Try this learning loop

State a prediction before changing the controls. Identify which assumption changes the result, reset and test a second case. These deliberately small models explain one mechanism at a time.

Continue learning