Swipe for more columns →
How the sentiment score works. Built only from each token's main pair's on-chain trades as reported by DexScreener. No social data, no opinions.
buy pressure = buys ÷ (buys + sells), averaged over the last 1h and 24h (24h only if the last hour had fewer than 5 trades)
buyer breadth = unique buyers ÷ (unique buyers + unique sellers), 24h (DexScreener's public data feed doesn't include unique buyer/seller counts, even though its app shows them, so this is usually dropped; see below)
momentum = 0.5 + (24h price change, capped at ±50%) ÷ 100
score = 100 × (0.4 × buy pressure + 0.3 × buyer breadth + 0.3 × momentum). 60+ leans bullish, 40 or less leans bearish, in between is neutral. If unique buyer/seller counts are missing, buyer breadth is dropped and the other two are reweighted to 4/7 and 3/7. Pools with fewer than 20 trades in 24h aren't scored, and neither are stablecoins or wrapped/major assets (USDC, USDT, WCRO, WETH, WBTC…), since their price is pegged or set off-chain.
This is a description of recent trading activity, not a prediction.
UltFans