Where This Came From

It started as a math tangent, not a game idea. I was thinking about a chain of four connected segments — lengths 1, 2, 3, 4 — each pointing up, down, left, or right, anchored at the center of a grid. The question was simple: which coordinates can the far end actually reach?

The answer turned out to be a clean, almost elegant shape: a Manhattan-distance diamond of radius 10 (since 1+2+3+4=10), filled in a checkerboard pattern when all four segments are used, and nearly fully filled — with exactly 8 missing lattice points — once you allow skipping any segment.

That reachability puzzle turned into a movement vocabulary. A knight's move isn't a special piece anymore, it's just "2 perpendicular to 1." A rook-like slide is "3+4 in the same direction." Chess movement rules become emergent expressions of a small vector language instead of hardcoded per-piece behavior.

So I built a 2-hour POC: Vector Harvest: 1·2·3·4. Four movement links, one route per turn, pickups scattered on a 20×20 board.

The Obvious Failure Mode

Give a player four movement vectors and a board full of point values, and the naive design writes itself: harvest everything you can reach, forever, and the score goes up forever. That's not a game, that's a treadmill.

The interesting design question wasn't "how do you move" — the vector chain answered that. It was "what stops a player from just fetching points forever?"

The Answer: Continuous Taxation, Not Scarcity

The obvious fixes are the boring ones: cap the number of pickups, make them run out, add a timer. Vector Harvest does something different — it taxes movement itself, continuously, regardless of whether you collect anything.

  • Every cell you enter costs tax, whether or not you harvest there.
  • Passing through a cell costs half its displayed tax. Ending a segment there costs the full amount.
  • Tax scales with the round number: 0.0025% × round, with visible 2× and 5× outlier cells.
  • At the end of each round, total capital — not just that round's harvest — gets taxed again: closing capital = floor((opening capital + harvest) × (1 − wealth tax rate)).

This means bigger, farther-reaching routes aren't automatically better. A length-4 segment pays half tax on its first three cells and full tax on the fourth — so reach has a built-in cost curve, not a flat one. And because the entire capital pool is taxed at round end, simply accumulating points is actively penalized. Hoarding is the losing strategy, not the winning one.

Why This Matters More Than It Looks

The naive version of this game would let skill translate directly into an unbounded score. The tax system breaks that assumption on purpose. Instead of "more efficient route = strictly higher score," the system asks: is this specific route worth what it costs to run?

That's a different kind of judgment than pure optimization. It's closer to portfolio thinking than speed-running: some routes are locally efficient but not worth the compounding tax exposure. The playtest numbers make the shape of this visible:

  • First vanilla run: 66
  • Practiced vanilla run: 99
  • Search-assisted vanilla run: 157
  • First full scaling campaign (with upgrade choices): 11,781

The jump from ~157 to ~11,781 isn't from moving better. It comes from the Treasury mechanic: cumulative tax paid converts into a small public benefit (extended pickup lifetime), and round-end upgrade offers (yield, market density, timed lifetime, tax shelters) let you actually spend into reducing future tax exposure. The system doesn't remove the tax pressure — it gives you legitimate ways to negotiate with it.

The Design Lesson

A system that only rewards "collect everything you can reach" doesn't need a player, it needs a script. What makes Vector Harvest a game instead of a treadmill is that the cost structure competes directly with the reward structure, at every single move, not just at some end-of-level checkpoint.

The tax isn't a penalty bolted onto the fun part. It is the game. The four-vector movement language gave the project its shape; the continuous taxation gave it a reason to stop being a fetch quest.

POC Boundary

This build deliberately excludes enemies, inventory, tax-reset pickups, and narrative progression. It tests one thing: whether movement-driven harvesting stays interesting when endpoint tax terrain and player-chosen scaling compete against continuous wealth taxation. That's the whole experiment.

Play it: b1c3game.github.io/vector-harvest