There's a weak correlation between later years and higher correct percentages (r=.43), but years after 1000 are definitely better known (post-1000 years have a mean correct percentage of 55, vs 37 for pre-1000). I suspect that knowledge of history between about the time of Constantine and 1066 is generally pretty weak.
The regexes were a bit tough to get right without writing a script. AI was useless.
It split quickly, reforming as the HRE by Otto in 962 on one side, and the Kingdom of France for the other part.