The check-in is a short status you record before or after your session, typed or spoken. Tradoshi turns it into a dated state attached to the period's trades, which becomes an analysis dimension like any other. It is not a diary, it is a variable.
Contents
A check-in is a short, dated note on your state at that moment. You can type it or dictate it, and in both cases what is read is the content: the words you use, not the sound of your voice.
What the app does with it is simple, and that is where the value sits: it extracts a state, dates it, and attaches it to the period's trades. From there the state is a column like the market or the hour, and P&L can be sorted by state.
The only question this data exists to answer is this one: do your trades taken in a certain state lose more often? Fifty observations are usually enough to settle it, and the answer is always personal. It is rarely the one you would have guessed.
The rule that decides everything is regularity of entry, not subtlety of wording. A check-in filled only after a bad session produces a variable recorded in one case only, and the correlation you think you see there is just collection bias. Three lines every day beat a page once a week by a very long way.
Then keep a short, stable vocabulary. The state exists to make groups, and fifteen nuances produce fifteen groups of three trades on which no gap is interpretable. Three or four recurring states are worth more than a fine taxonomy.
The classic mistake. Filling it only on bad days. A variable recorded in one case only compares to nothing, and the crossing will then show that your bad states coincide with your bad results, which proves only the order in which you decided to write. The ordinary day is the one that gives all the others their value.
It is a short, dated note about your state, taken before or after a session. It is read to extract a state attached to the period's trades, and that state then becomes an analysis dimension: P&L can be sorted by state just as it is sorted by market or hour.
Whichever you prefer, the result is the same. In both cases it is the content that is read, meaning the words you use. Dictating saves time in the morning, and that is its only advantage.
Three lines are enough, and are even preferable. What makes this data valuable is not the length of the text but the regularity of entry: a short note taken every day produces a usable variable, a weekly page produces none.
Around fifty, provided they also cover ordinary days. Below that, a state declared three times coinciding with three losses says nothing at all, and it is very easy to read a rule into it. The number of observations matters less than how they are spread.
No. What is read is the content of what you say or write, the words themselves. Dictation is there to capture faster, not to make anything beyond the words audible.
The check-in happens in the Daily check-in screen, and two ways of reading it coexist. Oshi Flash spots the keywords of your state right in your browser, with no AI: instant, free and unlimited. Oshi Analyse has an AI read the content, drawing a finer reading and a recommendation from it, and consumes AI credits per minute.
The resulting state then shows up in the Psychology screen, which compares your rule adherence on the days your check-in is steady against the days it is degraded. That crossing is what justifies the entry, and the measured gap between those two columns is what says when to cut risk.
To go deeper on this: read the full guide. Every term is gathered in the glossary.