Guild Wars 2

We all know victory in WvW is defined by coverage more than skill.

- Want matches where coverage (including oceanics) is never a factor in who wins or loses?
- want WvW to be about skill of its players and how a server chooses to deploy its forces within a map?

There are huge problems in WvW at the moment which have their root cause in server coverage imbalances. The problems manifest themselves differently between the top and bottom tiers. The current system is killing the lower servers, and the masses at the top are trapped with lag/queues because they cant go anywhere else to get the most competitive matches.

Many low-tier servers are getting destroyed by this system, trapped in a spiral where guilds transfer out and up with no incentive to get adds, and I am worried about the long term health of WvW if half the servers continue to suffer these problems.

This post will present a solution to permanently resolve the population imbalance problem. This will achieve all of the above. Sorry if it has been discussed before. Below is a pic and links to spreadsheets. I have thought about this for a while, but with all the talk of leagues and whatnot, its clear that ANET is looking in the wrong place to solve the problem – hence this post.

The essence of this solution is to constantly monitor the number of players each server has on a particular map. this calculation is constantly updating, and its purpose is to 1) identify the server with the fewest players on a map, 2) define a cap for all servers on the map based on that number, and 3) prevent the other 2 servers from having significantly more than the lowest server. The excess is placed into the queue.

If this algorithm is added we can have even matches every week. But I dont know all the side-effects of doing this. I have analyzed most of them and always come to the conclusion that this is better than the system we have now. The worst side-effect is matches in the lower matchups can have queues where they never had them before. Do I want queues on my queue-free server? skritt-no, but giving and receiving blowouts isnt fun either. If ANET supports the system with transfers (see below), this will go away in time.

There is an offset value (5 in the example). ANET would have to calculate the actual value. This number is defined by the maximum rate-of-change in server population

Notice that this runs real-time so it will balance off-hours as well, servers wont lose because of coverage gaps. Nighttime wins will be decided by even player numbers just like daytime.

one option not in the spreadsheet is to set a lower limit on the player cap. It could be set to not go below (lets say) 20. that way if one server has zero people on, the other servers could have at least 20 each. obviously I dont see that happening, but it would prevent a server from calling a strike.

I also uploaded spreadsheet that shows how it would work if you used the lower 2 servers to set the cap (option 2 below). Its the halfway solution which wouldnt fix the problem (since coverage would still trump skill) but it would prevent the situation where one server blows out the other 2 solely because of numbers.

An obvious by-product of this is more queues for imbalanced matches. So it is critical that if ANET uses this approach that they enable free transfers to the servers which are repeatedly causing the pop-cap. one solution would be to enable free transfers to any destination server ranked lower than the source server until a new steady-state is achieved. Once in steady state ANET would always have to identify which server is the limiting factor and allocate a finite number of free transfers to that server. To ANET: if you cant do this part then don’t bother with this system, it wont work.

So Im tossing this out for conversation. Will i like it if its done, I dont know. Anet leaves many things half-implemented so I have my worries, but if we want a game of skill than we have to change things because right now skill is totally pointless when you match outside your “tier”.

Disclaimer: I know some of you have no reading comprehension skills and have been programmed by your baby-boomer parents to CONSTANTLY complain. hence the included attachments. if you are going to reply with a complaint at least read this post in its entirety and review the spreadsheet so you have a basic understanding how and why this works the way it does. You can download the file using the link below so you can try it for yourself.

the spreadsheet is from excel 2003. use the whole line in the link – first time i ever used this service, the web download lasts 7 days.
http://wikisend.com/download/350072/wvw balance example.xls

second spreadsheet showing the highest-of-the-smaller-2-servers option
http://wikisend.com/download/375626/wvw balance example option2.xls

So… What happens if a server has zero players on a map? Only 5 people are allowed on that map from either of the other two servers?

What about when people start to log off at night? Under this system is the server going to boot players to keep it even?

