Median win rate
The share of positively closed trades. The documented oscillator range is 60.19–60.89%. Win rate alone says nothing about the size of wins and losses.
UNDERSTANDING THE EVIDENCE
Numbers are useful only when it is clear what they measure.
This page therefore starts with the research question rather than a headline result. It then explains the metrics, the control group, historical APS findings and the limits of the current research state. Each number is presented only as strongly as the underlying test supports.
VERSION 1 · DOCUMENTED STATE
A broad test base with an explicit boundary on what it proves
Version 1 examines the decision core on end-of-day data across many markets and analytical methods. Breadth matters, but it does not replace realistic intraday simulation. Key figures are therefore presented together with their definition and methodological limits.
94real instruments
11market segments
60oscillators
5,640systematic individual tests
11.3mdocumented trade events
17years of data
Important: The 5,640 tests are not 5,640 statistically independent scientific experiments. Instruments are correlated, oscillators share inputs and formula families, and all tests use the same underlying architecture. The more relevant information is therefore cross-market consistency and comparison with identically treated random data – not the raw size of the number.
EVIDENCE STATUS
Each central statement receives a visible status
Rather than placing everything under a single label such as “validated”, the site distinguishes between documented findings, historical evidence, items that require revalidation and development objectives.
Documentedsupported by the current source or test baseHistorical findingavailable, but not presented as universal performance evidenceRevalidation requiredto be reproduced under Version 2 conditionsPlannedpart of the target architecture, not a finished product feature
Documented
Version 1 & control group
Results, metric definitions and limitations of the existing structural test.
Historical + documentedAPS Risk Layer
Mechanics, development history and historical comparison paths, clearly separated from revalidation.
Version 2Replication & extension
Intraday ordering, costs, APS integration and forward tests are added step by step.
1 · WHAT WAS TESTED?
Whether the same decision logic remains consistent across many markets
Version 1 examines a rule-based decision core on historical daily data from January 2009 to December 2025. Ninety-four real instruments across eleven market segments are evaluated with sixty oscillators under the same underlying logic. The 94 × 60 matrix creates 5,640 systematic individual tests.
The purpose of this breadth is not to create a large-looking number. It is to test whether the observed structure appears only in one market or one analytical method, or whether it persists across materially different instruments and oscillator variants.
94real instrumentsacross 11 market segments
60oscillators23 originals · 28 variants · 9 proprietary
5,640individual tests94 instruments × 60 oscillators
11.3mtrade events11,319,504 documented events
2 · WHY A CONTROL GROUP?
94 real markets are compared with six parametrically different synthetic trend regimes
An unusual metric can arise from genuine market structure or from the construction of the test itself. The 94 real instruments were therefore supplemented by six separately reported synthetic OHLC control series, creating a uniform matrix of 100 price series. All six controls come from the same generator family but use different settings for trend drift, trend switching, drift persistence and noise. The resulting paths range from near-pure noise to a highly persistent trend path. This design is deliberate because the Version 1 logic is trend-following rather than mean-reverting.
Documented
Version 1: 94 real instruments × 60 oscillators = 5,640 real-instrument tests; 6 synthetic control series × 60 oscillators = 360 control tests. Total: 100 price series and 6,000 test combinations.
HYPOTHETICAL / SIMULATED RESULTS
The quantitative values in the adjacent section are historical, simulated or backtest-derived research results unless expressly identified as realised live account results. They are therefore presented as hypothetical performance information for compliance purposes.
Material assumptions and classification: Version 1 is an end-of-day structural research test rather than a customer-account return simulation. Its hit-ratio metrics do not use an initial account balance or profit-reinvestment assumption. Unless a specific historical comparison states otherwise, the Version 1 baseline does not include commissions, slippage, market impact, intraday event ordering, complete stop/break-even/trailing logic or full APS integration; no management or incentive fees are modelled because no managed customer account is represented.
HYPOTHETICAL PERFORMANCE DISCLAIMER
HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH ARE DESCRIBED BELOW. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN. IN FACT, THERE ARE FREQUENTLY SHARP DIFFERENCES BETWEEN HYPOTHETICAL PERFORMANCE RESULTS AND THE ACTUAL RESULTS SUBSEQUENTLY ACHIEVED BY ANY PARTICULAR TRADING PROGRAM.
ONE OF THE LIMITATIONS OF HYPOTHETICAL PERFORMANCE RESULTS IS THAT THEY ARE GENERALLY PREPARED WITH THE BENEFIT OF HINDSIGHT. IN ADDITION, HYPOTHETICAL TRADING DOES NOT INVOLVE FINANCIAL RISK, AND NO HYPOTHETICAL TRADING RECORD CAN COMPLETELY ACCOUNT FOR THE IMPACT OF FINANCIAL RISK IN ACTUAL TRADING. FOR EXAMPLE, THE ABILITY TO WITHSTAND LOSSES OR TO ADHERE TO A PARTICULAR TRADING PROGRAM IN SPITE OF TRADING LOSSES ARE MATERIAL POINTS WHICH CAN ALSO ADVERSELY AFFECT ACTUAL TRADING RESULTS.
