Repeatability

Repeatability is the quiet property that holds the other two up: the same inputs must produce the same call. A discretionary trader is a rule too, but a rule that drifts — the same chart can become a buy on a confident morning and a pass on a nervous afternoon. A model removes that drift by specification: the decision is determined by the inputs and the rule, not by the trader's state of mind. That is what makes its record a clean test of the rule rather than a diary of moods.

The discipline is simple to state and hard to keep, which is exactly why a model has the advantage: its rule does not get nervous. The hardest version of repeatability to fake is a call whose every field a stranger can confirm was set before the trade resolved.

How a timestamp before the outcome locks repeatability in

The strongest form of “the same call, fixed in advance” is a record a third party can verify. On the tested models here, the entry, target, stop and grade are written into a hash anchored to Bitcoin at publication — as a single SHA-256 hash. Because every field, the grade included, is inside that hash, none of it can be quietly revised after the fact: any change would break the receipt. Repeatability stops being a promise the operator makes and becomes a property a reader can check, call by call, before the outcome is known.

How do you actually know a strategy works?

Repeatability is the property that makes the on-chain check worth running at all: there is no point freezing a call in public unless the rule behind it would have produced the same call anyway. The weakest answer is a backtest, because a curve can be tuned until it fits the noise in old data and still fail the moment it meets a live market. A stronger answer is a forward record: a continuous series of calls, each logged before its outcome, with the losers left in. The strongest answer of all is a forward record a stranger can re-check without trusting the person who kept it.

That is exactly the standard the tested models here are held to. Every published call carries an A-to-D conviction grade, and the whole call — its entry, target, stop, grade and signal time — is condensed into a single cryptographic fingerprint that is anchored to a public ledger the instant it goes out. Touch any field afterward and the fingerprint no longer matches the receipt, so a confirmed match proves the call existed in that exact shape before the trade resolved. A record built this way is not a story you are asked to believe; it is a claim you can audit, call by call. Reading such a record line by line is a skill worth practising on any model, tested or your own.

What a bad version of this looks like

Repeatability leaks away the moment a single field is left to the moment rather than the rule. The common ways it breaks:

• Improvising the exit. Setting the stop or target after the entry — or moving it once price approaches — means the trade you are running is no longer the trade you tested.
• Sizing on feeling. Letting conviction-in-the-moment set the position size, rather than the stop distance and a fixed account percentage, makes two identical setups into two different bets.
• An unhashed grade. A conviction label that can be edited after the fact can be quietly upgraded once a call wins, which turns the grade into a marketing dial instead of a fixed input.
• A record kept after the close. Logging the call once you know how it went, rather than before, is not repeatability — it is memory, and memory keeps the winners and loses the losers.

Repeatability is what makes testability and measurability trustworthy: a record built from consistent, pre-committed calls measures the rule, while a record built from inconsistent ones only measures the trader's mood.