Don’t get me wrong, its a great idea, but it’s far from the ideal solution. You’re assuming that players from every server always have people on a map. Not to mention players who are “doing their own thing” on a map and not working toward progressing their server’s score.

I’m not sure why this suggestion doesn’t get more traction. I essentially posted this same concept almost a month ago and only go two responses, both being more or less positive. https://forum-en.gw2archive.eu/forum/wuv/wuv/Plus-Ten/first

Someone please point out the downside to this.

@Clockradio,
There would be a minimum starting threshold, so if the lowest server has zero, the others can still have up to X players; once the lowest server reaches the threshold is when the queue rules kick in.

To your other point, I think typically peaks times and off times are similar among the competing servers. (right now rankings are based on coverage, so that makes this work even better) People will log in and out roughly around the same time periods, keeping queues low. If it does become unbalanced because a large zerg decided to leave the map, the system doesn’t boot players off the now higher side, but rather queues them from getting even more and allows the lower map population catch up, thus creating a fair fight again in a short amount of time.

Anyway, great post Plains of Ashes. I hope more people look into this option.

Looks interesting. But are you saying open up free transfers on demand? For a time period of how ever long the player wants to play in that match?

Take SoS and IoJ for example. They have an out of proportion Ocn/Sea population compared to the rest of their timezones and compared to their closest competition. So for example, it’s Ocn prime time, the following are map pop:

SoS: 100
CD: 50
EBay: 30

And if there are players in queue from BG, they will be able to see this and say, hey, there are open slots on EBay, I can go play in there until my queue on BG pops? Or even players in the SoS queue can say, hey, there are open spots in the IoJ/GoM/DH match, I’ll go to GoM for awhile? Or players not even in a queue, but say, in a blowout match like TC/Mag/FA/ who want to have actual fights rather than PvD?

Now that would be cool.

As you say, it would only work if Anet was able to implement free, instantaneous on-demand transfers to those servers facing the high pop server at the time. It would also depend on being able to communicate the open slots available to those waiting in queue. And it would end up depending on whether players would want to transfer to assist an undermanned server.

@Johje Holan
I don’t think the OP was implying what you described. I believe he was suggesting is that if you were on a server that was over populated, due to people transfer to a winning server, that you could transfer to a lower ranked server for free.

The free transfer threshold being set by ArenaNet and would probably around the lower third of servers (because ranks are population based now). This would be needed because in this system the queue’s would be longer for those over populated servers since the competing server couldn’t keep up with their numbers. (which is the problem in the first place).

Once people realized they could go to a “lower tier” server and be competitive (because the maps would have a near equal player base at all times), they would transfer and thus eventually create nearly equal queue times on all servers.

Also, in your example, the map numbers of 100, 50, 30 would never be that far apart in this system. In an extreme case it would be more like A-50, B-40, C-30. (and that could only happen if server C had 15 people leave all at once [in the OP’s plus 5 example])

Johje Holan.4607:

Looks interesting. But are you saying open up free transfers on demand? For a time period of how ever long the player wants to play in that match?

Take SoS and IoJ for example. They have an out of proportion Ocn/Sea population compared to the rest of their timezones and compared to their closest competition. So for example, it’s Ocn prime time, the following are map pop:

SoS: 100
CD: 50
EBay: 30

And if there are players in queue from BG, they will be able to see this and say, hey, there are open slots on EBay, I can go play in there until my queue on BG pops? Or even players in the SoS queue can say, hey, there are open spots in the IoJ/GoM/DH match, I’ll go to GoM for awhile? Or players not even in a queue, but say, in a blowout match like TC/Mag/FA/ who want to have actual fights rather than PvD?

Now that would be cool.

As you say, it would only work if Anet was able to implement free, instantaneous on-demand transfers to those servers facing the high pop server at the time. It would also depend on being able to communicate the open slots available to those waiting in queue. And it would end up depending on whether players would want to transfer to assist an undermanned server.

That goes against all of what Anet are intending, the whole purpose is so that you can fight for YOUR server through thick and thin and establish server loyalty (doesn’t mean much atm though).