THERE ARE NUMEROUS OTHER FACTORS RELATED TO THE MARKETS IN GENERAL OR TO THE IMPLEMENTATION OF ANY SPECIFIC TRADING PROGRAM WHICH CANNOT BE FULLY ACCOUNTED FOR IN THE PREPARATION OF HYPOTHETICAL PERFORMANCE RESULTS AND ALL OF WHICH CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS.
3 · WIN/LOSS COUNT RATIO
1.5725 is not a percentage, a return or a profit factor.
The metric used here is the winners-to-losers count ratio: number of winning trades ÷ number of losing trades. It has no unit. A value of 1.0 means equal numbers of winners and losers; 1.5 means 1.5 winners per loser. Values below 1.0 mean more losers than winners.
Across the sixty oscillators, the median for real-market data is 1.5725. The synthetic control group has a median of 0.6012. None of the 5,640 real-instrument tests falls below 1.0, while 346 of 360 random tests do.
Do not confuse this with profit factor: the count ratio measures only how many trades finish positive or negative. Monetary profit factor is gross profit divided by gross loss and therefore reflects the size of gains and losses.
Median winners-to-losers count ratio winners ÷ losers · no unit
What is shown?
The real test matrix and random series sit on different sides of the 1.0 neutrality line in the documented dataset.
How should it be interpreted?This is a structural Version 1 finding. It is not a return promise and does not replace intraday-realistic execution testing.
4 · OTHER KEY METRICS
The important figures with their actual meaning
Median trading rate
The share of evaluated situations in which the rules actually produce a trade. The documented range is 44.82–48.31%.
Count-ratio range
Across the sixty oscillators the aggregated ratio varies by only 0.0485 points, roughly three per cent of the median. This indicates limited dependence on one oscillator, not statistical independence.
Tests below 1.0
No real-instrument test in the documented dataset falls below a count ratio of 1.0. This is an in-sample consistency observation, not a guarantee for new data.
5 · AN IMPORTANT LIMITATION
5,640 tests are not 5,640 independent experiments
The number accurately describes the breadth of the test matrix. It must not be interpreted as if every test were statistically independent. Markets are correlated, oscillators partly share price inputs and formula families, and all tests use the same pivot and decision architecture.
Shared market structure
Related instruments can move similarly, so the raw test count contains more observations than truly independent information sources.
Related oscillators
Several analytical methods share inputs or mathematical families. Low dispersion is informative, but sixty oscillators are not sixty fully independent models.
Shared rules
The same pivot and decision logic underlies the entire matrix. That is intentional for consistency testing, but it limits an interpretation based on independent experiments.
6 · HISTORICAL APS COMPARISON
One paired 81-trade path illustrates mechanics – not universal superiority
A documented 2019 comparison evaluated the same 81-trade path once conventionally and once with APS protection logic. The figures therefore illustrate how APS affected that specific path. They are not presented as general evidence of future performance.
Outcome of the 81-trade path
Conventional€3,391.96
With APS€5,446.04
The APS path ended 60.6% higher in this historical comparison.
Gross profit ÷ gross loss
Conventional1.31569
With APS1.46985
Monetary profit factor increased by 11.7% on this path.
Largest decline from a prior peak
Conventional€1,867.54
With APS€1,773.50
The reconstructed maximum drawdown was approximately 5.0% lower.
Status: historical finding · revalidation required
Version 2 is intended to evaluate APS under identical intraday execution, cost assumptions and an unchanged baseline. Only that isolated comparison can quantify its incremental contribution to the full framework more robustly.
7 · WHAT VERSION 1 DOES NOT PROVE
Open questions are part of the result, not footnotes.
No complete intraday simulation
Daily bars do not always reveal which price level was reached first. That ordering can change both result and risk path.
No full cost model
Commissions, slippage and potentially market impact need to be represented in execution-level testing.
APS not yet fully integrated
APS has its own historical test base. Its incremental contribution is to be isolated under Version 2 conditions.
Forward testing still pending
Previously unused data and rolling walk-forward tests are part of the next validation stage.
DOCUMENTATION
A reader should be able to trace a result back to its basis.
A number alone is insufficient for professional review. A material finding should therefore carry the data used, the rules in force, the system version, test conditions and known limitations. It should also be clear whether the result is historical, independently reproduced or still awaiting Version 2 revalidation.
The public layer explains meaning and limits; deeper documentation is intended to show how the finding was produced technically and methodologically.
Documentation principles →NEXT
How Version 2 is intended to address the open questions.
Look up technical terms
Definitions of profit factor, intraday, slippage, walk-forward, high-watermark and other terms are available in the glossary.