The ideal WvW world would have even player representation from all timezones to allow for structure upgrading, smooth transitions of command and to eliminate periods of time where the numbers are so low that all assets are lost. In other words the ideal world has even coverage.
Most discussion about coverage seems to revolve around recruiting more guilds but rarely do these new guilds neatly fill in a hole in the coverage map. The result is spikes in activity during certain periods which results in massive zergs, queues and lag issues.
One possible solution to this problem is to completely rethink world occupancy by increasing the granularity of the transfer mechanism.
Currently, each world has a population level which results in it being assigned a population rating (e.g. ‘very high’ or ‘high’). Transfer to the highest-population servers is not possible unless a ‘blackout’ is arranged, usually in the form of people being encouraged to log out for an hour so that the population levels are artificially lowered which, in turn, creates a window for some guild to transfer players onto the server.
Instead of the total number of players logged in last hour (or whatever time period is actually used) being the metric limiting transfers, instead:
1) track the amount of WXP earned by each player on an hourly basis (similar to the way guild activity is tracked on an hourly basis)
2) once a week, crunch those numbers to profile the mean times-of-day/days-of-week that the player plays (24*7=168 identifiable periods)
3) track average WvW queue lengths for each world on an hourly basis and aggregate to build a similar weekly profile of congestion
4) if a player wants to transfer to another world, compare the player’s WvW activity profile to the world’s congestion profile and
4a) if the two do not overlap allow the player to instantly transfer to the new world for free
4b) if the two do overlap then determine the degree of overlap and determine a cost accordingly (e.g. 10 gold or 1 gem per hour of overlap)
That’s it.
Such a system would make it easy for players to move to fill in holes in coverage maps, and penalise players (possibly quite severely) if they desire to relocate to a world which is already congested and has queues during their demonstrated play times.
Example 1: A player plays 1700-2200 Mon-Fri and 1200-1600 Sat-Sun. The desired world experiences congestion during 1200, 1300 and 1400 on Saturday only. Three hours of overlap means that the cost to transfer to the world would be (3h*1d*10g=) 30 gold.
Example 2: The same player thinks of transferring to a very popular server with congestion periods 1500-2400 every day of the week. Overlaps for Mon-Fri are 1700, 1800, 1900, 2000, 2100 and 2200 but no overlaps for the weekend. Cost to transfer would be (6h*5d*10g=) 300 gold. He decides against it.
(Of course the cost/hour would need to be fine-tuned and 10g was an arbitrary amount chosen for illustrative purposes only.)
This is not intended as a way for existing guilds, in fixed timezones, to transfer between worlds.
Guilds form because players, on the same world, playing together at the same time, organise themselves in a formal fashion. By having a transfer mechanism which preferentially directs players to fill coverage holes, new guilds will form where none existed before.
Thoughts?