The mechanics change during a release

Your analysis might be right and you can still lose, because the market you are trading is not the market you backtested in. Around a high-impact release:

Spreads widen several-fold, sometimes for a minute or more. A trade that was cheap becomes expensive instantly.
Slippage increases — market orders fill far from the displayed price.
Stops are not guaranteed — in a gap, your stop becomes a market order at the next available price. Your stop is not a limit on loss unless it is a guaranteed stop.
Margin requirements often rise — some brokers widen margin on CFDs around major events.

Three legitimate approaches

Stand down — flatten or avoid initiating in the window around the release. This is what most professionals do, and it costs you nothing but optionality.
Reduce size — if you must hold through, cut risk to a fraction of normal so that a gap cannot do serious damage.
Trade the aftermath — wait for spreads to normalise and trade the follow-through. You give up the first move and keep sane execution.

Sizing into an event

The arithmetic is unforgiving. If a release can plausibly gap your stop by three times its normal distance, then trading your usual size means your actual risk is three times your intended risk. Either cut the size to compensate, or accept that you are running three times the risk and call it what it is.

Widen stops or cut size — never both

Widening the stop to survive the noise while keeping the same size silently multiplies your risk. If you need a wider stop, you must take a smaller position. This is exactly the same discipline as Part 3 of the MQL5 course: cut the size rather than stretch the stop.

The honest caveat

Reducing size around events lowers both your risk and your expected return. There is no free lunch — the volatility you are avoiding is also where the opportunity is. Decide deliberately rather than by default.

How to check this yourself

Take your normal position size and your normal stop. Now assume the spread triples and your stop gaps by twice its distance — which is what a release does. Recompute the loss at that size. Then work backwards: what size keeps the loss at your intended risk? Write it down before the event, not during it.