Trading psychology, with numbers attached

Advice about trading psychology is free everywhere and almost never actionable, because it describes feelings and your file records behaviour. Every state below has a signature you can measure.

The problem with psychology advice

Almost everything written on this subject follows the same shape: a description of a feeling, followed by advice to be more disciplined. Both halves are useless in practice, and for the same reason — you cannot check either against anything.

Read a description of revenge trading and you will recognise yourself, because everyone does. Read the advice — take a break, follow your plan — and there is nothing to do differently tomorrow, because you already agreed with it yesterday.

The missing piece is evidence. Not "do you feel like you overtrade" but "did your high-volume days actually finish worse than your quiet ones, and by how much".

What changes when a state has a signature

Each page here describes a state twice: how it works, and what it leaves in your file.

The traces are ordinary data. Trade counts per day. The gap between one exit and the next entry. Position size relative to your own average. Whether a loss was followed by a bigger trade or a smaller one. None of it requires anything beyond the export your broker already gives you.

That turns an unfalsifiable conversation into three possible outcomes:

  • The pattern is there and it is expensive — now you have a size, and it is worth working on.
  • The pattern is there and it costs almost nothing — worth knowing, so you can stop spending attention on it.
  • The pattern is not there — you have ruled something out, which is progress that advice can never give you.

The third outcome is the one people never get from reading about psychology, and it is often the most useful.

What measurement cannot do

Being precise about the limit, because this section would otherwise become the thing it argues against.

A file records what you did, never why. Trading more after a loss and trading more because the setups appeared look identical in an export. A large position can be conviction or it can be tilt. The data narrows the question and hands it back to you; it does not answer it.

There is also a hard boundary at the edge of the file: whether you moved a stop after entry is not in there, and it is the single most common way a plan breaks. See what your statement hides.

Where to start

If you have an export handy, run it through the discipline check first. It applies six of these signatures at once and tells you which are present in your history — which is a faster route to the page worth reading than choosing one by how familiar it sounds.

In this section