UP SKILL FXBackTransparency
Where the prices come from, what they cannot tell you, and the exact arithmetic behind every readiness score.
Last updated 23 August 2026.
Up Skill FX connects to no live market feed, no broker and no paid data provider, and it never will. Every price you see is either free historical one-minute foreign exchange data or a deterministic teaching series generated by the platform. This is an architectural constraint, not a temporary limitation.
Replay sessions and case studies are built from free historical one-minute bars for major pairs and gold. The data is embedded in the application at build time, is never redistributed as a data product, and is used strictly to teach reading of real tape.
Simulator sequences and most structured drills use price series generated by a fixed algorithm. The same drill produces the same series for every learner and on every run, which is what makes results comparable and repeatable. They are not market data and they model no real instrument.
Every scenario carries a data-origin flag. The Evidence Report reports real-tape and synthetic results as separate segments so a strong synthetic record can never disguise a weak record on real historical tape.
Limitations of historical data
Free historical data is supplied as-is. Bars can be missing across illiquid hours, rollovers and holidays, and prices can carry vendor artefacts. Bars are never interpolated to hide a gap; a gap is shown as a gap.
Foreign exchange has no central exchange, so any historical file is one venue's view. Spreads, slippage, swap and commission are not present in the raw data and are applied only where a drill explicitly models them.
Every replay and case study was selected after its outcome was known. Selection bias is disclosed rather than hidden: sessions are chosen to illustrate a reasoning process, not because they represent typical conditions or a profitable strategy.
Nothing in the historical record supports a claim about future prices. A profitable replay or simulated sequence is not evidence of skill and a losing one is not evidence of error — which is exactly why scores measure process, not profit.
How scores are computed
No capability score, mastery percentage or readiness verdict is derived from profit or loss. They are computed from what you demonstrated: recall tested under assessment, application under scenario commitment, and process followed under observation.
The same inputs always produce the same score. No random factor, no model judgement and no hidden weighting enters the arithmetic; the mentor comments on scores but never sets them.
Each observation is blended into a prior score as prior × (1 − weight) + observation × weight, and each evidence kind carries a ceiling it can never push a score past — a self-reported baseline or a lesson checkpoint cannot on its own demonstrate mastery.
Module mastery is a composite: instruction 30 points, assessment 55, application 15. A module quiz at or above 80% is treated as reliable reasoning; at or above 65% it counts as evidence but not as mastery. Mastery also requires at least one committed applied case where a case exists.
Assessed reasoning ages. After 90 days evidence is treated as stale and is progressively discounted, bottoming out at 60% of its original weight by day 270. A score that falls is not a penalty; it is the platform declining to claim something it can no longer see.
Section readiness comes from the section assessment plus the completion state of every lecture in that section. A section unlocks the next only when it passes at or above your configured mastery bar, which can be raised but never set below the pass mark. Failing verdicts schedule targeted replays of the exact lectures that authored your misses.
Every capability movement is traceable in the Evidence Report drilldown to the attempt, the questions, the authoring lectures and the prior-to-after blend that produced it. If a number cannot be explained that way, treat it as a defect and report it.
It is not a qualification, not a licence, not a credential recognised by any regulator, and not a prediction that you will make money. Certificates record work completed inside this platform and nothing more.
| Evidence | Weight | Ceiling | Confirms |
|---|---|---|---|
| Self-reported baseline | 1 | 60 | Claim, awaits measurement |
| Lesson checkpoint (recall) | 0.4 | 70 | Measured |
| Assessment (tested recall and application) | 0.4 | 100 | Measured |
| Scenario reasoning (applied) | 0.5 | 100 | Measured |
| Journalled process (observed behaviour) | 0.25 | 85 | Measured |
Educational platform only. Nothing here is personal financial advice or a recommendation to trade. Leveraged foreign exchange and derivatives carry substantial risk of loss. Past performance and historical relationships do not guarantee future results. Simulations and assessments are educational; you remain responsible for your own decisions.