# Felted private-table walkthrough: method and rehearsal

Checked September 7, 2026. This package accompanies [Poker With Friends in Felted](https://gtogecko.com/blog/felted-private-table).

This is a first-party, source-reviewed setup guide for ordinary private Play with Friends tables, associated with Felted 1.7.0. It is not a simulator, a tournament scheduler or a live service test. GTO Solutions AS publishes Felted and GTO Gecko.

## What was reviewed

Current official US [Apple](https://apps.apple.com/us/app/felted-texas-holdem-poker/id6759221162) and [Google](https://play.google.com/store/apps/details?id=com.felted.felted&hl=en_US) listings corroborate two-to-six-player private tables, six-character codes, separate table chips, host-selected settings and the Rail. [Official support](https://gtogecko.com/felted/support) establishes internet requirements and support-report guidance.

Authorized first-party implementation review established the English controls, initial settings, automatic submission on the sixth code character, host start conditions and creation-based table limit. Local widget tests used simulated table services. They check particular UI paths; they do not establish that real phones can connect to a deployed server. Source code, synthetic test credentials and private implementation snapshots are not distributed in this public package.

The initial setup/join/waiting-room run passed 15 tests and failed one seat-slider accessibility semantic-value assertion in the available Flutter environment. See `checks.json` for the final executed suite counts and environment. No full-app accessibility, security, network-performance or purchase guarantee is made.

## Repeat the editorial first-deal rehearsal

Read the current article and open `host-card.html` in a browser. Complete the invitation blanks manually, keeping the real code with your intended group. This file does not send messages, create rooms, store answers or contact Felted. It can be printed or copied. Our examples below are constructed role scenarios, not observations of real users.

| Case | Setup | Expected check |
|---|---|---|
| Two players, room for six | Host plus one friend seated, both with chips and sitting in | Two is enough to start; four open seats are not blockers. |
| One player plus a watcher | Host seated; friend on the Rail | Still one seated player; cannot use the watcher as the second player. |
| Watch from the start | Guest intends only to spectate | Select Just watch instead before completing the code; the sixth character submits the selected intent. |
| Late arrival | Guest tries to take a seat during a hand | A seat must wait for a permitted between-hands opportunity. The offered watching route is conditional on Rail access. |
| Full table | Every available seat occupied | Watching does not reserve a seat. Wait for an opening to play. |
| Room created early | Current capped table created ten minutes before the group is ready | Ten minutes of its 30-minute cap have elapsed; the first deal does not restart the cap. A hand already underway may finish before closure. |

## Files and limits

`role-flow.svg` explains guest control order; `hero.svg` is conceptual cover art. Neither is an app screenshot. The article supplies the same essential information in text. `host-card.html` is an original manual coordination aid, not a measured intervention. `checks.json` records bounded checks; `SHA256SUMS` checks this package's bytes.

The sole arithmetic example is 2,500 starting chips / 10-chip big blind = 250 big blinds. It does not predict duration, an optimal setup, strategy, or results. No random sample, seed, solve, model or customer dataset was used.

Current source defaults and access limits can change or differ in older builds. Scope excludes Clubs, public tables, offline campaign rules, real-money arrangements and private tournaments. This rehearsal is useful only if the current app permits the selected role/settings. Consult the displayed controls and support when they differ.