While we cannot predict as per command $3, we can foresee the future. For this we use scenario planning as per command $4. Yet, we have to give each scenario a probability. For if a scenario were deemed to be impossible, we would not entertain that scenario. So if you ever think: this stock can only go up, you are acting too stupid and you violate command $9. You are perfectly free to decide that the probability of a stock dropping is 1% and of it going sideways is also 1% and that the probability of the stock going up is 98%. That is perfectly fine as long as you also date these scenarios.
So let’s take that as an example. Say we think that the probability of stock A going up at least 1% next week is 98% with sideways and down each 1%. Besides planning what to do if this is indeed the case per command $2, only react and never predict, we now can also measure the gap between your prevision and reality. You can use Brier’s Rule for this. Basically you get only 4 penalty points if you are right, but 9801 penalty points if you are wrong. If you rake up a lot of penalty points, you are probably using a bad model or no model at all.
If you are unhappy with your results, you can either create a better model to foresee the future, or change your trading strategy. Be careful though with trying to create a better model. While model fitting, i.e. updating your model as to better foresee what just happened, is perfectly fine in the early stages of modelling, once you are settled on a model, more model fitting will actually make things worse.