The Oshi Score is a score out of 100 computed from your real trades. It is the average of six performance indicators, each brought onto the same scale. It is not a discipline score: that one exists, it has another name, and it measures something entirely different.
Contents
The Oshi Score answers a simple question: where do your results stand, on a single scale? Six classic trading indicators are each brought onto 0 to 100 by a fixed rule, and the overall score is the average of those that can be computed.
The normalisation rules are ours and they are deliberately strict in the middle of the scale. A profit factor of 1, meaning break-even, scores 0; it takes 2 to reach 100. An average win to average loss ratio of 1 scores 50. A recovery factor of 3 scores 100. A drawdown of 0% scores 100 and a drawdown of 50% scores 0. Consistency compares your best winning day to the total of your winning days: if a single day makes the whole gain, that axis falls to zero.
And it has to be said straight away what this score does not do, because its name suggests otherwise: it looks at none of your rules. Not your time windows, not your daily loss limit, not the presence of a stop. All of that is measured, but elsewhere, by the discipline score.
The overall number is a mean, and a mean of six axes behaves exactly as you would fear: an axis at 10 vanishes behind five axes at 70. The useful reading is therefore never the score, it is the radar beside it, where the slipping axis shows at a glance. The detail shows the real metric next to each axis, not its normalised version, so you can check what the score did with your figure.
A score that barely moves while your P&L swings wildly is a signal in itself. The six axes are ratios, not amounts: they do not change when you double your position size. An account whose result takes off without the score moving has won bigger, not better, and the two do not age the same way.
The classic mistake. Comparing it to another trader's. The thresholds turning a profit factor or a drawdown into a score out of 100 are ours, and they are the same for everyone while styles are not. A scalper and a swing trader have no reason to share the same distribution across these six axes. It measures you against yourself last month, never a ranking.
It measures your results, on a single scale from 0 to 100. It is the average of six normalised classic indicators: win rate, profit factor, average win to average loss, recovery factor, drawdown control and consistency of gains. Everything is computed from your real trades.
No, and that is the most common confusion about it. None of its six axes looks at your rules: not your time windows, not your daily loss limit, not the presence of a stop. Discipline has its own score, the discipline score, which measures nothing else.
The question does not arise in absolute terms, because the normalisation thresholds are ours and they are identical across styles. What can be read is your score against your score a month ago, and above all the radar axis that is slipping. An axis at 10 behind five axes at 70 gives a reassuring mean and a real problem.
Because the six axes are ratios, not amounts. Doubling your position size doubles the result and leaves win rate, profit factor and the win/loss ratio exactly where they were. The score is then telling the truth: you won bigger, not better.
No. It is computed from your trades by a fixed, documented formula, and it consumes no AI credits. That is true of every statistic in this glossary: AI is there to comment and analyse, never to produce a figure.
The Oshi Score is a Dashboard card you can add, move or remove like the others. It shows three things: the gauge for the overall score, the radar of the six axes, and below the radar the axis-by-axis detail with the real metric next to its score, for example 'Profit factor 1.80' facing its 80.
It also appears in the recap that follows an import, which often makes it the first figure a new account sees about itself. While there are not enough trades, the card says so instead of showing a score built on nothing.
To go deeper on this: read the full guide. Every term is gathered in the glossary.