BSP / Buying analysis
How to Read a Signal Track Record Without Mixing Evidence Types
Verification begins by naming the record and what it covers. A model percentage, a competition result and a follower account are not interchangeable.
Name the record before judging it
A backtest applies rules to historical data. A forward instruction log records messages as they appear. A provider account records that account's execution and funding. A follower account records the subscriber's actual outcome. An external competition result adds the organiser's period and rules. State which object you have before making a comparison.
The source may establish one useful fact without establishing the rest. An organiser placement can corroborate a named competition entry; it does not demonstrate the return of a separate subscription. An instruction log can support publication history without proving an attainable fill.
Define the denominator and time boundary
Require the eligible population, start and end time, handling of open trades and treatment of amendments. A headline win rate without the counted population cannot answer how selective the archive is. Adding raw percentages across independent calls is not a funded account growth curve.
Keep the stated method fixed while reviewing the record. Choosing whichever target makes a past message look successful introduces hindsight. An unresolved multi-target or entry-range instruction should remain unresolved until a predeclared rule explains its treatment.
Link the instruction to execution separately
To make an account-level claim, obtain the relevant account record and funding context. Reconcile instructions with filled, rejected and skipped operations. A follower can legitimately differ from the source because of configuration or broker conditions; that difference should be explained, not hidden.
Costs, concurrent exposure and available capital belong in the account analysis. A sequence of gross isolated results does not by itself describe what one subscriber could hold at once. Keep that distinction when comparing providers.
Read independent verification narrowly
Identify the verifier, period, materials and scope of any audit. Do not expand a scoped attestation into a claim that every advertised plan was audited. Cryptographic timestamps likewise concern covered bytes and attestation timing; they do not certify profitability or fields absent from the commitment.
When source documents are missing, report the gap. Our public-documentation notes are not a substitute audit and do not manufacture a score from incomplete records.
Worked case: attractive history, unknown account
Hypothetical case: a page adds percentages from isolated instructions and labels the sum a portfolio return. The buyer has no funding schedule, simultaneous-position rule or broker record. Preserve the reported aggregate as a source statement, but do not reconstruct an account curve from it.
A different page shows a source account on a copying platform. That record may be useful for identifying the source, yet the follower's allocation, start time and operations remain separate. A platform-tracked source is not automatically an independently audited subscriber experience.
Inspect unresolved events before totals
Select examples of canceled, amended and still-open instructions, not only closed winners. Ask how the method treats them. A consistent population rule matters because unresolved cases can change which observations enter a headline.
Keep the counting policy prospective. Do not choose an entry or target interpretation after seeing which one yields the better result. If the original instruction and rule cannot resolve a case, label it unresolved and state how it affects the claim you can make.
Preserve evidence independence
Trace repeated claims to their original artifact. A provider statement quoted by several publications is one underlying statement, not several independent validations. A source that examined a record and a source that merely linked to it should not receive the same label.
The useful conclusion names the strongest supported scope and its remaining boundary. For example, original publication may be documented while broker execution is untested. Do not combine partial facts into a stronger account claim that no source established.
Decision Record
| Decision question | Record to retain |
|---|---|
| Which object is this? | Simulation, instruction, source or follower record |
| What is counted? | Eligible population and outcome policy |
| Who checked what? | Named verification scope and period |
Continue the Comparison
Source Documents and Scope
Primary documents support the stated mechanisms or attributed records. The buying procedure is BSP editorial analysis, not a paid-product, latency or subscriber-performance test.