Here’s a fun article on PMI.org. By omission, it gives some very bad advice about estimation. What is it missing? Asking the people who are going to do the work!!! Any estimation method or approach that fails to ask the actual human beings who are going to do the work about the effort required is going to be badly wrong. (Of course, even asking the people who are going to do the work is no guarantee of good estimates.)
The starting point of the linked article is a study that showed 90% of projects having cost overruns. To me, this just shows the futility of predicting the future, not anything about how we can (and should?) make better estimates.