Deployer rug history — what the numbers mean

Of all the signals Soliscope tracks, deployer history is the most predictive. Past behaviour is the best predictor of future behaviour — a wallet that has rugged six of its last ten launches is not reformed. The token name, the narrative, the community activity: none of it matters if the person pressing deploy has a 60% rug rate.

Soliscope groups the deployer wallet with every linked address into a deployer cluster and shows the history:

• Launch count — total tokens created across all wallets in the cluster
• Rug count — how many of those tokens ended with a price crash of more than 80%
• Rug rate — rug count divided by resolved launches

What you will not see is the wallet that funded the deployer. We built that signal, tested it (see the methodology), could not show it predicts anything, and removed it — the funder position turns out to be dominated by legitimate infrastructure: CEX withdrawals, bridges, launchpads.

There’s no universally safe rug rate, but Soliscope uses these as starting thresholds:

• 0 rugs, ≥3 launches — meaningful clean history. Not a guarantee, but positive signal.
• 1 rug — yellow flag. One data point, but worth noting.
• 2+ rugs — red flag. Pattern of behaviour.
• Rug rate >50% — exit signal regardless of launch count.
• New deployer (0 launches) — no history is not safe. Fresh wallets rug as often as veterans.

A new wallet with zero launches is genuinely ambiguous. Don’t interpret it as clean — interpret it as unknown. The other signals matter more when history is absent.

The intuition is appealing — serial ruggers rotate deployer wallets to keep any single address clean, so the wallet that funds those fresh deployers ought to be a hard-to-fake tell. We built the signal, shipped it, and then measured it honestly against tracked outcomes.

De-confounded (leave-one-deployer-out, so a signal is judged against launches it did not itself train on), the funder-chain flag did not predict rugs. The reason is unsurprising in hindsight: the funder position is dominated by legitimate high-volume infrastructure (CEX withdrawals, bridges, launchpads), which funds many things and rugs almost none of them. So we removed the funder flag from every verdict. Full write-up on the methodology page.

For a while afterwards we kept showing the funder address as neutral provenance. In August 2026 we removed that too, including from the paid API. Our own rule is that a number must survive de-confounded validation on productiondata before we publish it, and no funder measurement ever did — the two we ran were on local snapshots and disagreed with each other. Displaying it anyway, even labelled “not scored”, invited people to read something into it that we could not stand behind.

Deployer history is a strong negative filter — it eliminates bad bets efficiently. But it doesn’t tell you which token will graduate. A clean deployer launching a token that gets bundled, has 90% concentration, and zero organic buyers is still a loss. Use deployer history to eliminate the worst options first, then apply the remaining signals to find something worth trading.

If you want to understand what happens after you clear the deployer check, read how to find gem tokens on Solana.

See deployer history live on every token → Open the Soliscope feed