Small scope, real game: ten mechanics you can ship this weekend

The reason most weekend projects do not ship is not difficulty, it is that they were never weekend projects. A game with two systems has four interactions between them, and finding the two that are broken is what eats Sunday.

So: ten games with one verb each. All ten have been done before, which is the point. A known shape means you spend the weekend on feel instead of on discovery. Each one gets the twist that makes it worth playing and the part that will cost you more than you budgeted.

1. One-button runner

You hold to rise and release to fall. That is the whole input.

The twist worth adding: the level is generated from your last run, so the obstacles you barely cleared come back tighter.

What will take longer than you think: the gap tuning. One-button games live or die on whether the gaps are reachable at every speed, and the only way to know is to play it two hundred times.

2. Tile placement with a shared pool

Drag a tile onto a grid. Score for adjacency. The pool is shared, so what you take is what your opponent cannot have.

The twist: make the opponent a simple bot that always takes the highest-scoring tile. Predictable enemies are more fun than smart ones at this size.

What will cost you: the scoring feedback. Players need to see why a placement scored, and drawing that clearly is most of the work.

3. Word game with one daily seed

Everybody gets the same puzzle, the seed is the date, and the result is a shareable string of blocks.

The twist: the mechanic does not have to be guessing. Ordering, grouping, splitting and pathfinding are all under-used at this size.

What will cost you: the word list. A good curated list of a few thousand words is the difference between a fun daily and a frustrating one, and no generated list is good enough.

4. Physics parking

Drive a vehicle into a marked space using controls that are slightly wrong for the job.

The twist: take away a control everybody expects. No steering, only handbrake and momentum. No reverse.

What will cost you: the collision feel. Arcade physics that are satisfying rather than punishing is a tuning problem, not a code problem, and you cannot skip it.

5. Rhythm typing

Words fall, you type them, the tempo increases.

The twist: the words are drawn from something with meaning (a poem, a recipe, an error log), so typing them accumulates into a text you can read at the end.

What will cost you: input latency. Rhythm games are unforgiving about it and the browser adds some for free. Budget an evening on it.

6. Idle with one number

One resource, one upgrade path, one prestige. Ten minutes to an ending.

The twist: the prestige changes the rules, not the multiplier. Each reset alters what the number means.

What will cost you: the curve. Idle games are entirely a curve, and getting one that feels generous without collapsing takes a spreadsheet and several passes.

7. Local multiplayer on one keyboard

Two players, one device, two clusters of keys. No netcode, no lobbies, no matchmaking.

The twist: asymmetry. One player is faster, the other can build. It is more interesting than a mirror match and it is less balancing work than it sounds.

What will cost you: simultaneous key handling. Cheap keyboards drop simultaneous presses in ways you will not reproduce on your own machine.

8. Drawing game with a constraint

You draw, but you are only allowed a fixed number of strokes, or one continuous line, or a palette of three colours.

The twist: show other people's attempts at the same prompt afterwards. Comparison is the whole reward and it requires no accounts if you keep it to a local session.

What will cost you: stroke smoothing. Raw pointer input looks terrible, and the interpolation that fixes it is fiddlier than expected.

9. Deduction on a small grid

A five-by-five board with hidden information and three or four clue types. Minesweeper's family tree, not Minesweeper.

The twist: guarantee solvability without guessing. It is a generator constraint and it doubles how good the game feels.

What will cost you: exactly that generator. Write the solver first and generate boards by rejecting anything the solver cannot finish.

10. One-room escape

A single screen. Six objects. Four of them do something.

The twist: make one object change what another object does, so the room has a state rather than a checklist.

What will cost you: the hint economy. Players get stuck in places you cannot predict, and a game with no recovery from that is a game with a ten percent completion rate.

Picking one

If you have not shipped a game before, take four, six or ten. They are the three with no timing-critical code, which removes an entire category of bugs from your weekend.

Whichever you take, write the cut list before you write any code: no accounts, no settings menu, no second level type, no music, no tutorial. Then build the ugly version, play it on Saturday night, and spend Sunday on feel rather than on features. The version of this that ships is the version where Sunday was not for adding anything.