Dynamic Population-Based Scoring Evaluation Frequency
First, nightcapping is not the fault of the players. They're just playing at a time they and their friends can get together. It's also not anyone's fault that there is an EU server which dramatically increases the scoring potency of nightcapping because it creates a window when anyone who would want to play, would play on the EU server rather than NA. I play with a lot of folks at many times of the day (including primetime EST and AEST) and everyone I met just wanted a fun time in PvP.
-Scoring Evaluation frequency is tied to current population
-Population is determined by the majority of factions (2 factions at 1 bar means the next scoring evaluation will be based on 1 bar population)
-Frequency for each population level should be determined by the people with the metrics to know how it affects the campaign, aka ZOS.
--Example: If the server average is poplocked at 450 players for 6 hours a day, that is 2700 player hours. If it's at 1 bar 12 hours a day, then that is 1800 player hours. ((50+50+50) * 12). For easy math, let's say the last 6 hours going into and leaving primetime of 2-3 bars is a total of 1800 player hours ((100+100+100) * 6). Scoring frequency could be 9 per hour in poplocked, 6 per hour in 2/3 bars, and 3 times per hour in 1 bar times.
This rewards map swings during the hardest time to change the map, but not so much that those playing in off peak times don't matter. In the above example, off peak players directly contribute to 28% of the total score in the lowest population times and indirectly by starting off in the best place during the mid-population times.
This scoring can be leveraged to help the BIGGEST COMPLAINT OF CYRODIIL, LAG. Specifically, lag caused when people don't spread out. If the scoring evaluation is at a high frequency, other smaller objectives can be magnified.
--Example: Reduce what a keep is worth in scoring if it doesn't hold its resources. 100% earned if all 3 resources are held, 80% for 2, 60% for 1, 40% for not owning any resources. So not protecting resources during poplocked times will cost a lot of scoring points, and with a scoring interval of 6.5 minutes, small groups can spread out and change the scoring. This is just an example of how a more frequent population-based scoring evaluation can make otherwise unneeded parts of the map really matter and not meant as the main theme of this post.