Cards on Record
Next post ·Today's Cards 07:50 ET
Paper record · 1 Sep – 22 Oct 2026

Called before.Graded after.Nothing deleted.

+11.25R net of fees 70 cards 50 / 20 wins / losses 71% won
Drag to turn · pinch to zoom · tap a light
Two posts a day

Called at 07:50. Graded at 17:45.

The morning post names the card before it can trigger. The evening post grades it, net of fees, and replies to the morning post.

Today's CardsAs posted · Fri 23 Oct · posted 08:03 ET
ES is sitting on the week's shelf, so today it has a card each way. The short fires on a break of the shelf, and the long fires on a reclaim of the overnight high. Neither can fire before 10:15 ET, and each one's stop is the other's trigger. The SPX put spread rides the short. Yesterday's bitcoin short is still open.
▸ ES short2026-10-23-ES-1
ES loses the week's shelf: the leader's break spreads · 15m close below 7,676 from 10:15 ET, stop 7,714.25 · Near trigger at posting.
▸ ES long2026-10-23-ES-2
ES reclaims the overnight high: the fourth test bought · 15m close above 7,714.25 from 10:15 ET, stop 7,676
▸ SPX short2026-10-23-SPX-O1
SPX 7,650/7,600 put spread: the ES short's vehicle · put debit spread buy 7,650P / sell 7,600P, debit 23.5, options expiry 2026-10-30 · underlying 15m close below 7,637.36 from 10:15 ET, stop 7,674
+ 6 macro and crypto basket calls, graded separately · 2 cards still working from earlier posts
The RecordAs posted · Thu 22 Oct · graded 22 Oct 17:07 ET
Two small wins today, both from cards that hit their first target and then gave the rest back. The NQ short paid early before the market bounced. Yesterday's bitcoin long closed just above breakeven after fees. The SPX call spread lost as stocks fell. This morning's bitcoin short is open and slightly against us.
▲ WIN2026-10-22-NQ-1
NQ short, T1 hit, then stopped: +0.13R after fees (+0.13R gross, fees under 0.01R)
▲ WIN2026-10-21-BTC-1
BTC long, T1 hit, then stopped: +0.06R after fees (+0.11R gross, fees −0.05R)
▸ BTC short2026-10-22-BTC-1
open, marked −0.12R at the close of grading (not a result)
▸ BTC short2026-10-17-BTC-1
expired untriggered: no position, no grade
▸ GC short, open, marked +1.17R at the close of grading (not a result), T2 hit2026-10-20-GC-1
shadow card: graded, kept out of the record
2 more in the post · + 6 macro and crypto basket calls, graded separately · Paper record (pre-launch): 70 cards · 50W 20L · +11.25R net (+12.60R gross) · 71% won
The board

70 cards. 20 losses. All on it.

Every card the desk has graded, oldest first. Drag the playhead or press play; the line, the totals and the receipt follow it. Pre-launch paper record, 1 September 2026 to 22 October 2026.

+11.25Rnet of fees
+12.60Rbefore fees
0.02Raverage fee a card
50 / 20wins / losses · 71% won
−3.63Rworst stretch so far
Gold · shadow 8 cards · 3 / 5 · −3.88R net of fees · 38% won Kept out of the record above from 17 October 2026: still called before and graded after, losers included, until a 20-card forward run shows it pays.
The record, card by cardR net of fees
Receipt
← swipe for the next card →
A day on the desk

Four steps. Same every day.

Tap a step. The morning post is the proof: its timestamp comes before the bar that triggers the card.

The grading spec

Rules the code follows.

No person and no model decides whether a card won. The grade comes from the bars and these rules.

Bars15-minute bars from the feed the card is graded on: spot for BTC, ETH and SOL; for the futures the contract the card was written on, kept across a roll; leader stocks on regular hours only.
TriggerClose triggers on the bar's close; touch triggers on its high or low. Bars before a card's start time cannot trigger it.
FillAt the triggering bar's close. Beyond the chase price there is no fill, and nobody chases a gap.
AmbiguityOne bar touches stop and target: the stop counts. A bar that opens past the stop exits at the open.
Scale-outEqual parts at each target. Stop to entry after T1, to T1 after T2.
FeesEntries and stops as market orders, targets as resting orders, in R of the card's own risk. WIN or LOSS is decided net.
UnitR. 1R is trigger to stop. No dollars, no account sizes. The record leads with net R and the card count.

How the cards are made

Cards written each morning by a model analyst under a method and rules I design and approve. Code grades every card each evening against real 15-minute bars, fees and losers included.

What the desk never does

  • ×Place trades or manage money
  • ×Give advice about your account
  • ×Quote dollars or position sizes
  • ×Delete a losing card or a post
Free · Telegram and X · @CardsOnRecord

Get the cards at 07:50. Check the grade at 17:45.

Follow on Telegram Follow on X

The public record starts the day the channel goes live. Everything on this page is the pre-launch paper record.