Matchmaking and Invite Problems

Prove which stage fails

Treat invitation, visible party membership, and matchmaking as three separate checkpoints. An invite that is sent but never produces a visible member belongs to the first two stages. A complete roster that enters a queue and waits belongs to matchmaking. A party that breaks only after a failed queue or reconnect creates a fourth observation about current group state.

Record the platform mix, region, time with timezone, installed version, party leader, intended mode, crossplay setting, and last successful checkpoint. Do not invent a timeout threshold or error-code meaning; the official snapshot used here does not publish either value. The stage is the evidence that selects the next reversible check.

Verify access and version before rebuilding

Ask each player to confirm the displayed game version and the online multiplayer access required by their console account. A Steam PC participant should verify the current account session and broadband path. These checks establish visible prerequisites; they do not prove that platform services, game services, and every network hop are healthy.

Cancel the stale invite, choose one leader, and send one new invitation after all players return to the party surface. Verify every intended member by name before queueing. If one player still does not appear, continue with the friends and private lobby guide rather than using repeated matchmaking attempts to diagnose a missing party member.

Use crossplay as one controlled comparison

The dated August 13 issue list associated crossplay disabled with longer matchmaking. When the complete group accepts mixed-platform play, enable crossplay for one comparison while keeping party, mode, and time window unchanged. This is one comparison of the current queue condition; it does not prove that the queue is fixed or that the toggle was the sole cause.

If the enabled queue enters while the disabled queue did not, record both attempts and their start times. If neither enters, restore the group's intended preference and continue to the wider-status check. Do not publish a guaranteed wait time, region population, player count, or claim that one crossplay choice always matches faster, because none is supplied by the current sources.

Rebuild stale party state once

After a failed queue or disconnect, return every player to the party screen. If the visible roster is inconsistent, disband once, have one leader recreate the group, and confirm each member before starting one fresh queue. That controlled rebuild checks current party presentation without deleting local data or changing multiple account and network variables.

A successful rebuild shows that the party can be assembled at that moment. It does not establish why the prior state failed, and it does not guarantee the next match will remain connected. If one member repeatedly disappears at the same stage, preserve that player's platform and account-session result for a focused report.

Separate queue scope from a wider incident

Check the newest official feed and compare its publication date and version with the installed client. If a notice explicitly matches matchmaking, a platform, or a region, follow only the current instruction in that notice. An older issue entry is historical evidence, not proof that the same condition remains active.

Use the server and known-issues guide when multiple independent parties or platforms show a shared failure. This page has no live queue telemetry, so silence in the feed cannot establish normal service and a social report cannot establish a global outage.

Escalate a bounded matchmaking report

Provide the platform mix, game versions, required online-access checks, region, time, party size, leader platform, mode, crossplay state, invite result, roster result, and one controlled queue comparison. State whether solo queue or a different party was tested, but do not expose account tokens or private identifiers.

The goal is to show the smallest repeatable boundary: one account cannot join, one party cannot queue, one platform combination fails, or a broader set appears affected. Support can investigate that evidence; this manual cannot assign service priority, promise a fix, or infer a permanent platform restriction from the captured launch-window notes.