Results follow the seat, not the name. Toggle someone off if they leave. Late arrivals join the next round.
Fake data, real tokens. Only reachable by typing the hash.
Same face, same size, one variable each. Which of these look cropped on the phone? That is the whole question — a, b, c and d isolate it to the fill, the filter, the transform, or none of them.
Two digits, not four, on purpose: d has no scaleY so its font-size is raised 1.25× to match the height, which makes a four-digit face 465px wide on a 390px screen. A variant clipped by the screen edge is the exact false answer this test exists to kill.
The suspect worth naming: WebKit rasterises background-clip:text at the pre-transform size, so a scaled element can shear at the edges in a way no geometric measurement catches. If a and c crop and b and d do not, that is the answer.
White, radius 22, no border, no shadow. Separation is surface tone.
One segment per round, wider finals segment. Olive done, blue current.
Initials on a fill picked from the name, deterministic. Six fills from the tokens; avatar colour is content, not chrome, and never counts against the one accent.
Header pattern: quiet label left, count or one action right. Never a caption.
Golden point at deuce, if it's on. Buzzer point when time runs out mid-game, the next rally decides it.
One name per line or commas. Numbering gets stripped. Order becomes seeding.
On the other phone: Settings, Claim keeper, enter this code. Valid 5 minutes.
Ask the keeper for the handoff code in Settings.
Every secondary flow lives in one of these. Springs from the bottom, never a modal, never a new page.
Results go to the season ladder and every player's record. Saving again replaces, never doubles.
Buzzer point. The next rally decides the game.
Tap to dismiss