ALL THE STOPS ◆ gamification concepts
Four ways to make scoring more rewarding. All four are now built and live in both levels. This page is kept as the record of what each one is for, and why it is shaped the way it is — the panels below are the real components, not imitations of them. The routes are computed from the actual scorers.
Concept 1Score card feel pass
No new systems at all — the same card you already have, revealed instead of just appearing. Score counts up, bars fill in sequence rather than snapping, stars stamp in, and a near-miss line appears when you're close. That last one is the actual retry hook: "2 points from three stars" is much harder to walk away from than a bare 88.
Built. This is not a mockup — it is src/ui/scorecard.js,
the same component both levels render through, so this page and the game
cannot drift apart. Each star lights as the counter crosses its real
threshold (50 / 75 / 90), which is where the suspense comes from: you
watch 88 climb toward 90 and stall.
Concept 2Progress and personal bests
Best score and stars per level, kept in localStorage and shown on the level select. This is the backbone of a retry loop — without a remembered best, there is nothing to beat, and "retry to improve without being forced to" is the exact signal PLAN.md section 9 says to watch for.
Built — src/game/progress.js, shown on the landing page.
Note what it does not do: nothing is locked, nothing is gated.
Stars stay pride, not progression. Storage is injectable so it is testable
in Node, and a browser that refuses to write costs a player their history,
never the game.
Concept 3Route collection
The distinctive one. audit-levels already knows every settings
combination that three-stars a level — Dusk Rider has exactly six. So the
game can show you how many you've found, and list the ones you haven't as
blanks.
This matters because the obvious gamification works against you: rewarding "higher score" pushes players to hunt one best number, which quietly turns a trade-off space back into a quiz. Rewarding route discovery pushes the other way — it sends you looking for a second good answer unlike your first, which is the entire lesson of the level.
Concept 4Show me a better shot
Already in PLAN.md section 5 as "the ghost of better". After scoring, a button spins the dials to a three-star answer and re-renders the scene, so you watch the photograph change rather than reading about it.
The catch: it hands you an answer. Built after route collection it can reveal one route you haven't found and say so — "one of the six" — which teaches that there are others. Built before, it reads as the answer and undoes concept 3's work. That's why I'd sequence it last.
Built — src/ui/ghost.js. It picks a route you have
not found, prefers one at a different aperture from the shot you
just took, and does not tick it off: shown is not found. It stays
blank until you shoot it yourself.
OutcomeAll four, in that order
Built 1 and 2 first — the feel pass is the biggest change per hour of work, and personal bests give it something to aim at. Then 3, the one unique to this game, which needed persistence to be worth anything. Then 4 last, so it could reveal one of the six rather than the answer.
The risk named at the start still stands: all four make the score more prominent, and this is a teaching game. If playtesters start optimising the number instead of reading the tips, pull back. Concept 3 is the hedge, not a guarantee — and it is the first thing to watch in the six-person test.