I had this idea that wouldn't leave. What if finding your NBA team felt less like a quiz and more like the screen itself was deciding? Not radio buttons and a results page. The viewport physically subdividing with each answer. Top half yes, bottom half no. Pick one, and your half becomes the next question while the rejected team's colors lock behind you.

Four questions. Sixteen teams. By the end, the screen is carved into strips of color — your team glowing at the center, the rejected teams stacked behind you like geological layers.

The first question. Celtics green fills the whole screen.

The mechanic is the point

It's a binary tree. That's all it is. But you're not looking at a tree diagram — you're inside it. The screen is the data structure. Every answer physically subdivides your viewport. The questions aren't scientific. "Do you think you're better than everyone?" sends you toward the Lakers. "Is BBQ a personality trait?" points you at the Spurs. "Do you like cheese?" lands you on the Bucks.

I wanted the humor to do the work. The mechanic creates the tension; the questions release it.

Three decisions that mattered

The early version split the screen 50/50 with each answer. Mathematically clean. Felt completely wrong. After four splits, your team occupied 6% of the screen — a tiny sliver. The experience felt like being squeezed into a corner instead of zeroing in on something.

I switched to 75/25. The chosen path takes three-quarters of the space. The rejected team gets a thin strip. Now your team grows as you commit. That single change transformed the whole feel.

Mid-quiz. The rejected team gets a thin strip. The certainty meter builds tension at the top.

Second: I started collapsing old rejections after two answers. Without this, the screen fills up with every path you didn't take. Removing old strips keeps the focus forward — just your current question and your most recent rejection visible.

Third: a certainty meter. Thin progress bar at the top. "16 TEAMS" at the start, "50% CERTAIN" halfway, "100% CERTAIN" at the end. Small thing. Builds real tension. The narrowing-down feels tangible in a way the splits alone don't quite sell.

The reveal

Tap your final team's section and it explodes full-screen. Team colors, huge name, a one-liner — "Showtime runs in your veins" — and confetti. Then a TRY AGAIN button to run it back and see where different answers take you.

The payoff. Full-screen team colors, confetti, and a tagline.

The whole thing is one HTML file. No framework, no build step, no dependencies. Bebas Neue loaded from Google Fonts. CSS and JS embedded inline. Open it in a browser and it works.

There's no business case for Decidr. I built it because the idea was interesting and I wanted to see if the screen-as-data-structure concept actually worked. It did. That's enough.