Vapor sweeper and puzzle principles

Vapor Sweeper is chill version of minesweeper where you never have to guess, and there's no penalty for hints or undos. This is the best way to learn minesweeper or to relax with it if you're already a pro.

You can play now at 2weeks.games: Play Now

Motivations

I really like minesweeper. I think it has a really elegant puzzle design, but it’s a little bit hard to learn, especially because traditional minesweeper often requires guessing, and usually hint systems just reveal cells without explaining how you can figure it out on your own.

Foge, my co-founder, doesn’t like minesweeper. He bounced off of it when he was young and decided it wasn’t for him. My selfish goal was to make our version helpful enough that it could get folks over that initial hump and help people learn what’s so elegant about it. So far Foge says he still doesn’t like it, but he grudgingly admits that Vapor Sweeper has taught him some techniques and he’s getting better at it. I think he might get there.

If you’ve never played minesweeper, our version is a great way to learn. If you already like minesweeper, I think you’ll dig our version. If you already don’t like minesweeper, it might change your mind!

Working on Vapor Sweeper also helped me crystallize the principles we follow at 2weeks when designing puzzle games. Here’s the writeup from our internal wiki:

2weeks puzzle principles

Different game types are rewarding in different ways. Some games reward your reflexes or your time investment or your narrative choices. Great puzzle games reward your reasoning ability. A good puzzle experience makes you feel smart. A bad puzzle experience makes you feel dumb, bored, or both.

Players are most likely to have a bad puzzle experience right at the beginning of learning a new puzzle game. It’s easy to get scared off. For everyone that loves Solitaire or Minesweeper as a relaxing way to engage their brain, there’s someone who got demotivated because it was hard to internalize the rules or learn techniques and decided it simply wasn’t for them.

We want to build puzzle games that help people fall in love with puzzles. To support this, we follow these principles:

• Prevent failure states. If players get to a stuck position, give them the ability to undo their way to an unstuck position and try again. Avoid generating unsolvable puzzles or puzzles that require guessing.
• Hints should provide techniques, not answers. When a player is out of ideas and asks for a hint, don’t just provide a random valid move - the player learns little from that. Show the players the chain of logic that leads to that next move, and ideally let them perform the move themselves.
• Don’t disincentivize hint-taking. Give players the freedom to use their judgment about how many hints they take. Correctly-designed hints teach players techniques and help them build intuition. Don’t limit the number of hints or penalize their use in non-competitive modes (and always have non-competitive modes).
• Provide automation for the mechanical stuff. It’s not particularly interesting to note-take all the possible number placements on a Sudoku board, or to stack all of the cards on the foundation in Solitaire after they’ve all been revealed. If it’s a rote operation for the player and the computer can do it trivially, give players the option to have the computer take care of it.
• Only initiate automation when the player asks. Sometimes the mechanical stuff still feels rewarding, or even meditative, especially after overcoming a trickier puzzle state. Don’t trigger it automatically and risk robbing the player of the satisfaction of the action.