
A Cancel Is Also a Decision
Cancelling a working order is an active choice: the trader re-evaluated the plan and withdrew the instruction. In a live market that choice has consequences — the position that never existed, the risk never taken, the opportunity declined. In simulation, the cancel can be rehearsed and studied with zero cost.
Because cancellations leave no fill record in most simulators, they are the easiest decisions to lose. The journal is where they survive.
What to Capture in a Cancellation
A cancellation entry needs four facts: what order was working, when it was cancelled, what the simulated market showed at that moment, and why the cancel was chosen. The 'why' should be written as observed, not as justified — 'price approached the limit and I withdrew it' is a valid entry even when the reason was discomfort.
- Order parameters at time of cancel (type, limit, quantity).
- Elapsed time the order was working.
- Simulated market state when cancelled.
- Stated reason, recorded honestly.
Distinguishing Discipline From Hesitation
Two cancels can look identical in a log yet mean opposite things. Cancelling because the invalidation condition triggered is discipline executing the plan. Cancelling because the price drifted close and nerves intervened is hesitation overriding the plan. The facts are the same; the reason field separates them.
Review sessions should count both types. A rising count of discipline-cancels is a healthy trend; a rising count of hesitation-cancels is a pattern to address in the next study cycle.
Filing the Cancel in the Reviewbook
Give cancelled orders the same status as filled ones: a full row in the Review Sheet with 'Simulated fill status: Cancelled before fill.' The artifact of the entry is not a trade — it is a documented decision to withdraw.
Over months, the cancellation log becomes a map of how the learner responds to uncertainty: which conditions trigger withdrawal, whether the trigger was planned, and whether the plan itself needs revision.
The Cancel Log — Fields
| Field | Example Entry | Purpose |
|---|---|---|
| Original plan reference | Entry ID the order came from | Links cancel to intent |
| Cancel reason (written) | 'Spread widened beyond assumption' | Separates decision from drift |
| Market state at cancel | Synthetic price, spread, volatility tag | Context for later review |
| Time in book | 3 synthetic minutes | Measures patience vs. plan |
| Replacement action | None / new order / exit | Closes the loop |
Journal Checklist
- Is the cancel linked to the original plan entry?
- Is the reason written before the outcome is known?
- Is the market state at cancel moment captured?
- Is 'no replacement' recorded as a choice, not an omission?
- Does the log distinguish 'cancelled by me' from 'expired by model'?
Review Note
Never delete a cancelled order from the journal. Deleting it does not erase the decision — it erases your ability to learn from it.
Frequently Asked Questions
Because cancellation is a decision with a reason and a consequence. Unlogged cancels are the most invisible bias source in any journal.
Log it as 'model-expired' — a different event with different lessons (your expiry setting vs. your willpower).
Same as fills: did the cancel follow a pre-written rule or react to noise? A rule-based cancel that 'missed a winner' is still correct process.
Continue Exploring
Comparing Planned and Simulated Execution
Building a plan-versus-result comparison that measures deviation without turning into self-judgment.
Trade Review Without Hindsight
Techniques for reviewing simulated decisions without letting the known outcome rewrite the reasoning.
Costs and Assumptions in Simulation
Commissions, delays, and model limitations: how to document what the simulator charges, assumes, and quietly omits.