I agree with Bestbard's suggestion. If you spent time in one run preparing the exploit for the next run, that previous run should count towards the time you took to complete the new run. Sounds perfectly fair and reasonable.