Out of 373,309 wallets in CoinCheck's launch ledger, these are the ones that graduated from the bonding curve most often. A minimum launch count is required, because a single lucky launch is not a track record.
Check a tokenGraduation rate rises with launch count: about 0.2% for creators with two launches, around 0.5% at four, and past 1.5% by eight. Repeat launchers are measurably better at getting a token off the curve than one-shot creators.
That is the one deployer signal this data supports without cross-chain work, and it is reproducible from the ledger. It is still weak evidence about a wallet, and a creator rotating wallets each launch defeats it entirely.
What these numbers are. CoinCheck's 94.6% aggregate is a bonding-curve graduation predictor over 79,519 resolved calls. It is not a rug-detection accuracy rate, and the bullish bucket scores 1%.
What graduation is not. Leaving the bonding curve means a token got enough buying interest to clear a low bar. It is not a safety endorsement, and a deployer who graduates often is not thereby running a good project.
Independence. No paid promotion. No ranking is bought.