For many subjects you can record metrics that can be analyzed over time and then use statistical techniques to predict future values for those metrics.
The question then becomes inferring what may / will happen if the metrics hit their predicted values.
Then you have second and third order effects that come from the predicted first order effects. That is, if metric X hits the predicted value and Y occurs as a result, what will party A do as a result? or how will Z be impacted?
>> Do you have any ideas on how to train this kind of thinking so that it becomes second nature?
Practice. You must track and analyze the metrics data and practice making predictions and doing analysis.
It is very subject-dependant. Predicting the weather is quite different from predicting stock markets which is quite different from predicting geo-politics.
Whatever decision you’re facing - build a decision tree as far as it makes sense and actually depends on your input. By asking a tree of “what if” questions or something like that.
In a nutshell it’s a risk management