Why this is not a score out of 100
Composite scores are the standard product in this corner of the market. Take win rate, drawdown, average hold time and a few others, multiply each by a weight, present the total as 73/100.
The problem is not the arithmetic. It is that nobody derived the weights. They were chosen so the output would feel reasonable. A number built that way cannot fall for a reason you can name or rise from an action you can take, and two traders with identical scores may have nothing in common.
So this page returns something less impressive and more usable: six independent checks, each with its own threshold printed next to its result. If you think one of them is wrong for how you trade, ignore that line. The others still stand, which is not true of a weighted total.
The six checks
Position size stays consistent. How much your trade sizes vary relative to their average. Sudden size is the most common way a good month is undone, and it shows up in a file without any interpretation.
No blow-out days. Your busiest day against your typical day. A day at three times your median trade count is usually not a plan.
You do not speed up after a loss. Your typical pause between trades, compared with the pause specifically after a losing one. Speeding up after a loss is the mechanical signature of trying to win it back.
Your worst loss is not an outlier. The largest single loss against the typical one. A loss several times the usual size is not bad luck; it is a position that was too large or a stop that was not respected.
A stop was actually set. What share of trades carried a stop-loss at all.
The stop held when it was tested. Of the trades that moved against you, how many finished past the level you had set.
Unscored is not failed
The last two checks need the stop-loss column, which only MetaTrader exports carry. Other brokers drop it.
When the data is absent, those rows say not scored — and the denominator shrinks with them, so a file without stops is judged on four checks rather than punished on six. Counting a missing column as a failure would mark someone down for their broker's export format, which has nothing to do with how they trade.
What passing actually means
A file records what happened. It does not record what you meant to happen, and it has no idea whether you followed your own rules or simply had a quiet month in which the rules were never tested.
So a pass means the behaviour is absent from this export — not that the intention behind it was right. Someone who plans badly but executes their bad plan consistently will pass most of these.
And the single habit most worth catching leaves no trace at all here: moving a stop after entry. A closed statement stores the final stop level, so a widened stop looks identical to a respected one. If you know you do that, no page reading an exported file will tell you about it, this one included.
Frequently asked
Why are the thresholds round numbers?
Because they are conventions, not findings, and pretending otherwise would be the same dishonesty as the weighted score. Three times your median day, four times your typical loss, 90% of trades with a stop — each is a readable line in the sand, printed beside the result so you can judge whether it fits your style rather than trusting it.
I failed "position size stays consistent" but I scale in deliberately.
Then that check does not describe you, and the page is built so that you can drop it without the other five collapsing. Deliberate scaling and impulsive sizing look the same in an export; only you know which one it was.
Does it work with fewer than ten trades?
No, and it says so rather than scoring anyway. Under ten closed trades you are describing a week, not a way of working.
Are costs included?
Yes. Every P&L figure behind these checks is net of commission and swap.