Live forward test · updates itself

Every call we make, scored in public.

The radar logs each signal before the outcome is known, then a bot scores it against live prices. Nothing is edited, nothing is removed — the misses stay on the board. Direction accuracy and plan profitability are scored separately under a versioned futures contract. Earlier research estimates are not presented as proof.

If you had risked $100 on every signal +$131 across 8 closed trades, after a 10bps round-trip cost. 5 of 8 reached target before stop (62.5%, against a 38.5% breakeven).
Too early to judge: under 20 closed trades nothing is proven either way.

Below is a different question, and the two can disagree without either being wrong. Direction asks only whether price was on the signal's side four hours later. The money above asks whether the trade plan reached its target before its stop within twelve hours. A call can be right at four hours and still have been stopped out on the way.

4h direction · prospective 87.5% 7 of 8 calls ended in the signalled direction from the observable live entry. Separate from whether the trade plan made money, below.
Avg 4h move / call +0.52% Mark-to-market move in the signal's direction at the fixed endpoint, before costs. It is not a trade P&L claim.
Cleared 0.25% bar 75.0% 6 of 8 calls moved more than the declared profit bar by the four-hour endpoint.
8Signals logged
0Open right now
12hAvg time to target
7 · 1Direction right · wrong
12h plan record 8 plans scored · gross average +0.273R. Sensitivity after assumed round-trip execution cost: 0bp +0.273R · 4bp +0.229R · 7bp +0.196R · 10bp +0.163R · 20bp +0.053R . Fewer than eight weekly blocks: no interval or profitability claim yet. These are cost scenarios, not claims about a user's fee tier or fill.
young record This board restarted on Jul 31 14:00 with the newest engine version. 8 signals is a start, not proof — judge any record over hundreds of calls, not dozens. We show it from day one anyway, because that is the point.
Version history

Old evidence stays visible

Engine versionRows preservedFirstLastStatus
execution-v4.0.0-usdm-futures-long-only 8 Jul 31 14:00 Aug 01 00:04 prospective futures contract
execution-v3.0.0-usdm-futures 79 Jul 25 10:24 Jul 30 13:45 older definition or market · excluded from the {{ ENGINE_LABEL }} aggregate
legacy 31 Jul 20 05:45 Jul 24 12:02 older definition or market · excluded from the {{ ENGINE_LABEL }} aggregate
legacy-archive_1784526131 34 Jul 17 06:03 Jul 20 05:02 older definition or market · excluded from the {{ ENGINE_LABEL }} aggregate
legacy-archive_1784210126 88 Jul 13 11:30 Jul 16 13:10 older definition or market · excluded from the {{ ENGINE_LABEL }} aggregate
Why you can trust this page

Three rules, no exceptions

1
Logged before the outcome

Every eligible signal enters the ledger with its observable entry, source-candle time, configuration fingerprint and engine version.

2
Two definitions, kept separate

Direction is scored at the four-hour endpoint. The plan is scored by first stop/target touch over twelve hours. Neither can overwrite the other.

3
Misses stay on the board

Rows are append-only. A rule change creates a new engine version; earlier versions stay preserved instead of being reset out of view.

The ledger

Recent signals, exactly as fired

ObservedContractCallSetupLive entry4h exit4h moveDirectionPlan
AAVE LONG capitulation 93.6900 92.5800 -1.19% wrong stop · 56m
HYPE LONG capitulation 52.4170 52.5580 +0.27% right expired · -0.54R
LDO LONG capitulation 0.3447 0.3463 +0.46% right stop · 9h 31m
BTC LONG capitulation 62,650.40 63,064.90 +0.66% right expired · +0.82R
LINK LONG capitulation 8.1210 8.1760 +0.68% right expired · +1.02R
TRX LONG capitulation 0.3260 0.3261 +0.04% right expired · +1.06R
BCH LONG capitulation 206.340 208.930 +1.25% right target · 12h
OP LONG capitulation 0.08520 0.08690 +2.00% right expired · +0.23R

Don't take our word. Watch the board.

Or ask about any USDⓈ-M perpetual contract yourself. Free while in beta, no signup.

Start beta chat →