Which experiment is more likely to be reliable, and why?

Since there's nothing here other than what you've asked i'll guess, an experiment that is more likely to be reliable will not use the first values it gets, will be done multiple times and the values will be close in number