shillscore
Crypto-Twitter signal accuracy. BTC-excess returns on every token mention, ranked by damped median.
90-day BTC-excess. Sparse today; fills out as seed mentions cross the 90-day mark.
Insiders — score leans on one token (≥50% of matured calls). Overwhelmingly project accounts ranked on their own coin (e.g. @superformxyz → UP). Useful to see who shills their own bag well, not who finds alpha.
90d leaders
ranked by damped median BTC-excess · follow them on X
When you sign in with X, this strip will flag accounts you don't yet follow. For now the follow links open X in a new tab.
how the leaderboard is built
running mean BTC-excess per account · each dot = one matured call · curve ends today
Every dot is one matured call. Click a dot to see the tweet, the token it called, and how much that call moved the account's line.
| acct | n | mean (line) | median | dampened |
|---|---|---|---|---|
| @AskVenice | 5 | +497.7% | +434.3% | +307.1% |
| @superformxyz | 17 | +164.6% | +141.2% | +124.1% |
| @TurboToadToken | 62 | +23.8% | +19.4% | +18.7% |
| @injective | 20 | +23.3% | +20.3% | +18.2% |
| @CryptoHayes | 14 | +19.1% | +14.2% | +12.2% |
| @cryptocom | 16 | -2.6% | +0.8% | +0.7% |
| @Crypto_Crib_ | 39 | +0.5% | -0.3% | -0.3% |
| @pangolindex | 25 | -3.9% | -6.4% | -5.8% |
| @okx | 6 | -12.9% | -20.2% | -14.9% |
| @MetaMask | 6 | +13.9% | -21.6% | -15.9% |
Why three numbers? The chart plots the running raw mean — easy to read off as a curve but volatile. Each dot is one matured call; click it to see the tweet and how much it nudged the mean. The leaderboard sorts by dampened = median × √(N / (N + 5)), which trims tails and penalizes thin samples. Hover or click a row to isolate that account.
who caught the 90d winners
top 9 tokens by BTC-excess return over 90d · coloured dots = top-10 accounts · grey dots = other tracked accounts
Survivor-biased by design. Each panel was selected because the token beat BTC over the window — so any dot looks good. The interesting signal is which accounts got there first vs late on each call, and how often the same accounts recur across panels. Skill lives in the leaderboard table below; this view is about coverage and timing.
| acct | matured | median excess |
|---|---|---|
| @AskVenice | 5 | +434.3% |
| @superformxyz | 17 | +141.2% |
| @TurboToadToken | 62 | +19.4% |
| @injective | 20 | +20.3% |
| @CryptoHayes | 14 | +14.2% |
| @cryptocom | 16 | +0.8% |
| @Crypto_Crib_ | 39 | -0.3% |
| @pangolindex | 25 | -6.4% |
| @okx | 6 | -20.2% |
| @MetaMask | 6 | -21.6% |
| # | handle | matured | win % | median excess | damped | top token | followers |
|---|---|---|---|---|---|---|---|
| 1 | @AskVeniceVenice | 5 | 100% | +434.3% | +307.1% | VVV100% | 63,803 |
| 2 | @superformxyzSuperform | 17 | 100% | +141.2% | +124.1% | UP100% | 125,713 |
| 3 | @TurboToadTokenTurbo 🐸 | 62 | 97% | +19.4% | +18.7% | TURBO100% | 110,655 |
| 4 | @injectiveInjective 🥷 | 20 | 95% | +20.3% | +18.2% | INJ100% | 544,694 |
| 5 | @CryptoHayesArthur Hayes | 14 | 93% | +14.2% | +12.2% | HYPE50% | 791,773 |
| 6 | @cryptocomCrypto.com | 16 | 50% | +0.8% | +0.7% | BTC63% | 3,065,086 |
| 7 | @Crypto_Crib_Crypto Crib | 39 | 46% | -0.3% | -0.3% | BTC59% | 206,037 |
| 8 | @pangolindexPangolin (PangoPerps is out!) | 25 | 24% | -6.4% | -5.8% | AVAX72% | 118,616 |
| 9 | @okxOKX | 6 | 17% | -20.2% | -14.9% | XAUT83% | 5,695,477 |
| 10 | @MetaMaskMetaMask 🦊 | 6 | 33% | -21.6% | -15.9% | XAUT83% | 1,433,613 |
| 11 | @zksyncZKsync | 10 | 10% | -22.0% | -17.9% | ZK100% | 1,472,688 |
| 12 | @PolymarketMoneyPolymarket Money | 20 | 30% | -20.3% | -18.1% | META50% | 98,410 |
| 13 | @rainbowdotmeRainbow | 6 | 17% | -25.8% | -19.0% | RNBW100% | 117,745 |
| 14 | @unusual_whalesunusual_whales | 9 | 0% | -33.0% | -26.5% | META89% | 3,735,628 |
| 15 | @worldlibertyfiWLFI | 7 | 0% | -44.0% | -33.6% | WLFI100% | 779,560 |
| 16 | @MANTRA_ChainMANTRA | The EVM L1 for RWAs | 55 | 0% | -40.1% | -38.4% | MANTRA100% | 380,722 |