One Body, Three Personalities

The same articulated character as Floaty, Heavy and Precise — a controlled comparison, a real play-test, a revision, and a defended design choice.

Due: 21 May 2027 · Weight: 50%

The brief#

One character. Three personalities. You decide which one is right — and show your working.

This is the same shared jump engine every week has used, at full stretch. The Jump Lab starts you with Floaty, Heavy and Precise; you’re free to use them as given or as a starting point for your own tuning, as long as the final personalities are clearly distinguishable from each other by more than a slider label.

The comparison at the centre of this brief is deliberately narrow: change one parameter between your A and B trials, not several at once, and replay them with the Lab’s A/B tool so the difference is something a marker can watch rather than take your word for. A wide, unfocused comparison (“I changed gravity, air control and pose together”) makes it impossible to say which change caused what — that’s the entire reason week 10’s fair-test discipline exists.

The play-test is not optional and not simulated. Hand the Lab to someone else — a classmate, a housemate, anyone who hasn’t tuned these presets themselves — and watch them try the jump without coaching them through it. Write down what actually happened, even if it’s mundane. “They jumped the gap on the first try and stopped paying attention to the character” is a usable observation; a problem you expected but didn’t observe is not.

What you submit#

1. Preset comparison table — Floaty, Heavy, Precise, each with several real trials and their measurements.
2. One single-variable A/B comparison — a saved replay link, the one parameter you changed, and what you expected versus what the replay showed.
3. Play-test account — who played, what you asked them to do (or deliberately didn’t explain), and the specific problem they hit.
4. Revision — the setting you changed in response, and a before/after trial demonstrating it.
5. Final explanation (roughly 400–600 words) — the feel you were designing for, and why your data supports it. Distance jumped and whether the gap was cleared are not, by themselves, evidence of quality: a jump that clears every gap by pure air-control correction and a jump that clears none because it’s too heavy are both measurable, and neither is automatically “better” without you saying what you were optimising for.