WvW Structures & System Plus Populations Issues — Guild Wars 2 Forums

WvW Structures & System Plus Populations Issues

SkyShroud.2865SkyShroud.2865 Member ✭✭✭✭
edited January 26, 2018 in WvW

This is plainly a educational post about how the WvW is structured and designed plus the populations issues accompanying with it currently. The purpose is to educate players about the current WvW systems and help facilitate related decisions making. The explanations given are based on my personal understandings resulted from observations and reading related sources from either the devs themselves or logical assumptions from fellow players over the years. If there are any portions that you feel is inaccurate or can be misinterpreted or left out, feel free to point them out.

This is not a population balance suggestion thread so please post your population balance suggestion(s) on thread(s) of your own.


Abstract

Current WvW is a server based design in which each server has their own population base made up of Pvers, PvPers & WvWers while the server population fixed threshold is based on WvW itself. The current WvW design and past events contributed to populations disparities between servers.

History (Timeline)

This is a timeline base on anet's decisions regards to servers or WvW. Hopefully from this timeline you will see that the current WvW issues are partly because of the decisions of the past.

2012 August 28 - Gw2 Launch, all servers are free to transfer with weekly restriction
2013 January 16 - Announcement of Guesting feature and ending of free transfers
2013 January 28 - Guesting feature implemented, free transfers ended
2014 March 18 - Second WvW tournament pre-sales, discounted and free transfers to selected servers
2014 March 28 - Start of Second WvW tournament and end of pre-sales
2015 July 23 - Announcement on server's population threshold to be base on WvW
2015 July 28 - Server's population threshold is now base on WvW
2016 April 22 - Start of World Linking beta
2016 September 06 - Announcement of Skirmish Score System
2016 September 09 - Implemented Skirmish Score System
2017 February 24 - New algorithm for server's population threshold implemented
2017 May 31 - Announcement of Competitive Pack (Reward Patch)
2017 June 06 - Implemented Reward Patch

Before 2015 July 28, there were observations that selected servers have their populations threshold increased.

Current WvW Server Structures

Despite having megaserver, it doesn't change how the server is structured. Players still have to choose a server and therefore belong to a specific server, server's threshold is based on WvW population itself thus WvW is now currently structured in this manner

When we put different servers side by side with each other, we can have something like this

Just by looking at the images, we can easily determine the obvious issues.

Base populations

Base populations are all the players in that server, regardless if is PvErs or WvWers or PvPers or inactive or even bots. According to the timeline, we can see that there were a lot of chances where populations can move around and get to uber stack on any servers. All the past events resulted in a disparity between servers' base populations and since WvW populations is correlated to the server populations, it will affect the populations balance.

Fixed Threshold

The fixed threshold is universally same for all servers which theoretically sounds fair on the paper but practically speaking it isn't. A threshold isn't a hard cap, it is simply a indicator on when to open and when to close a server. Therefore, it is possible for any servers to exceed that threshold by large amount which can become population imbalance issue to servers that exceed the threshold by small amount. This can happen because of the base populations. There are servers with huge base populations where many initially don't play WvW but that doesn't necessary means they will never play WvW so if they do start playing WvW, it can cause the server to exceed the threshold by large amount. There are servers too with large amount of inactive players and that doesn't necessary means they will always be inactive so when they do return to play, they too can cause the server to exceed the threshold by large amount.

This fixed threshold has another flaw due to its fixed nature. There are servers that are very low in WvW count therefore it will takes a lot to hit that threshold. This huge space allow what we called stacking and inorder to stack, we have to destack from elsewhere. It may looks fine since they will still hit the threshold but let's not forget that we also need take into account of the unpredictable server's base populations and the limitations of the threshold algorithm which will be explained next.

Population Threshold Algorithm

Population threshold algorithm was based on populations alone and later modified to include play hours. The logic behind this change is that there are players and servers with many of such players who don't play as long hours as others thus resulted in certain empty timezones. The pro of this change is that it is a possible fix to the empty timezones. The con of this change is that it is also a cause for unbalance timezones.

Let do a logical assumed scenario of let say that server ABC has a WvW populations of 1000 that plays 1 hours therefore resulting in data value of 1000. Then, server XYZ has a WvW populations of 250 that plays 4 hours therefore resulting in data value of 1000. Both servers hit the threshold therefore is locked. If players of server ABC suddenly feel motivated and put in more hours, the population balance logic will completely collapse at that point onward.

