Hunt conditions

Every rule a suggestion passes before it reaches the screen — and, just as important, the rules it does not pass. Companion to the 0DTE engine docs (v2 / v3 / v4) and the Option Bounce rules.

Two independent sources feed this page

1. Market scan (/api/hunt-scan) — the rows labelled “market scan”. These pass every gate in the table below: cap floor, liquidity, volume surge, move capacity, earnings, and a multi-day expiry.

2. SPXW coach (/api/zero-dte-coach) — the row labelled “SPXW coach”. This is the v4 engine, and it is a separate system. The hunt mode only chooses how far out-of-the-money to go; none of the gates below are applied to it. Its contracts are

0DTE
by definition — they expire the same day.

If you are looking at an SPXW contract on a hunt desk, monster’s rvol, capacity and market-cap rules had nothing to do with it. Judge it by the coach’s own score and confidence, documented in the v4 engine notes.

Per-mode gates — market scan
A candidate must clear all of these. Failing any one drops it silently.

worm

+30%

Orderly movers only. Names already running hard are rejected — they belong to the other desks.

Min score50
Min RVOL1.2×
Min move capacity2%
Max move already made8%
StrikeATM
Min days to expiry3 (next Friday weekly)
Imminent earningsskipped entirely
Ranked byconfidence

witch

+100%

Confirmed momentum that still has room left in the move.

Min score55
Min RVOL1.5×
Min move capacity4%
Max move already madeno cap
Strike1 step OTM
Min days to expiry1 (next Friday weekly)
Imminent earnings−20 confidence
Ranked byconfidence

monster

+200% or more

Violent eruptions only. An empty list on a quiet tape is the correct answer, not a fault.

Min score45
Min RVOL
Min move capacity7%
Max move already madeno cap
Strike2 steps OTM
Min days to expiry1 (next Friday weekly)
Imminent earnings−15 confidence
Ranked byexplosiveness
Universe and liquidity
Unlimited suggestions
Shared market-scan settings for Worm, Witch and Monster Hunt.

Loading Hunt settings…

Which lane is running
Chosen automatically by the New York clock.

Live tape · 9:30–16:00 ET — dip and jump models on live prices. A dip needs an explicit CALL/PUT side (bounce or breakdown confirmed); a jump must be JUMPING or COILED, never merely LOADING.

Extended hours · 4:00–9:30 and 16:00–20:00 ET — after-hours prints and premarket gaps of ±3% or more, mapped to gap-continuation candidates.

Planning mode · weekends and overnight — ranks the last completed session by where each name closed inside its own range. Top 30% on heavy volume is a CALL continuation candidate, bottom 30% the mirror; midrange closes are dropped as indecision. These are candidates to confirm at the open, not entries.

What the hunt does not do

It does not predict earnings direction. Names reporting within a day are skipped on worm and confidence-penalised elsewhere, because technicals cannot see a scheduled event.

It does not read intraday structure. The scan uses daily-resolution data — today’s and yesterday’s bars plus the last trade. The 5-minute ceiling/floor detector in live commentary is separate and covers only the charted symbol.

“3x–5x” describes capacity, not an outcome. It means the move has room if it completes. Far-OTM options also go to zero faster when it does not.

It keeps no record of past suggestions. Nothing is persisted, so the hunt’s historical hit rate cannot currently be audited — by anyone. Treat performance claims about it as unverified until that changes.