Rules
Free entry. Five predictions per UTC day across views and likes combined.
Before you predict
- Participate directly from your X account. Current blue verification is required; gold and gray badges do not qualify.
- No website signup, wallet, deposit, or token purchase is required.
- Use someone else’s public, unedited post, at most 24 hours old.
- A views prediction needs at least 1,000 starting views. A likes prediction needs at least 100 starting likes.
- Choose one metric and one whole-number target or decimal multiplier within 2–64× its starting count.
- The bot must be accepting entries. Check service status.
Daily limits
- 5 accepted predictions per account per UTC day, across both metrics combined. The limit resets at 00:00 UTC.
- One accepted prediction per account and source post in total. You cannot add a likes prediction after a views prediction on the same post, or vice versa. The daily reset does not allow another prediction on that post.
After all five entries are used, another otherwise valid prediction from a currently blue-verified, public account can receive a daily-limit card. There is at most one card per account per UTC day. It creates no prediction and uses no entry. The card shows the next reset, or says no entries remain this season if entries close at or before that reset.
There is no platform-wide player or prediction quota. Provider rate limits, API budgets, or service availability can still pause new entries.
Season entries
Season 1 stays in draft until REACH starts it. The first season runs for 48 hours, accepting new predictions only during the first 24. The final 24 hours let existing predictions finish; unused daily entries cannot be used during that period. Test predictions remain in their separate season. A new season does not allow another prediction on a source you have already used.
Every prediction gets 24 hours
The clock starts when REACH records acceptance, not when you post on X. A prediction accepted Monday at 23:30 UTC has a Tuesday 23:30 UTC deadline. The midnight limit reset does not shorten it.
Your metric, target, starting count, deadline, immutable X identity, scoring version, and season are fixed at acceptance. Predictions cannot be edited or canceled. Reaching a target needs recorded official evidence within that 24-hour window.
Final season rankings wait for the existing final-check allowance and for every season prediction to finish. A late-processed result stays in its accepted season. Completed season rankings, lifetime points, and prediction history are retained.
When a message is unclear
Use one metric, one target, and one source post. Typos or other languages can be understood only when the meaning is unambiguous. Unclear or ineligible predictions receive no reply. REACH does not guess a target, metric, or duration.
Edits, deletions, and missing data
A recorded edit to an accepted X prediction command incurs a −100 point penalty, even after a recorded hit. It takes precedence over a source edit or missing data. Deleting the command does not cancel the accepted prediction; its original terms still apply.
For reach-v3 and reach-v2, an unavailable final source, a recorded source edit, or no usable final check within five minutes after the deadline voids the prediction for zero points. A count first meeting the target after the deadline is also void: the crossing time is unknown. REACH does not guess missing evidence.
Read the measurement and exception policy for the full schedule and outcome rules.
Fair play and account limits
Do not buy or coordinate artificial views or likes, use controlled accounts to evade limits, or interfere with data collection. Suspected manipulation can move a prediction into review.
A verification badge does not prove one person has only one account. X counts do not prove unique people or organic interest. Some manipulation may go undetected.
Rule versions and service requirements
A receipt records the accepted terms and scoring version. New predictions use reach-v3; existing reach-v1 and reach-v2 predictions keep their original settlement contracts. New rules cannot silently change accepted targets or regrade older predictions. Reviews need an auditable reason.
The legal operator and public legal contact have not yet been provided. No legal approval is claimed. X data access and bot operation depend on their provider requirements being met.