with 1, 2 or some number of the form 3*2N come into a 4×4 playing field from a direction determined by the player’s previous move, and the game provides an indication as to what tile is coming. 1&2 tiles can be combined to make a 3; otherwise only like tiles can be combined. The game ends when the board is full and no further combinations can be made. At the end, 1&2 tiles give no points, while tiles with a 3*2N value give 3N points. This reflects the fact that in principle the effort involved in producing each higher tile is also exponential, i.e. it takes 1 combination step to make a 3 (1+2), 3 combination steps to make a 6 (1+2,1+2,3+3), 7 to make a 12 (1+2,1+2,1+2,1+2,3+3,3+3,6+6), and so on (obviously we are going 2(N+1)-1). In reality the number of combinations needed is lower because tiles of 3 and higher value stream onto the board frequently, but the scoring is broadly fair. There’s also some light music and dopey voices to help give the game a bit of personality.
As is standard for the genre the game is one where skill is stymied by randomization. Even an optimal strategy will sometimes not be resilient to getting a string of 5 2 tiles or the untimely arrival of a 96 tile at the top of a sea of 3’s and 6’s. The tile preview helps ward off the feel of being completely at the mercy of the RNG, and the low re-entry barrier makes it easy to lose a lot of time to this little app. The only hiccup here is the slow default speed of the scoring period and the weird idea of “signing” a high score (admittedly for families playing on an iPad this makes more sense). It’s not a revolution in any way, but Threes is a pretty good time-waster.