Hi Michael!

Agree that what is troublesome is when estimates become commitments. In situations where forecasts can be done quickly, cheaply and provide value, estimates are useful off course. If the tools is used correctly. I believe some environments/situations are better suited for the estimate tool than others. Especially when you have fixed time projects.

In the end estimates are a tool to reduce risk.

I still believe that when things are complicated or predictable estimates can be a successful tool. when used appropriately. As you write.

I also believe that in complex environments, where predictability is low, a better way to reduce risk could be to not make estimates.

A better approach could be to make a hypothesis, start working on the hypothesis. Have fast feedback loops. Use a fixed period of time. Use fixed cost instead.

The period should be set from how much money you are willing to invest before you need to take a decision to continue. If value progression isn't according to expectations then abort. If you believe you can have a ROI invest more money.¨

The only estimate you need to do is to ask the team if they think anything of value can be completed in the fixed period. If not you don't even start. Note the estimate is about whether test the hypothesis can be tested successfully in the period , not the content.