The trigger
{{d.trigger}}
The call
{{d.call}}
"I think you missed my turn."
Killer is made for a crowd: many people in one game at the same time, so it's purely social. Many pool halls run Killer Nights to bring people in on slow nights. Classic Killer runs on paper and playing cards.
- 01Each player is assigned a playing card
- 02Lives and turns are written on a board
- 03Each player flips the next player's card to pass the turn
Cards get forgotten, missed or lost. Turns get argued about. And one organizer carries all of it: the board, the cards, the pot, and every "wait, whose turn is it?"
What I changed
- No card flipping, and no forgotten, missed or lost cards. The phone knows whose turn it is.
- Nobody comes up saying "I think you missed my turn." Everyone can see the rounds and the pot rollover.
- The pressure comes off the one person organising: players find games, sign up and follow along on their own.
Group 1Serious playersThey come to play for money. They need exact lives, a clear turn order, an entry fee and a prize pot everyone can see.
Group 2NewcomersThey come to have fun. Draw cards like Extra Life, Shield and Steal a Life keep them in the game and make it social.
One game, set up for either crowd: lives, house rules, entry fee and draw cards change per group, so the same app runs a money night and a casual night.
Turn order · hiddenSerious players shouldn't know who's up next. If they did, they could team up against someone, and that's unfair when money is on the table.
Turn order · shownNewcomers in a big group lose track of whose turn is coming. Showing who's next helps them remember their turn and be at the table on time.
So it's a setting, not a rule: Public turn info lets the organizer hide or show who's next for each game.
I owned the product and design decisions. AI-assisted development brought it to life.
I initiated the product, defined its functionality, designed the user experience and interface, and made the key product decisions throughout development. I used AI-assisted development to build, test and iterate on the application, validating decisions in a real pool-hall environment.
I defined what to build, evaluated alternatives, reviewed AI-generated code changes and tested the application in real-world conditions. The code itself was generated by Claude Code, not written by hand. This let me iterate quickly while remaining responsible for the product's functionality and user experience.
- Spotted the problem in pool halls and defined the core functionality and rules
- Tested it with 47 players at the pool hall. No bugs were reported
- Reviewed and merged 9 AI-generated pull requests, one change at a time
- Designed the player, organizer, spectator and TV experiences
- Made the calls on turn visibility, late arrivals, game setup and draw cards
- Published the iOS app and ran every Apple, Firebase, Cloudflare and GitHub account step myself
The app, shot by shot
Build 39 · click a screen to enlarge
On the TV
The whole room sees whose turn it is.
The organizer casts the game to the pool hall TV with Chromecast. It shows who's up, their lives, the round, how many players are left and the prize pot. With Public turn info on, it also shows who's next. A QR code lets anyone follow on their own phone.
Chromecast receiver, 1920 × 1080, Public turn info on
Players don't depend on the organizer anymore.
[ Screenshot to add: {{x.title}} with a live game ]
Key decisionsWhat I said no to
Most fixes came from seeing it in the pool hall, not at a desk.
{{s.tag}}
{{s.text}}
Business model · planned, not shipped
I priced it by who has a reason to pay, not by which features exist.
Day 1 · replaced
One "Premium" plan, $4.99/mo
Four unrelated features bundled behind one price.
Next step before building the Venue plan: talk to five pool hall owners in Da Nang. Two or more yes answers, and it gets built.
{{n.value}}
{{n.label}}
Next
Back to all work →
Work with me