How Lexicon keeps setup tests structured and reviewable.
This page explains the trust-first methodology behind the Lexicon Beta: deterministic rule definitions, historical testing workflow, anchor and timing logic, and the limits that keep research conclusions in bounds.
A structured research process, not a signal machine.
State the setup as explicit, deterministic logic — indicators, anchors, sequence, and entry conditions.
Run the definition against historical data under visible assumptions about fills, slippage, and structure.
Read the output as evidence to examine, not as prediction. Note what held and what did not.
Adjust one deliberate variable, then retest — so every change to the research stays visible.
Setup definitions are explicit, repeatable, and encoded with a unique DNA signature to ensure every historical test is perfectly reproducible.
Lexicon is built around deterministic rule definitions so the same setup can be reviewed, rerun, and compared without silent drift between tests.
The goal is not to describe a vague trading idea. The goal is to define the rule clearly enough that the logic, comparisons, sequence, and entry conditions can be evaluated the same way each time.
The core workflow is simple: define the rule, run the test, review the output, refine one variable, retest.
Define the rule. Run the test. Review the output. Refine one variable. Retest.
Historical testing is meant to support a disciplined research loop. A trader defines the setup, runs the historical test, reviews the output, adjusts one variable, and retests.
That structure matters because it keeps research changes visible. It is easier to understand what changed, what stayed stable, and whether a result improved because of one deliberate adjustment or because the logic drifted in multiple directions at once.
Multi-step setups need ordered logic and must-hold conditions.
Many setups are not one-condition events. One condition can happen first, another can follow, and the setup can be required to persist long enough to matter before an entry is allowed.
That is why sequence and persistence matter. The methodology assumes that ordered logic such as AND, OR, THEN, FOR N bars, and ENTER should remain visible in the QDL setup definition instead of being compressed into a vague signal label.
Setup context should reflect the levels and timing references traders actually watch.
Anchor and timing logic matter because many setups depend on prior-day levels, opening ranges, session highs and lows, VWAP context, first-window highs or lows, and other time-aware reference points rather than on one indicator alone.
Reference anchors are meant to stay tied to information that would have been available at the time of the test. The purpose is to keep setup context realistic and reviewable, not to let future information leak backward into the rule.
Historical tests are useful research evidence, but they do not remove uncertainty.
Historical tests depend on assumptions about price data, bar structure, fill behaviour, slippage, commissions, and how a rule is interpreted. Those assumptions need to stay visible because they affect what a result can and cannot support.
Lexicon does not present historical output as prediction. Historical tests are evidence to review, not predictions. They can help a trader examine a setup definition, but they do not guarantee future edge or prove that a rule will behave the same way in changing market conditions.
Research software still requires judgment, skepticism, and user responsibility.
Lexicon is in Private Beta. It is research software only. It is not signals, not financial advice, not a profit system, and offers no guaranteed results.
Users are responsible for reviewing assumptions, challenging conclusions, and avoiding the mistake of reading backtest output as certainty. The methodology is designed to support structured research, not to replace responsibility or turn historical evidence into a recommendation.
If methodology clarity matters to you, apply to review the Beta workflow.
Private Beta is aimed at traders who care how a setup is defined, what assumptions are being made, and whether the resulting research can be reviewed carefully instead of taken at face value.