Another flaw of this algorithm is it doesn't account for EOTM. While it is logical to exclude EOTM because EOTM is a mix of servers but at the same time it also become a exploitable area. Since the population threshold algorithm only takes into account of WvW, then, in the event that guilds or individuals decided to hang out in WvW due to whatever reasons, the accuracy of the population calculation will take a hit. Also, it is likely possible for players to game this algorithm just by getting everyone to hang out in EOTM for weeks while waiting for the algorithm to update in order to stack the server further.

The algorithm is observed to be updated by using data over a period of an month. Whether this data is entity based or collectively based, it is hard to determine. However, we can say that the data will take a month to be reflected on the status. While this can possibly prevent massive bandwagon, it also a flaw where any major guild wars 2 events that capable of pulling populations off WvW will end up opening the servers for that long, including previously extremely full servers, tipping the balance even more when the events are over.

Growth & Attrition

In WvW, we all want our servers or guilds to be able to grow to the point of able to match the numbers of other servers but is that possible? Using the image referenced from above, how likely are servers that slightly exceeding the threshold able to grow to able to compete against servers that far exceed the threshold? Unlikely because their growth is limited by the threshold. How long does it take for attrition to kick in for servers that far exceed the threshold? It is unpredictable because players can play multiple years and let's not forget about the returning players and players from base populations.

World Linking

In essence, world linking is a soft merger between servers where it give anet the flexibility to swap servers around if need to. However, world linking doesn't resolve population imbalance, it only resolve population decline. This is because every server is different in their populations across timezones, it will be quite a miracle if anyone is able to find matching servers for all the servers. This already difficult decision to minimize population differences is made even more chaotic when large groups of players performed transfers between servers or deliberate boycott WvW to tank populations, throwing populations data into disarray and thus take months for it to become accurate once again. At the mean time, the rest of the populations have to deal with the extreme inaccurate populations linking decisions made base on it.


Again, this is not a population balance suggestion thread so please post your population balance suggestion(s) on thread(s) of your own.

Founder & Leader of Equinox Solstice [TIME], a Singapore-Based International PvX Guild
Henge of Denravi Server
www.gw2time.com

--

Explanations of WvW Structures & Populations Issues

Comments

  • CattivoUomo.7198CattivoUomo.7198 Member ✭✭
    edited September 20, 2017

    Agree with pretty much all of this. Population discrepancies throughout the course of a day/week/month vary drastically. Those worlds with 24/7 coverage where they typically outnumber their opponents are naturally going to control the points per tick. Another 'player induced' issue not covered in your post is persistent 2 vs 1 situations where the 3rd world is constantly hammered on by player alliances. Now this is understandable, and is clearly the intent of having 3 worlds, but it does lead to players shying away or avoiding WvW altogether, thus leading to even more population imbalances. I would be curious to hear thoughts on system enforced alliances, meaning alliances that get triggered when the world scores deviate passed a certain threshold. For instance, say server G is 20 points ahead of server B and 30 points ahead of server R. An enforced alliance could then be triggered that makes server B and R friendly to each other thus forcing them to fight only server G for a period of time. This by the way is not a population balance suggestion, simply an idea for breaking a persistent 2 vs 1 scenario and keeping scores a little more competitive.

  • @CattivoUomo.7198 said:
    Agree with pretty much all of this. Population discrepancies throughout the course of a day/week/month vary drastically. Those worlds with 24/7 coverage where they typically outnumber their opponents are naturally going to control the points per tick. Another 'player induced' issue not covered in your post is persistent 2 vs 1 situations where the 3rd world is constantly hammered on by player alliances. Now this is understandable, and is clearly the intent of having 3 worlds, but it does lead to players shying away or avoiding WvW altogether, thus leading to even more population imbalances. I would be curious to hear thoughts on system enforced alliances, meaning alliances that get kitten when the world scores deviate passed a certain threshold. For instance, say server G is 20 points ahead of server B and 30 points ahead of server R. An enforced alliance could then be kitten that makes server B and R friendly to each other thus forcing them to fight only server G for a period of time. This by the way is not a population balance suggestion, simply an idea for breaking a persistent 2 vs 1 scenario and keeping scores a little more competitive.

