Which Jeremy was trying to fix.Yes, I hadn't noticed that one. On CCT it gives you and average xx solves of yy attempts where the DNFs are left out.
It should be a good representation of a typical time for the session, so some good options are:
- DNF-dependent trimmed average (where the number of trimmed times from both ends equals the number of DNFs)
- Average of your middle 50% of solves.