Transparent · no black box

How we grade a fair launch

Every A–F grade you see on the radar is one number, computed the same way, from public on-chain data. Here is the entire rubric — every signal, every penalty, published — so you can check our work on any token.

The model, in one line

Start at 100. Subtract for what's unfair.

A launch begins with a perfect score. We read five distribution signals off-chain and subtract a fixed, published penalty for each. What's left is the fairness score — which maps to a letter grade.

100
every launch starts here
5
on-chain penalties
=
A–F
fairness grade
The five signals

What we penalize, and by how much

These are the on-chain fingerprints most rug pulls leave behind: supply reserved before you can mint, and supply pooled in too few wallets. Each penalty below is exactly what the code subtracts.

1 · Pre-mint

up to −40

Supply the deployer reserved for themselves before the public mint opened. Zero pre-mint is the mark of a true fair launch; a big pre-mint hands the deployer a pile of supply to dump. We subtract 2 points for every 1% of max supply pre-minted, capped at −40 (reached at 20% pre-mint).

Pre-mint (% of max supply)Penalty
0% — none reserved0 🟢
Any pre-mint−2 per 1%
≥ 5% (flagged red)−10 and up
≥ 20% (cap)−40 (max)

2 · Top-wallet concentration

up to −35

The share of minted supply held by the single largest wallet. One wallet holding half the supply can crash the price alone — the heaviest penalty in the model, because it's the clearest concentration risk.

Largest wallet holds…Penalty
≥ 50% — extreme−35
≥ 30% — heavy−24
≥ 20%−15
≥ 10%−8
< 10%0

3 · Deployer self-allocation

up to −25

How much of the supply the deployer's own wallet holds right now — separate from pre-mint, because a deployer can also accumulate by minting heavily. A deployer sitting on a large stake is a wallet that can walk. Check any address's full record on its deployer track-record page.

Deployer wallet holds…Penalty
≥ 30%−25
≥ 15%−12
≥ 5%−5
> 0% but < 5%0 🟢

4 · Holder spread

up to −15

How many wallets hold the token at all. A launch with a handful of holders has no real market — a broad holder base is what liquid, distributed supply looks like. Early launches are naturally thin, so the penalty is gentle and fades as the base grows.

Holder countPenalty
< 5 holders−15
< 15 holders−7
< 50 holders−2
≥ 50 holders0 🟢

5 · Top-5 combined

up to −12

The share of minted supply held by the five largest wallets together. This catches distribution that looks fine at the very top but is quietly pooled across a small cartel of wallets just below it.

Top 5 wallets hold…Penalty
≥ 85% combined−12
≥ 65% combined−6
< 65% combined0

The final score is clamped to 0–100 and rounded. Every per-token report shows exactly which of these fired, in plain English, under a "why this grade" breakdown.

Score → letter

What the letter means

The final 0–100 score maps to a grade on these fixed thresholds — the same cutoffs everywhere the grade appears.

A
85–100
Broadly fair
B
70–84
Mostly fair
C
55–69
Some flags
D
40–54
Concentrated
F
0–39
High risk
One source of truth

The same grade, everywhere

The radar, reports, leaderboard, API and badges don't each guess — they all import one shared scoring module. So a token can never score A on the badge and F in the report. Every surface reads the exact rubric above.

Read this part

What the grade can't tell you

A transparent score is only honest if it's honest about its blind spots. A high grade is a good sign, not a green light.

It's a smoke detector, not a guarantee.

The grade reads on-chain distribution — the signals most rugs leave behind — but a fair-looking launch can still fail, and a low grade can still moon. It is never financial advice. Always do your own research.

It can't see wallet clustering.

Supply split across ten wallets that all belong to one person looks more distributed than it is. On-chain data alone can't prove two addresses share an owner, so genuine concentration can hide behind many addresses.

It's a snapshot, not a forecast.

The grade reflects distribution right now. A launch that's fair today can concentrate tomorrow as wallets accumulate — which is exactly why the radar and alerts re-read it live.

It scores distribution, not intent.

"Fair launch" describes how supply is spread, not whether the team is honest or the project is any good. Zero pre-mint and broad holders can't tell you if anyone will keep building.

Now grade one, then mint it for less.

Once a launch clears the rubric, the cheapest honest way in is a non-custodial mint — you sign from your own wallet, keys never leave your device. The dominant bots quietly take 5%; the Fair-Launch Radar mints the same on-chain token for a flat 1.75%.

1.75%vs 5% on the bots
Check our work

Run the rubric on a real launch

You've seen every penalty. Now open the live board and watch it grade real KRC-20 launches A–F from on-chain data — every "why this grade" line traces straight back to the table above.