Cash-Up Without the Arguments
A tiny calculator I built so poker night ends with everyone getting paid correctly, not with someone double-checking the maths on their phone.
We play a regular home game. Everyone buys in for the same amount, gets the same starting stack, and at the end of the night the banker collects all the chips and pays everyone out based on what they finished with. The maths is simple: divide the buy-in by the starting chips to get a per-chip value, then multiply each player's final count by that number.
Simple, but at midnight after a few hours of poker, someone always ends up calculating it twice on the iOS calculator and someone else ends up recounting their chips. It takes longer than it should.
The Actual Problem
The problem is not the maths -- it is the flow. You want to enter the game setup once (starting chips and buy-in), see the per-chip multiplier lock in, and then just type in each player's chip count one at a time and read off what they're owed. One number in, one number out, no fiddling.
Every general-purpose calculator app forces you to re-enter the multiplier for each player, or store it somewhere, or do it in your head. And if someone miscounted their chips and needs to recount, you're starting over.
What I Built
It is a single screen. Two fields at the top: starting chips and buy-in. Below those, it shows the per-chip value in real time. Then a single field for final chip count, and a big gold result box showing the dollar amount.
That is the whole thing. No history, no multi-player tracking, no currency conversion. Just the number the banker needs to hand someone. The poker night version of the right tool for the job.
You can use it here.
Why Build It at All
There is a version of this argument where I should have just bookmarked a generic percentage calculator. But purpose-built tools are faster to use, and faster matters when you are trying to wrap up and get home. The setup takes five seconds, the per-player calculation is instant, and the dark green poker aesthetic is genuinely more fun than a white calculator UI.
The whole thing took about 20 minutes to build with Claude Code. That is probably the stronger argument for building it: the cost of a bespoke solution has dropped far enough that "just build the right thing" is often the correct call, even for a problem you solve once a fortnight.