The note claims one game should be one message. This chapter is the whole mechanism: the live card the group sees, four player screens, three admin screens, and the per game math. Screens are sketches of what PlaySpaces could do, not shipped software.
WhatsApp is the campfire.
It should never have been the database.
The community, the banter, the "Hennur El Clasico", that stays in the group. It's why this thing works, and it's the acquisition engine. What moves out is state: who's in, who paid, who's waiting. One game becomes one message with a live card, the link stops being a checkout page and becomes the roster itself.
I sketched the four screens a player would touch. Each one exists to make one failure from chapter 1 impossible, and the verbatim quote it retires sits under it.
A name appears the moment its payment clears, atomically, for everyone. There is no version of this list to copy, so there is no version to copy wrong. A held spot shows who is mid payment, so nobody takes it by accident.
Half this group arrives in squads, so booking should not pretend otherwise. Friends come from your squad list or as named guests, and one UPI payment locks every spot together. No more sharing the link and phoning three people to make sure they each checked out in time.
You fronted ₹540, the app sends each friend a UPI collect request and quietly reminds them until it's settled. The awkward "bro pay me back" call becomes a status chip. Guests who keep playing get nudged to make an account after their second game, that's the acquisition loop, automated.
Dropouts happen, rain, traffic, moms. Now a dropout is a button, not an apology: the leaver gets an instant refund, the first person in the queue gets a timed offer, and the game stays whole without a single message in the group.
Nothing here removes the admin. It removes the clerk work: chasing, auditing, reconciling, refunding. What stays is the part that actually needs a person: when to add a game, when to hold one, who gets the benefit of the doubt.
The Wednesday 9v9 has run nine times, so it's a template, not a retype. Format, turf, price, and the safety rule carry over; publish drops exactly one message into the group. Recurring games post themselves every week until told otherwise.
Everything Trisha currently keeps in her head, on one screen: paid, held, queued, revenue. The robot does the chasing, nudges at T-6h, waitlist offers, the auto-cancel rule armed in the corner. The only humans surfaced are the two who actually need a decision. And because a spot is a payment, "who paid without adding their name" is a question that can no longer be asked.
Every player has a position and a quiet form rating from past games. One tap drafts balanced bibs/non-bibs; drag to override, squads who want to play together stay together, and the sheet publishes to the game page at a set time. "We played 2 games but both were unbalanced" stops being a post-game autopsy.
Left column measured from the export, 20 May – 30 July. Right column is what the same ten weeks look like when state lives in the product.
| Per game | Today, measured | With the live card |
|---|---|---|
| Roster pastessame 22-line message, re-sent | ~19 1,755 across 90 games |
0 one live card, posted once |
| Admin messageschasing, auditing, reposting | ~20 1,759 total · 37% of the whole chat |
~2 the post + the human calls |
| Payment reconciliationticks vs. backend | Manual "let me figure out who paid without adding their name" |
None a spot is a payment, desync can't exist |
| Group bookingsyou + friends | N calls one payment per phone number |
1 tap squad checkout + split & collect |
| Late dropout8:27 PM, raining | A hole apology in chat, spot dies |
~15 min release → refund → waitlist №1 offered |
Roughly 1 in 5 games currently wobbles, downgraded or cancelled late, with discounts owed and refunds processed by hand. Timed holds, auto-waitlist, and the safety rule convert most wobbles into either a locked-in game or a clean, automatic refund at 5:30, not a 7:25 PM apology. At ~₹3,200 a game, every rescued night is real money, and every rescued night protects the streak that keeps this group alive.
This is the sharper point: the current model needs one heroic Trisha per ~9 games a week. That's the scaling ceiling of the whole platform. Move the clerk-work into the product and one community lead runs several venues, new city, new turf, same playbook. The WhatsApp group remains the front door; guest players become accounts after their second game, which turns every squad booking into acquisition.
No migration, no announcement, no risk. The group keeps its chat. Three numbers get measured and they do the arguing.
The Wednesday 9v9 gets posted as a live card instead of a template. Everything else about the group stays untouched. If it flops, delete the message, nothing else changes.
Every game runs on cards. Squad checkout and the waitlist go live. Trisha keeps full override on everything, the robot proposes, she disposes.
If pastes didn't hit zero, if filling got slower, or if the group's banter thinned out, we stop, and the chat goes back to exactly what it was. The export says it won't.