forex-scalping.com

It is tempting to risk more on setups that look perfect and less on those that look marginal. The problem is that our confidence is poorly calibrated. Setups that feel perfect often have recent wins fresh in memory or a strong narrative, but their actual win rate may not differ much from the others. If you risk three percent on a feeling and the trade fails, a single event can erase weeks of progress. I use a fixed fractional risk on every trade, such as one percent, and I let the strategy's edge play out over many trades. If I want to differentiate between setups, I use objective criteria written in advance, such as the number of confirmed conditions, and I test whether the higher condition group actually performs better. Only if the data supports it would I adjust size, and then modestly. Consistency in sizing also makes it easier to evaluate performance, because the results are comparable across trades. Variable sizing based on mood muddles every statistic you try to compute later.

Fixed risk is the safer approach. Confidence is not a reliable guide, as you say. If a setup really performs better, the data will show it over time, and then a small adjustment based on objective criteria makes sense.