You will be sometimes wrong and sometimes right. So don’t punish yourself when you are not sure.
Always be transparent over how confident you are in your analysis and what will be the consequences if you are wrong.
Try to always focus on the practical implications of what you are analyzing.
Always keep in mind what are you trying to prove/disprove. If you to dismiss a business idea for example, you do not need to analyze every possible aspect of you can rule out the opportunity based on napkin calculations.
I worked with pricing ML models for some time, and if a problem has a 40% irreducible error rate you shouldn’t account for uncertainties that sum up to 0.3% of it.
Try to find similar but easier problems.
Analysis is just an activity, and is not the whole job.
Always be transparent over how confident you are in your analysis and what will be the consequences if you are wrong.
Try to always focus on the practical implications of what you are analyzing.
Always keep in mind what are you trying to prove/disprove. If you to dismiss a business idea for example, you do not need to analyze every possible aspect of you can rule out the opportunity based on napkin calculations.
I worked with pricing ML models for some time, and if a problem has a 40% irreducible error rate you shouldn’t account for uncertainties that sum up to 0.3% of it.
Try to find similar but easier problems.
Analysis is just an activity, and is not the whole job.