    I think on a high level that sounds like something worth trying as an experiment at least. The problem is, what should that be based on? Overall war points? Skirmish war score?

  • The example above outlines why people often feel like they are outnumbered (because they are in their timezone) but the overall population is the same. Leaving the balance portion up to players hasn't resulted in populations evening out either. In other games they will try a method like opening up a new server or only having the linked servers accept new players. I'm afraid that doesn't really tend to work over the long term. Then those same players will just get farmed by the comped groups who run 30-50 deep. There really needs to be a viable reason to destack and that hasn't been provided. A larger server can farm pips and enjoy their time without any real fear of competition. To the OP's point, rarely (if ever) have I encountered a match where the 2nd and 3rd place server team up to take on the 1st place server. Great post, I'm hoping it generates some discussion.

    I think folks are curious sometimes about why someone would stay on a lower ranked server. Personally, I go to where I feel like I can have the best experience for my playstyle. Being in last place isn't terribly fun but we still have a good time on Tarnished Coast. I log in, get into fights, down players, have to use even more strategy to attempt to compete, you get the idea. To me it is way more challenging (frustrating too).

  • FogLeg.9354FogLeg.9354 Member ✭✭✭

    @Fatherbliss.4701 said:
    There really needs to be a viable reason to destack and that hasn't been provided.

    Yep, as long having more players means more points and better chance to win we can keep jumping around different methods of dividing players into "servers" and they still find ways to get into few more populated ones. Only way to make sure all servers get equal population would be redesigning WvW systems so players themselves want to go into less populated servers.

  • Sovereign.1093Sovereign.1093 Member ✭✭✭✭

    the idea therefore if you want to have a link is play at yourutmost best for 2 hours then log off until linked. then return to as long as you want =)

    Slayers XD. Love wvw? Prove it. Join NSP.

  • I just want to be able to play with my guild again. They transferred months ago, but during this time I was unfortunately dealing with a death in my family so I wasn't thinking about the game - nor did I think I'd not be able to play my preferred game mode with my preferred guildies. The balancing attempts by Anet has simply blocked stranded players from really having any fun and I already know a number of people who just kind of gave up on the game because they feel like they can't enjoy it anymore. I myself have barely found desire to put time in beyond checking my daily rewards, maybe a story event or two, or simply just to see if the server opened yet. It's just not fun like this - and I don't think I would be upset if they seemed like they were actually figuring out a good way to balance the game - but it just seems like they're barricading the way for -any- real change.

    On one end - I understand that there is probably a huge avenue of problematic algorithms and mathematical equations that they are trying to impose in the game in hopes they get the expected results - but i simply don't think this current system is working fast enough or even enough period to provide a reasonable explanation as to why good standing players who just want to enjoy the game with the people they enjoy playing with in the style they like to play have to wait on the sidelines for it to 'work out'. I'd rather spend my time playing an unbalanced game with all my friends then playing an 'almost kinda sorta' balanced game without the friends.

    Unfortunately, many replies I see lately are people telling the player to convince their guild to move down in tiers just so they can collect the few players locked out of the system. Which will both cost the guild a small fortune if its significantly large and to separate from the other guilds they have formed bonds with and likely created a synergy with. It is an unreasonable solution for most people - and so we wait.

  • Sovereign.1093Sovereign.1093 Member ✭✭✭✭

    @ZalonWolf.3291 said:
    I just want to be able to play with my guild again. They transferred months ago, but during this time I was unfortunately dealing with a death in my family so I wasn't thinking about the game - nor did I think I'd not be able to play my preferred game mode with my preferred guildies. The balancing attempts by Anet has simply blocked stranded players from really having any fun and I already know a number of people who just kind of gave up on the game because they feel like they can't enjoy it anymore. I myself have barely found desire to put time in beyond checking my daily rewards, maybe a story event or two, or simply just to see if the server opened yet. It's just not fun like this - and I don't think I would be upset if they seemed like they were actually figuring out a good way to balance the game - but it just seems like they're barricading the way for -any- real change.

    On one end - I understand that there is probably a huge avenue of problematic algorithms and mathematical equations that they are trying to impose in the game in hopes they get the expected results - but i simply don't think this current system is working fast enough or even enough period to provide a reasonable explanation as to why good standing players who just want to enjoy the game with the people they enjoy playing with in the style they like to play have to wait on the sidelines for it to 'work out'. I'd rather spend my time playing an unbalanced game with all my friends then playing an 'almost kinda sorta' balanced game without the friends.

