How SER Score works

Original, token-relevant content

Posts that are clearly about the token and say something of your own. A clear reference to the token, original thinking and media that adds context all help.

Genuine engagement from real audiences

Replies, quotes, reposts, bookmarks and likes from real accounts. Only engagement that passes SER’s quality and fairness checks counts.

Consistency

Showing up over time — recurring, qualified contribution across a scoring window rather than a single burst.

Holding the token

Being invested in the token you post about is recognised.

Long-term reputation

Your track record on SER across projects and scoring windows. It builds with consistent, clean participation and travels with your SER Passport.

Attributable onchain impact

Onchain actions SER can technically attribute to you — for example a wallet that arrives through your tracked link and goes on to hold the token. Correlated activity, such as a price move after your post, is never treated as caused by you.

• Volume for its own sake. Posting more doesn’t beat posting better.
• Bought or coordinated engagement. Engagement the fairness system excludes is removed before scoring.
• Duplicates and spam. Copied, templated or near-identical posts don’t qualify.
• Off-topic or out-of-window posts. Posts that aren’t about the tracked token or campaign, or that fall outside its window, are rejected — with the reason shown.
• Paid status. Tiers can’t be bought.

• Your SER Score, tier and rank, and how many points you need for the next tier — on your dashboard.
• For every tracked post: its status and the reason, its final score, and plain-language signals describing what helped it — for example Original content, Token referenced or Engagement: strong.
• Your fairness status and reward multiplier, taken from the published penalty table.
• Your public SER Passport: score, tier, rank and history.

SER doesn’t publish point-by-point breakdowns, weights or thresholds. If a result looks wrong, open the post on your dashboard and appeal — a reviewer sees the full detail.

Before rewards are allocated, SER applies your epoch fairness status for that campaign or token. The statuses and their effect on rewards are published in the penalty table on fairness; how statuses are determined is not. Fairness status is separate from long-term SER Trust, explained in trust and appeals.

Your SER Score places you in one of six tiers:

Tiers below APEX are split into sub-levels I, II and III. Tiers reflect performance and trust: they are earned, can go down as well as up, and can’t be bought. Your dashboard shows your own progress toward the next tier; the thresholds themselves are internal configuration.

• Anti-gaming. A published formula becomes a target. Keeping weights, limits, thresholds and detection rules private keeps the incentive on genuine contribution and protects reward pools for the sers who earn them.
• Same rules for every ser. Every account in a scoring run is scored with the same active version. Scores are never edited by hand; corrections happen through reviews and appeals.
• Server-computed and reproducible. Scores are computed by SER’s servers with deterministic arithmetic, so the same inputs always produce the same result.
• Accountable. You always see your own results, statuses and the reasons behind them, and you can appeal a post classification or a fairness decision.

• Scoring parameters live in versioned, validated configuration stored by SER — not in the app.
• Every reward epoch records the scoring version and the fairness-rules version it used. Historical epochs are never silently rescored.
• Reputation snapshots published to SER’s reputation registry contract carry a fingerprint of the configuration that produced them, so a parameter change is always detectable onchain — without publishing the parameters.
• Version contents are internal and managed by SER administrators.

Token pages show a SER Social Score and a social velocity status. The social score summarises qualified social activity around a token: how many qualified creators post about it, the quality and originality of that activity, how many distinct people engage, its growth, and its reach after spam filtering. Velocity — Surging, Rising, Stable or Cooling — compares recent qualified activity with the token’s own recent baseline. Component weights and status boundaries are internal.

Both describe social activity, not price. They are not a rating of the token and not investment advice.

The steps below describe Robinhood Chain, where rewards from pons tokens are claimed onchain. On Solana, rewards from pump.fun coins are allocated by the same scoring run, in the coin’s pair token (SOL by default), and are sent to your saved wallet instead of being claimed — see getting paid.

1. The epoch window closes; the scoring run uses the active configuration version.
2. Each eligible account’s epoch score is computed, including its epoch fairness adjustment. Activity from sources that aren’t reward-eligible still counts toward reputation but is excluded from allocations.
3. The funded total is allocated in proportion to score with exact integer (bigint) arithmetic and largest-remainder rounding, ties broken deterministically by address — allocations sum to the funded total exactly.
4. Allocations become leaves of a Merkle tree; the root, asset, total, window and claim deadline are published to the distributor contract.
5. You claim with your proof. Each allocation can be claimed once; after the deadline, unclaimed funds can be swept.

index is your position in the epoch’s allocation list and amount is in raw units of the epoch’s asset. Anyone may submit a claim, but funds always go to the account in the leaf.