4gravitons make an interesting analysis of four different kinds of experiments and their respective likelihoods of discovering something interesting.
Some experiments are almost sure to discover something new because we know nothing about something new except that it is out there to measure. Some experiments are strongly expected to reproduce the status quo and are just dotting i's and crossing t's to more fully establish the range of the status quo with the off chance that an unexpected anomaly is discovered.
In between are theory testing experiments. These can be powerful or merely paradigm confirming, unless a theory can be easily modified to adapt to any set of results, or there are many similar theories, not all of which can be falsified simultaneously.