I'm interested in examining the efficiency of the scientific process. Part of this involves examining what actually goes into making a discovery. One common objection is:
The implication is that these things can't be analysed and we should just throw money and minds into a pot and let them do science.
I have a suspicion that due to the availability bias we tend to discuss surprise discoveries more and therefore overestimate their importance.
Has anyone published data on how frequently unexpected (chance) discoveries actually occur? Ideally, it would include a weighting on how important the discovery was.