How to quote a pass rate so a stranger can check it

"87% success rate" is the most common number in the prop firm passing market and the least meaningful, because it does not say what it is a rate of. A percentage with an undefined denominator is not a statistic. It is a mood.

Here is what has to be true before a pass rate can be compared with another one.

1. Define the denominator, in purchases

The candidate denominators are very different things:

• Evaluations attempted — every account that was worked on, including ones abandoned.
• Purchases — every time a client paid. This is the honest one for a buyer, because it is the unit they are deciding on.
• Clients — one person who bought four evaluations counts once, which can flatter or deflate depending on how they did.

A service quoting a rate per client while selling per evaluation has changed the meaning of the number without changing the number. The question to ask is literal: per purchase, or per client?

2. Include the resets, or the rate is survivorship bias

A retry policy creates an obvious accounting temptation: the first attempt fails, the retry passes, and the pass is counted while the failure quietly leaves the sample. The rate goes up and nothing about the service changed.

If retries are covered — and at a flat fee they usually are — then the honest figure counts resets as failures, because each reset is a purchase-equivalent the client did not have to make at another provider. Publishing "X of Y purchases reached funded status, including Z resets" is a weaker-sounding sentence that is worth more than a clean percentage.

3. Date the window and say what happened to open accounts

A rate over an undefined period is not reproducible. Rates drift with firm rule changes, market regime, and the mix of account sizes being sold.

Two minimum requirements: the window ("January–October 2026"), and a stated treatment of accounts still in progress. Excluding them inflates the rate. Counting them as failures deflates it. Either is defensible if it is written down; silence is not.

4. Make the claim checkable by someone who does not trust you

A number in a headline cannot be verified. What can be verified is an artefact a firm will confirm:

• A certificate per pass, with the client name redacted and an account number the firm can look up.
• A checked date, so a reader knows which version of the firm's rules applied.
• The failures, listed with their cause, in the same place as the passes.

That last item is the one that separates a record from an advertisement, and it is also the one that is uncomfortable to publish, which is why it is informative.

5. Prefer the wording that cannot be read as a promise

"87% success rate" implies a probability about your evaluation. It is not one — the outcome depends on the account size, the firm, the rules in force that month, and the trader. What a service can actually claim is a historical count over a defined population.

The defensible sentence has four parts: count, denominator, window, treatment of resets. For example — "63 of 71 purchases reached funded status between January and October 2026, including nine resets." Every word there is a fact about the past, and a reader can ask to see the 71.

The disclosure

I run a passing service, ElitePropX, so the standard above is the one I am applying to myself. The passed challenges page lists the accounts that reached funded status with client names redacted, and the evaluations that were reset alongside them — roughly one in twenty. The methodology page states what a checked date means and how a correction is published when a firm changes a rule.

Whether the numbers on those pages are good is a separate question from whether they are legible. Legible first: a reader who can count the denominator can decide for themselves, and that is the only kind of claim worth making in a market where every competitor has a nicer percentage.

The one-line test

Before you compare two services, ask both: "your rate — per purchase or per client, what window, and are resets counted?"

The service that answers in one sentence is not necessarily better. It is the only one whose number you can actually use.