
A timer that is set, not assumed
If the timer is set after midnight, it is not a timer. The desk confirms a timer line exists in the live settings, not in a marketing panel.
Review criteria
Six axes. No star score, no winner wall. The method is published so a reader can repeat it on the next screen capture.

| Axis | What is checked | How the desk checks it |
|---|---|---|
| Onboarding | First-screen clarity, language, ID ask | Fresh install walkthrough |
| Table chrome | Format name, drop numbers, joker visibility | Screenshot of live table |
| Wallet clarity | Payment rail, KYC ask, hold text | Receipt date stamp recorded |
| Rule help | In-product help for declare and drop | Search inside the live app |
| Support path | In-product care and on-site chat | Test ticket without payment |
| Play controls | Session timer, deposit cap, break rule | Settings walk, not a marketing claim |
Sample: a single Speedwin build on a single device. Window: the date stamped on each receipt, not a calendar quarter. Weighting: equal weight per axis. No axis is folded into a star score. Re-run rule: a reader can repeat this on the next install and the next screen capture. Anything that cannot be re-run is not an axis. It is a rumour.

If the timer is set after midnight, it is not a timer. The desk confirms a timer line exists in the live settings, not in a marketing panel.

A close-account path should be reachable without a support ticket. The desk notes whether the exit is visible in the live menu.

Search-ad phone numbers are not verified. The desk uses in-product care and the on-site chat at /contact/ as the verified route.
There is no star score. Star scores are not re-runnable. Six axes are re-runnable.
This desk does not assert a licence. If a licence line appears on a live screen, that line is the source. State rules still apply.
The criteria list is published. Repeat it on a fresh install. Anything that fails to repeat is a rumour, not a criterion.
Onboarding is the first screen after install. A Speedwin onboarding screen should name what the product is and what it is not. It should ask for an email or phone, an OTP, and a screen language. It should not ask for a one-time password to a random wallet. A first screen that asks for a one-time password to a random wallet is the wrong first screen, and the desk notes that as an axis-level finding.
Table chrome is the surface a player sees after sign-in. The table chrome should print format, drop numbers, joker status, seat count, and a timer line. A table chrome that hides the drop numbers is unclear, not a feature. The desk screenshots the chrome and notes what is visible and what is hidden.
Wallet clarity is the surface a player sees when opening a wallet menu. The wallet should name the payment rail, the document ask, and the hold policy in plain words. A wallet that asks for a selfie and a back-dated document in the same step is unclear. The desk screenshots the wallet and notes the ask.
Rule help is the in-product answer to declare and drop questions. A Speedwin app should answer declare and drop questions inside the app, not by a link to a marketing email. A reader who has to leave the app to learn the rules is the wrong kind of reader. The desk notes whether the rule help exists inside the live app.
Support path is the in-product care route. The desk uses in-product care and the on-site chat at /contact/ as the verified routes. A search-ad phone number is not verified and is not printed here. The desk notes whether the support path is reachable without a payment.
Play controls is the settings menu. The settings menu should print a session timer line, a deposit cap line, and a break rule line. A settings menu that hides these lines is unclear. The desk notes whether the lines exist inside the live settings.
The criteria list is published so a reader can repeat it. A fresh install on a different device, the next day, with a fresh account, will produce the same observations on the six axes. Anything that fails to repeat is a rumour, not a criterion.
The re-run rule is also the reason this desk does not publish a star score. A star score is not re-runnable; the six axes are. A star score is also the reason this desk does not publish a winner wall. A winner wall is not source.
If a reader runs the criteria on the next Speedwin build and gets a different reading, the desk notes the change on the next dated note. If the reading is the same, the desk does not invent a change.
Open the store listing on a fresh install. Walk the onboarding screen. Screenshot the first ask. Read the table chrome before the first discard. Open the wallet menu and screenshot the ask. Search the in-product help for "declare" and "drop". Open the in-product care menu. Open the settings menu.
On the onboarding screen: name the product, name the language, name the credential ask. If the first ask is a one-time password to a random wallet, that is an axis-level finding. Note it.
On the table chrome: name the format, name the drop numbers, name the joker status, name the seat count, name the timer line. If the chrome hides one of those lines, that is an axis-level finding. Note it.
On the wallet menu: name the payment rail, name the document ask, name the hold policy. If the wallet asks for a selfie and a back-dated document in the same step, that is an axis-level finding. Note it.
On the in-product help: name the declare page, name the drop page. If the help sends the reader outside the app, that is an axis-level finding. Note it.
On the in-product care: name the menu, name the response. If the menu hides the support path behind a payment, that is an axis-level finding. Note it.
On the settings menu: name the timer line, name the deposit cap, name the break rule. If the settings menu hides one of those lines, that is an axis-level finding. Note it.
A star score is not re-runnable. The criteria list is re-runnable. The criteria list is the desk's number.
A star score also folds a method into a single number, which hides the axes. The criteria list names the axes. The axes are the desk's reading.
Axis one is onboarding. Open the store listing on a fresh install. Walk the onboarding screen. Screenshot the first ask. The first ask should be a credential (email or phone) plus an OTP. If the first ask is a one-time password to a random wallet, that is an axis-level finding. Note the finding. Move to axis two.
Axis two is table chrome. Open the table after sign-in. Screenshot the chrome. The chrome should print format, drop numbers, joker status, seat count, and a timer line. If the chrome hides one of those lines, that is an axis-level finding. Note the finding. Move to axis three.
Axis three is wallet clarity. Open the wallet menu. Screenshot the ask. The ask should name the payment rail, the document ask, and the hold policy. If the wallet asks for a selfie and a back-dated document in the same step, that is an axis-level finding. Note the finding. Move to axis four.
Axis four is rule help. Search the in-product help for "declare" and "drop". Read the results. The help should answer declare and drop questions inside the app. If the help sends the reader outside the app, that is an axis-level finding. Note the finding. Move to axis five.
Axis five is support path. Open the in-product care menu. Read the response. The response should arrive inside the live product, not by an external phone number. If the response asks the reader to call a phone number, that is an axis-level finding. Note the finding. Move to axis six.
Axis six is play controls. Open the settings menu. Read the lines. The settings menu should print a session timer line, a deposit cap line, and a break rule line. If the settings menu hides one of those lines, that is an axis-level finding. Note the finding.
Six findings, one per axis. A reader who runs all six axes on a fresh install has re-run the criteria. A reader who runs fewer than six has re-run a partial criteria. A partial criteria is a rumour, not a criteria.
If the six findings on a fresh install match the six findings on the desk's reading, the criteria is stable. If they do not match, the desk notes the change on the next dated note. The desk does not invent a change.
A star score folds the six axes into a single number. The single number is easy to read. The single number is also easy to lose. A reader who reads the single number and forgets the axes has lost the method.
The desk publishes the six axes because the six axes are the method. The desk does not publish a star score because a star score is not the method.
A winner wall folds six readers into a single number. The single number is also easy to read. The single number is also easy to lose. A reader who reads the winner wall and forgets the criteria has lost the method.
The desk publishes the criteria because the criteria are re-runnable. The desk does not publish a winner wall because a winner wall is not re-runnable.
On the next install, the desk will re-run the six axes. If the six findings on the next install match the six findings on the current install, the criteria is stable. If they do not match, the desk notes the change.
A reader who wants the next reading should bookmark this page. A reader who wants the criteria should bookmark the criteria index. Both pages are re-runnable.