PATTERN+ PREDICTIONS
Updated

Setups · named patterns re-measured on our own prices every day

Setups: what usually happens next, counted

4 named setups are re-measured on our own price history every day; 30 fire today. Each one is a plain rule; we count every past time it fired, how often the price went the expected way, and whether that held up on the newest cases.

Prediction from past patterns, not financial advice.

Weak and bleeding

✓ Held up on newer cases · 62% → 60% lagged BTC

One of the weakest 20% of coins over the last 90 days and still trading below its 20-day average.

Over the next 10 days these coins did worse than Bitcoin 62% of the time (6,145 cases); the typical gap to Bitcoin was −3.2% (average −1.0%).

Win rate61.6%lagged Bitcoin
Cases6,14510-day hold each
Average−1.0%vs Bitcoin
Typical−3.2%the middle case · below 0 = lagged
Best−77.3%for the setup
Worst+502.2%against it

Firing today · Oct 3, 2026

TRX58% down in 30 days PAXG67% down in 30 days WLFI53% down in 30 days XAUT57% down in 30 days LUNC58% down in 30 days XPL65% down in 30 days 2Z53% down in 30 days DEXE58% down in 30 days RIF65% down in 30 days ALLO69% down in 30 days BARD64% down in 30 days REQ55% down in 30 days RVN53% up in 30 days AT70% down in 30 days OSMO60% down in 30 days HOME60% down in 30 days SAHARA55% down in 30 days SAPIEN55% down in 30 days ROBO50% down in 30 days CYBER60% down in 30 days BANK50% down in 30 days KAT60% down in 30 days HMSTR55% down in 30 days GNS50% down in 30 days SOPH55% down in 30 days F60% down in 30 days SHELL60% down in 30 days CITY55% down in 30 days

Volume spike in an uptrend

! Mixed: no clear edge in older cases · 50% → 44% higher

Trading volume jumps far above its 30-day normal while the coin trades above its 20-day average, Bitcoin is above its 50-day average and the coin is one of the 100 most traded.

Over the next 3 days the price was higher 48% of the time (2,389 cases); the typical move was −0.5% (average +1.3%).

Win rate47.8%were higher
Cases2,3893-day hold each
Average+1.3%price move
Typical−0.5%the middle case
Best+176.2%for the setup
Worst−81.2%against it

Firing today · Oct 3, 2026

ZRO65% down in 30 days SAND65% down in 30 days

Fresh exchange listing

✓ Held up on newer cases · 37% → 28% higher

Listed for spot trading on a major exchange we track in the last 3 days.

Over the next 7 days the price was higher 33% of the time (132 cases); the typical move was −11.5% (average −4.9%).

Win rate33.3%were higher
Cases1327-day hold each
Average−4.9%price move
Typical−11.5%the middle case
Best+309.1%for the setup
Worst−82.4%against it

Firing today · Oct 3, 2026

Not firing on any coin today.

Big unlock this week

· Too few cases to check (0)

A token unlock of 2% or more of circulating supply is due within the next 7 days.

No past cases in our price data yet.

Win rate—were higher
Cases07-day hold each
Average—price move
Typical—the middle case
Best—for the setup
Worst—against it

Firing today · Oct 3, 2026

Not firing on any coin today.

The detailsthe exact thresholds behind each setup and how its cases are counted

The exact rules

The exact rules

as of · data/setups.py · setups.rule

Weak and bleeding

ConditionValue
weakest share of coins over 90 days0.2
coins needed that day to rank10
price vs its 20-day averagebelow
held for (days)10
measured as10-day return minus BTC's 10-day return
counts as a win whenlags BTC
one open case per coin at a timeyes

Volume spike in an uptrend

ConditionValue
volume above normal by more than (standard deviations)2.0
volume compared with the last (days)30
price vs its 20-day averageat or above
Bitcoin vs its 50-day averageat or above
among the most traded coins (top)100
held for (days)3
measured as3-day return
counts as a win whenprice higher
one open case per coin at a timeyes

Fresh exchange listing

ConditionValue
spot listing within the last (days)3
exchangestracked
held for (days)7
measured as7-day return
counts as a win whenprice higher
one open case per coin at a timeyes

Big unlock this week

ConditionValue
unlock of at least (% of supply)2.0
within the next (days)7
held for (days)7
measured as7-day return
counts as a win whenprice higher
one open case per coin at a timeyes

Method

How setups are measured

as of · data/setups.py

Every rule reads only data dated on or before the day it fires (unlock schedules are published ahead, so their future dates are fair game). A coin can hold only one open case per setup: after it fires, it cannot fire again until its holding window has closed, so cases never overlap. A case counts once its window has closed.

The honesty check splits the cases in time order: the first 60% against the newest 40%. A setup "held up" when the newer cases still win more often than not on the same side and keep at least half of the earlier edge; otherwise it "faded". Under 20 cases there is too little to check.

Prediction from past patterns, not financial advice. A setup firing is a count of what happened the last times, never an instruction to trade.