The principle is the often cited 80/20 rule, where 20% of “x” controls 80% of “x”. Or 80% of the problems are caused by 20% of the people. Or 80% of the software errors are caused by 20% of the code.
Pareto optimization is about maximizing the outcomes of a system, where you cannot make any change that somehow doesn’t cause a worse outcome in some way. You’ve maximized all the parameters such that changing one can make it worse elsewhere.
For example, increasing an advertising budget may increase sales, but it will decrease the budget for engineering so less units will be produced for sale. Or you can only decrease quality so much that everyone stops purchasing.
Latest Answers