Sharing a device works when the handoff is part of the rule. Players need to know whose input the screen is requesting, which information should remain out of view, and when a reveal is ready for everyone. The same phone or laptop can support several people, but it does not create privacy merely because a panel is labeled host only.
This guide explains roster setup, public and private phases, a complete four-person quiz example, common recoveries, and local record handling. It also covers two, six, and twelve participants. The instructions describe the current Quirky Query Hub controls rather than a synchronized online room: people share the local interface and take turns through the actual game sequence.
Start with harmless fictional sample material. A robot's chosen snack is enough to demonstrate a private answer and reveal without collecting personal information. Once the handoff is clear, participants can decide whether to use their own preferences, keep the fictional version, or choose another format entirely.
Prepare the roster before private entry
Check Game Setup and the displayed player list before starting. Use names or nicknames the group can distinguish. Duplicate display names can exist, but two identical labels may make the scoreboard difficult for people to read. A small distinguishing nickname is often clearer than requiring the group to remember which identical label occupies which position.
Use the individual name controls when renaming an existing participant. The current implementation maintains player identifiers separately from display names. Bulk editing the comma-separated list is a setup action, not an instruction to transfer one person's identity by rearranging names. Set the membership and order deliberately, then confirm whose turn the screen shows before entering results.
Avoid changing membership halfway through collecting votes or guesses. A new roster can change what a completed round means and may reset relevant parts of the session. Finish or deliberately restart at a clear boundary if someone joins or leaves. The host should explain the decision rather than let people infer it from a changed counter.
Read each game's role arrangement. In a personal quiz, the person setting the secret answer should remain outside the guesser roster for that round if the intention is for only the others to guess. The browser collects a guess from every listed player; it does not infer that one name secretly belongs to a non-guessing author.
Name the phases aloud
Use three phase labels: public prompt, private entry, and reveal. A public choice can be read to everyone. A private answer is entered while others look away and the screen is positioned appropriately. A reveal is shown only when the game reaches that stage. Saying the phase aloud is more reliable than assuming every person understands the significance of a button.
Check the whole screen before handing it over. A closed host-only disclosure may hide one answer while a custom input still exposes the same text. Browser history, saved sets, or other visible panels may also be inappropriate to open during a private phase. Do not treat a social look-away agreement as a technical guarantee against someone determined to inspect the device.
Use a shared operator when passing the physical device is awkward. A participant can ask for the prompt to be read and choose how to enter a response. If the answer needs privacy, agree a method that actually preserves the intended phase or choose a public format. Do not promise anonymity while requiring someone to announce a private response to the whole room.
A complete four-person quiz round
In this fictional demonstration, author Mira chooses the question, “What snack does our imaginary station robot prefer?” The stored answer is pear. The guessers are Hassan, Nell, and Owen. Mira is not in the three-name player list for this round. All material is invented so the first trial needs no real personal preference or private story.
Mira opens How Well Do You Know Me, checks the roster, and starts the game as indicated. In the authoring phase she enters the question and correct answer. Others look away. She chooses Hide answer and start, then checks that the next screen requests the first guess rather than displaying her answer before handing the device onward.
Hassan sees the public question and his named turn, enters pear, and chooses Submit guess. The interface then requests the next remaining player's input. Nell enters Pear, and Owen enters banana. Each waits for the screen to show whose turn comes next. Nobody opens setup to edit another person's name or tries to reveal the answer early merely to check a theory.
After all three listed guesses are submitted, the game reveals the answer and compares the guesses. In the current implementation, letter case is ignored for this comparison, so pear and Pear match; banana does not. The game does not interpret arbitrary synonyms or decide how well someone understood the imaginary robot. Explain a text-match limitation separately from the human conversation.
They discuss why a fictional robot might prefer pears and then choose whether to continue. If Hassan becomes the next author, the group updates the next round's guesser roster before entering a new secret. Mira may now be a guesser, and Hassan stays outside that list. Rotating authorhood is a manual hosting arrangement, not a role the software infers from who last held the device.
The example shows a complete private-entry, three-guess, and reveal cycle. Use it as a practice procedure, not as a claim that personal knowledge can be measured by three matching words. Once the controls are understood, the group may continue with fictional questions or use voluntary ordinary preferences without treating an incorrect guess as a relationship verdict.
Distinguish game workflows
Two Truths and a Lie begins with an author entering two truths and one invented statement, then shuffling and hiding the marked lie. Guessers take turns choosing from the displayed statements. The reveal follows the collected guesses. Use a fresh fictional trio for the demonstration, and do not leave the author's entry screen facing the guessers.
Weird Would You Rather collects choices as the device is passed between voters. It is a different process from a quiz author hiding a correct answer. A participant's reason can remain spoken after the choice is entered. Do not imply the result represents an outside audience or a universal preference; it is the local round's recorded responses.
