← BACK HOME
AUDIT
Honest aggregate stats across every call the engine has fired. A winner is defined as a call that reached at least +20% peak gain since the call price. All numbers are computed from the live database, no curation.
TOTAL CALLS
6809
6806 with outcome
WIN RATE
45.5%
peak >= +20%
AVG PEAK GAIN
+77.4%
MEDIAN PEAK
+6.7%
half of calls above this
2X+ WINNERS
1030
peak >= +100%
1.5X+ WINNERS
1641
peak >= +50%
1.2X+ WINNERS
2400
peak >= +20%
WINDOW
7d
since 5/25/2026
BY CHAIN
BUCKETCALLSWIN RATE+20%+50%AVG PEAKMEDIAN PEAK
BASE54435.6%193102+116.6%+9.4%
SOL352960.4%21321539+127.9%+36.7%
HL273628.1%750+4.6%+2.6%
BY SCORE BAND
BUCKETCALLSWIN RATE+20%+50%AVG PEAKMEDIAN PEAK
50-59168726.6%8840+10.7%+2.5%
60-69242849.4%977665+70.3%+8.0%
70-79168958.0%892650+124.8%+25.2%
80+58156.3%303224+130.3%+24.3%
TOP 15 PEAK WINNERS
Highest peak gain since call. Click any row to open the token.
RANKCHAINTOKENSCOREPEAK GAIN
#1BASE$HETFHYPE-ETF
48+9898.73%#2SOL$CUMcummunity
75+9863.47%#3SOL$grail
87+7215.76%#4SOL$DATBIHGAHDat Bih Gah
79+6007.60%#5BASE$gSPEEDISHOWSPEED
50+5076.23%#6SOL$SPCXSpaceExplorationTechnologiesCorp
66+4597.58%#7SOL$ballish
75+3648.48%#8SOL$10mi identify as a 10m runner
79+3648.38%#9SOL$SYNDROME
75+3310.04%#10SOL$Berries
84+3058.80%#11SOL$MOMOももちゃん
84+2653.55%#12SOL$SCATSScetch Cats
83+2526.12%#13SOL$ALIENSaliens.gov
76+2420.76%#14SOL$Odysseus
78+2288.43%#15SOL$DOODLE
74+2212.48%METHODOLOGY
- A "call" is a signal row in the database. Each token can be called more than once if it crosses the threshold again after the 30-minute cooldown.
- Peak gain is the high-water mark since the call, tracked every minute by the peak tracker. It never decreases. A token that pumped to +500% and round-tripped still shows a +500% peak.
- Per-chain winner threshold: spot DEX chains (Base, Ethereum, Solana) count a winner at +20% peak. Hyperliquid perps count at +5% because mark-price swings rarely reach +20% in our scoring window.
- Filtered out of the aggregates: known quote tokens (USDC, WETH, USDT, WSOL, DAI, etc) that should never have been a call target; rows reading above +5000% peak (these are almost always pre-fix supply- denominator artifacts, not real outcomes).
- Calls without a recorded peak (very fresh signals the tracker has not ticked over yet) are excluded from peak averages but counted in the total.
- Numbers are recomputed on every page load with a 60-second cache. No curation, no cherry-picking beyond the two filters listed above.