Methodology

How Vesra reads a token

Every rule and threshold behind the scores, taken straight from the configuration that computes them.

vesra-score-0.4.0MomentumObserved riskData coverage

Principles

Three separate signals
Momentum, observed risk and data coverage are never blended into one number. High momentum does not hide risk.
Rules, not predictions
Hand-picked, deterministic heuristics. The same data always gives the same score. They are not a fitted or validated model and do not predict prices.
No guessing
Missing data is shown as “Insufficient data” or “Unknown”, never filled with zeros or estimates.
Addresses, not people
One person can use many addresses. A transfer is not a sale. No token is treated differently, including any Vesra token.

Which tokens are scored

Scoring reads holders and pool state from the chain, so it runs about every 10 minutes for at most 100 tokens: tokens launched in the last 24h with at least 10 trades, and older tokens while they are trending (at least 30 trades from 10+ addresses and 0.5 ETH of buys in 6h), busiest first. Everything else shows “Not scored”.

Momentum (0–100)

Four features over the last 6h, each turned into a percentile against fobo tokens in the same age bucket (under 1h, 1h–6h, 6h–1d, 1d–7d, over 7d). The score is the weighted sum. It needs at least 10 trades from 5 addresses and 8 comparable tokens; otherwise it reads “Insufficient data”.

New buyer addresses
Weight
30%
What it measures
Addresses that bought in the second half of the window (last 3h) but not in the first half.
Net buying
Weight
30%
What it measures
Buys minus sells in ETH, divided by buys plus sells (from -100% to +100%).
Continuity
Weight
20%
What it measures
Share of the 12 equal slices of the window in which net buying was positive: sustained, not one burst.
Buy distribution
Weight
20%
What it measures
1 minus the concentration (Herfindahl index) of buy volume across addresses: high when many addresses buy similar amounts.

Feature	Weight	What it measures
New buyer addresses	30%	Addresses that bought in the second half of the window (last 3h) but not in the first half.
Net buying	30%	Buys minus sells in ETH, divided by buys plus sells (from -100% to +100%).
Continuity	20%	Share of the 12 equal slices of the window in which net buying was positive: sustained, not one burst.
Buy distribution	20%	1 minus the concentration (Herfindahl index) of buy volume across addresses: high when many addresses buy similar amounts.

Observed risk (Low · Medium · High · Unknown)

Each check produces a finding with a severity. The risk level is the most severe finding. Until the holder distribution is computed, the level stays “Unknown” unless a check is already High. Holder shares exclude system addresses: the Uniswap V4 pool manager (which holds each pool's unsold supply), fobo contracts (hook, vesting, limit orders) and burn addresses.

Holder concentration
What it measures
Share of supply held outside system addresses by the largest address and the 10 largest.
Medium
top 10 ≥ 30% or top 1 ≥ 10%
High
top 10 ≥ 50% or top 1 ≥ 20%
Creator selling
What it measures
Share of the tokens the creator wallet bought in the pool that it has sold since. Selling more than it bought there, or selling without buying there, is High.
Medium
≥ 25%
High
≥ 60%
Sell price impact
What it measures
How much selling 1% of total supply now would move the price (fees excluded).
Medium
≥ 3%
High
≥ 10%
Serial creator
What it measures
Tokens launched by the same creator wallet in the last 24 hours.
Medium
≥ 10
High
≥ 50
Creator sold before
What it measures
In the creator's earlier tokens it traded in (at least 3), the share where it sold.
Medium
≥ 50%
High
≥ 80%
Early buyers
What it measures
Share of supply that addresses buying in the first 60s hold now, from real balances.
Medium
≥ 15%
High
≥ 30%

Check	What it measures	Medium	High
Holder concentration	Share of supply held outside system addresses by the largest address and the 10 largest.	top 10 ≥ 30% or top 1 ≥ 10%	top 10 ≥ 50% or top 1 ≥ 20%
Creator selling	Share of the tokens the creator wallet bought in the pool that it has sold since. Selling more than it bought there, or selling without buying there, is High.	≥ 25%	≥ 60%
Sell price impact	How much selling 1% of total supply now would move the price (fees excluded).	≥ 3%	≥ 10%
Serial creator	Tokens launched by the same creator wallet in the last 24 hours.	≥ 10	≥ 50
Creator sold before	In the creator's earlier tokens it traded in (at least 3), the share where it sold.	≥ 50%	≥ 80%
Early buyers	Share of supply that addresses buying in the first 60s hold now, from real balances.	≥ 15%	≥ 30%

Informational findings (a creator transferring tokens to other addresses, the size and lock of the dev buy, how far the price sits above the pool's floor) are shown but never raise the level.

Data coverage (Low · Medium · High)

How much the other two signals can be trusted. The level is the lowest of three factors:

History: share of the token’s life with indexed trades
High
≥ 95%
Medium
≥ 50%
Sample: trades in the window
High
≥ 100
Medium
≥ 20
Freshness: age of the newest indexed block
High
≤ 120s
Medium
≤ 15 min

Factor	High	Medium
History: share of the token’s life with indexed trades	≥ 95%	≥ 50%
Sample: trades in the window	≥ 100	≥ 20
Freshness: age of the newest indexed block	≤ 120s	≤ 15 min

Data sources and limits

Launches, trades, buybacks and dev-buy locks: fobo factory, hook and vesting events plus Uniswap V4 swaps on Robinhood Chain, read by Vesra’s own indexer after 20 confirmations. $FOBO pairs are valued in ETH at the $FOBO/ETH price of the moment.
Holders: each token’s Transfer history. Prices and floors: the token’s V4 pool and the fobo hook, read live. USD: Chainlink ETH/USD on Robinhood Chain.
Raw trades are kept for 30 days; trades by a token’s creator are kept longer because the creator checks use them.
Creator history counts every fobo launch; a token “ran 10×” when it traded at 10 times its start price. Launches through shared contracts are attributed to the sending wallet as they are resolved.

Changelog

vesra-score-0.4.02026-11-09
- Vesra now covers fobo launches on Robinhood Chain instead of Pons V2. Momentum, coverage and the creator, holder and early-buyer checks work as before.
- Momentum compares tokens of a similar age only: every fobo token trades the same way, in one hook-owned pool from launch on.
- Creator tax is no longer a check: fobo's creator share is a fixed 0.6% of every trade for every token.
- New information (never raises the level): the creator's buy in the launch transaction and whether it is locked in fobo vesting, and how far the price sits above the pool's floor.
- Creator history counts launches that traded at 10x their start price instead of graduations.
vesra-score-0.3.12026-11-08
- Creator selling: when the creator address sold more tokens than it bought on the curve or pool, or sold without buying there (tokens from the launch or from transfers), the finding is High and says so. Before, the first case showed a share above 100% and the second was not flagged at all.
vesra-score-0.3.02026-09-27
- The creator is now the wallet that sent the launch transaction, also when the token was launched through a shared contract (Multicall3, launch bots). Before, creator checks looked at the contract.
- New finding: the creator address launched many tokens in the last 24 hours.
- New finding: in its earlier tokens, the creator address sold in most of the ones it traded.
- New finding: addresses that bought in the first 60 seconds still hold a large share of supply (from real balances, not from trades).
vesra-score-0.2.02026-09-25
- Baseline published with Vesra V1: momentum against tokens of the same phase and age; observed risk from holder concentration, creator selling, sell price impact and creator tax; data coverage from history, sample size and freshness.

Scores are rule-based observations of on-chain data, not investment advice. About Vesra