In a ranking game, use the provided order controls and submit the intended ranking. A cooperative discussion that produces one order is a manual group choice, not twelve independent rankings automatically inferred from the conversation. Explain which result is being entered. The group should be able to connect each on-screen submission to the actual player or arrangement that produced it.
In 20 Questions Mystery, the host handles the private target and records responses to spoken or typed questions. The app does not interpret their meaning. Keep the secret disclosure closed when showing the public log. Guess is separate from the question counter, and the current candidate allows an incorrect pre-limit guess to continue rather than always ending the round immediately.
Recover from an accidental reveal
If someone sees a secret early, pause and choose a fresh question or target. They can also become a recorder if the others want to finish an unscored practice round. Do not require fake guessing or organize a public trial of whether the glance was intentional. The useful repair is a clear next procedure, not a verdict about the person's character.
If a wrong person submits during a turn, stop before collecting more responses. Read the displayed state and decide whether to restart that round as practice. Do not promise an undo or a transfer of the recorded response unless the actual interface provides it. A manual note can explain a practice result, but it does not change what the software stored.
If an answer fails text matching but the host recognizes the idea, distinguish the human acknowledgment from the stored result. Explain the expected wording without pretending the software performed semantic judgment. For a new quiz, use clear short answers and check how the current game compares them before collecting personal material. A harmless test is easier to correct than a disputed real answer.
If a participant leaves midway, avoid quietly adding a substitute name to finish their response. Membership affects the round's meaning. Use a fresh round or state the practice arrangement openly. Keep names stable within a collected set so the reveal is understandable to people reading it afterward.
Decide what remains on the device
Quirky Query Hub uses local browser records for game state, including relevant custom content and history. It is not a promise that spoken answers were never heard or typed answers disappear after the device is passed. Before personal entry, explain whether the group wants anything saved and who controls the shared device afterward.
The local backup controls can export records to a JSON file. Treat that file according to its contents: personal quiz answers and custom text may be included. A backup is a recovery copy, not an automatically private message or live multiplayer invitation. Do not send it to everyone merely because they participated in the session. Agree who should retain it and why.
Clearing local games is a data action, not a casual way to move to the next question. The current interface asks for confirmation and retains a pre-clear recovery copy in the browser. Do not describe it as eliminating every copy everywhere. An exported file or previously shared record remains a separate item that the local clearing action does not retrieve.
At the end, ask what custom material should remain and explain the actual available controls. Do not promise selective deletion if the interface only offers a broader operation for that data. You can choose less personal sample content from the outset, keep spoken answers out of fields, or use an unsaved manual activity where retention would not suit the group.
Restore deliberately, outside an active turn
Before a major data change, make a backup if the group wants to retain the existing session. When importing, read the file summary and choose the intended merge or replacement mode. Matching imported record keys can replace existing records even in a merge; do not assume that mode automatically combines every player, guess, and score into one result.
The current restore flow validates the incoming records and blocks play while the operation runs. Wait for the verified success message and remounted game before continuing. Previous timer callbacks are cancelled by that flow. This is a description of the implementation, not a claim that a browser recovery test has been performed as part of writing this guide.
If validation or a transaction fails, read the error and inspect the available recovery path rather than repeatedly confirming an uncertain import. Do not edit a downloaded JSON by guesswork merely to force acceptance. Keep the original file as a separate backup. Recovery should happen with deliberate understanding of the record being restored, not while players are waiting to submit a timed response.
Adapt for two, six, and twelve people
With two, one can author while the other is the sole guesser, then swap after a complete round. A public choice game can simply alternate device turns. Keep the roster appropriate to the actual role: the author does not belong in a guessing list merely because both people are physically present.
With six, a designated operator can read public prompts while private-entry phases still receive the necessary physical handling. Make the current player name clear before each input. A brief wait for the correct turn is preferable to collecting several answers under the same name and trying to reinterpret them later.
With twelve, consider the time required for individual entries. Short choices may work; long private answers may be better in a manual pair format or a smaller group. If using teams, explain that the entered team label represents a shared response under that manual convention. The software does not independently know which people contributed to the team's decision.
Run a short trial and keep a checklist
For a ten-minute trial, use two minutes for roster and roles, two for a fictional author demonstration, three for the private guess cycle, two for the reveal and a question about the procedure, and one to decide what to retain. The total is ten, with no requirement to enter personal content afterward. The trial is complete when people understand the handoff and can identify which screen belongs to which phase.
Before play, check the roster, author/guesser distinction, input method, private view, reveal condition, and intended record. Afterward, review saved content, exports, and any recovery action before returning the device to its owner. For deeper examples, read How to Create Your Own Personal Quiz, How to Play Two Truths and a Lie, and How to Score a Ranking Showdown. Make each handoff a clear action the people and the actual interface both support.