The trouble with error is that we have the tendency to dismiss it. When Kevin Dunbar analyzed the data from his in vivo studies of microbiology labs, one of his most remarkable findings was just how many experiments produced results that were generally unexpected. More than half the data collected by the researchers deviated significantly from what they had predicted they would find. Dunbar found that the scientists tended to treat these surprising outcomes as the result of flaws in their experimental method: some kind of contamination of the original tissue perhaps, or a mechanical malfunction, or an error at the data-processing phase. They assumed the result was noise, not signal. [Kindle location 1658]
In other words, many projects don't deliver the expected results; i.e., they are mistakes. And yet a lot of good comes from this. Johnson writes that there is much more "signal" in these mistakes than we expect, and we tend to overlook them or attribute them to our own errors, not some other factors that may be meaningful. This is yet another reason to revisit mistakes for learning. Yes, you will find things you did wrong, but you may also discover something entirely new.