I ran into this when trying to duplicate the behavior of the WCA website for our Speedsolving.com competition website:
Does anyone else agree with me that it seems odd that we consider an average to be tied if it is the same after rounding? I think it's strange that, for instance, a 12.333 average and a 12.327 average are considered identical (both round to 12.33), and therefore the tie is broken based on the lowest single, but in the case of comparing a 12.327 average (rounds to 12.33) and a 12.323 average (rounds to 12.32), the 12.323 average would be considered better without considering the single. I suppose all of this is somewhat arbitrary, but to me this seems particularly so. Anyone else agree?
Please let me know if I'm wrong about this. I ran through the calculations for several cases both in the event ranking pages and in the individual competition ranking pages, and it really seemed to me like this is how it is handled in both cases. I just found it surprising.