I feel like this form of estimation could be overvaluing estimation. It's an ESTIMATE, and it should be ok if you are wrong to some degree. There is a continuu

I feel like this form of estimation could be overvaluing estimation. It's an ESTIMATE, and it should be ok if you are wrong to some degree. There is a continuum of estimate accuracy. If you spend 2 minutes on an estimate you will have a wide error band. The more time you spend, the more accurate you can get. Let's say there is a task that will take 10 hours. If you spend 2 minutes estimating, you can probably tell it will take between 5 and 15 hours. You could conversely spend 2 hours estimating and conclude that it will take between 8 and 12 hours. But is that improved accuracy worth the time investment? You should balance accuracy against the value you are getting from it. What decisions are being made from these estimates? What is the cost if you are wrong? YMMV, but generally speaking in my opinion, if your team is spending 4 hours in a room each sprint "estimating", I think you are wasting time.