Measurement method — KursTradera.pl

Artefact 01 / 03

Broker risk-percentage protocol

You do not quote “70–80% of accounts are in the red” from memory. It is a property of one specific broker and the date of one specific disclosure.

Fill in 0 / 5

The broker or platform you use
The date you read the risk warning
The exact percentage of retail accounts at a loss (from that disclosure)
The product it concerns (for example CFDs)
The URL, or the place in the app, where the disclosure appears

If you cannot write that line down with a date, you do not scale your size. This is evidence-of-work hygiene, not a statistic of your strategy.

Artefact 02 / 03

Overnight cost sheet vs 1R

Before you work out the “profit from a setup”, work out what it costs to hold the position overnight. Overnight cost often swallows more than an average setup earns.

Fill in 0 / 4

The overnight cost of your instrument (swap or financing, from the broker’s fee schedule)
Your typical 1R in the same currency
How many nights an average position is “held”
Overnight cost as a percentage of your 1R

If overnight runs above a few per cent of your 1R, swinging that instrument becomes statistically expensive — regardless of whether the setup “looks good”.

Artefact 03 / 03

A mini-sample with no optimisation

You do not backtest “to be sure of a profit”. You take the rule, fix the sample in advance and test it once — without fitting the parameters to the result.

Fill in 0 / 4

The exact entry and invalidation rule (before you look at the data)
A sample fixed in advance (for example the last 20–30 occurrences of that type)
The result after costs: spread, commission, overnight
The result written down with no change to the rule along the way

Optimising on the past is fitting to noise. The mini-sample is there to prove that the rule is consistent — not that it “works out on average” after 40 rounds of tweaking.

Where the numbers and claims come from

We link the document, not the publisher: A link to a regulator’s home page confirms nothing. That is why a source points at a specific page: the supervisor’s announcement, the instrument specification, the publisher’s article. Before an address goes on the site, we open it and check that the document really does contain the term being cited.
A book is a source too — if we say so plainly: Some concepts exist only in the professional literature. Expectancy and R-multiple are terms from books about risk management, not from regulators’ documents. In that case we cite the author and the title and note that there is no online version to check — instead of propping them up with a random article from the internet.
A number with no source does not go into the text: A percentage, a cost, a limit, a deadline — every such value has an origin. If we could not confirm it, we describe the gap as unverified or drop the claim. Guessing a number is worse than not having it, because the reader cannot see the difference between data and guesswork.
The author is named, not anonymous: Material is signed by the editorial team together with the author’s name, not by “the team”. In the site’s structured data the author is a person linked to an editorial profile — so that responsibility can be attached to the content, and not only to a URL.

What we do not publish

Signals, entry calls and “setups of the day” — we do not run a service with recommendations.
Price forecasts and promises of profit. The outcome depends on the market, on costs and on the reader’s decisions.
Author details that cannot be confirmed: education, licences, the results of a personal account.
Other people’s content under our own name, and numbers with no traceable origin.

The review rhythm is explicit: content carries the date of its last change, and rankings the date their terms were checked. We make corrections in the text together with a note about what changed — an error found and described is part of a site’s credibility, not a disgrace. To report one, just send it through contact with the address of the page and the source that contradicts our claim.