    Unfortunately, many replies I see lately are people telling the player to convince their guild to move down in tiers just so they can collect the few players locked out of the system. Which will both cost the guild a small fortune if its significantly large and to separate from the other guilds they have formed bonds with and likely created a synergy with. It is an unreasonable solution for most people - and so we wait.

    my guild has 144 members if we move thats like 500 gold. 70k haha.

    this game should have no pop cap.

    Slayers XD. Love wvw? Prove it. Join NSP.

  • Strider Pj.2193Strider Pj.2193 Member ✭✭✭✭

    @ZalonWolf.3291 said:

    Unfortunately, many replies I see lately are people telling the player to convince their guild to move down in tiers just so they can collect the few players locked out of the system. Which will both cost the guild a small fortune if its significantly large and to separate from the other guilds they have formed bonds with and likely created a synergy with. It is an unreasonable solution for most people - and so we wait.

    There are definitely situations that most players would understand to allow bypassing caps.

    The problem is, too many servers have learned to take advantage of hibernating to allow whole guilds into servers. Stacking is definitely going on.

    People have noted some incentive for players to move down would be helpful.

    You are right that it is not cost effective for the guild to move down.

    Maybe Anet creates a player swap system where a server that is on the edge of their 'full' server status would create an option for players to swap spots: ie trading spots on full servers for the transfer fee.

  • Sovereign.1093Sovereign.1093 Member ✭✭✭✭

    swapping is actually a good idea. make it happen anet =)

    Slayers XD. Love wvw? Prove it. Join NSP.

  • To your point, Strider. I suggested that, in the past, Anet should give us the option to give up a WvW server world for a new 'unaffiliated' world status (which would have no access to WvW). Then allow unaffilated to Queue for WvW Server swaps - but i'm not sure how a queue would work - it'd likely have to be a 24 hour queue so people don't miss their chance to jump servers. Basically the idea works the same as swapping. I give up my spot on Tarnished Coast to allow someone from server B to move which allows server C to move and opens up my wanted spot on server D. In theory.

  • XenesisII.1540XenesisII.1540 Member ✭✭✭✭

    Wvw is an open join mode designed to hold large groups 24/7, there's always going to be an imbalance other than the first hour of reset. The only way you're ever going to get around that is by having smaller map caps, or controlled using instances like spvp to fill out the numbers, but even then it will still exist in some way. The system was also fully built around point scoring, which means there are players who will always want to be in the best situation for scoring, that means being on the biggest and best coverage server. It also means it created 2v1 for the wrong reasons, the 2 strongest beating on the weakest for points when it should be the other way around.

    Does anyone think BG would be as it is today if there wasn't a scoring system to motivate them to stack? All those free transfers did do quite a bit of damage to the system as well back then, although it could be put to good use these days to repopulate the lower servers, if anet could control it properly.

    Another derailing post. ^^
    EBG North Keep: One of the village residents will now flee if their home is destroyed!
    Maguuma: Free ppt, come and get it!

  • Strider Pj.2193Strider Pj.2193 Member ✭✭✭✭

    @ZalonWolf.3291 said:
    To your point, Strider. I suggested that, in the past, Anet should give us the option to give up a WvW server world for a new 'unaffiliated' world status (which would have no access to WvW). Then allow unaffilated to Queue for WvW Server swaps - but i'm not sure how a queue would work - it'd likely have to be a 24 hour queue so people don't miss their chance to jump servers. Basically the idea works the same as swapping. I give up my spot on Tarnished Coast to allow someone from server B to move which allows server C to move and opens up my wanted spot on server D. In theory.

    I think I saw that post. That would work in theory, and be hard to 'game' but of course I have no technical knowledge of whether it is an easy coding process.

    My thought would likely have been filled with potential abuses.....

    I think one of the biggest problems would be that a 'Full' BG does not equal a 'Full' TC, or even JQ which is the only other server right now without a link. So there might need to be more controls on it.

    The threshold for full has definitely moved.

  • ZalonWolf.3291ZalonWolf.3291 Member
    edited September 21, 2017

    Ah, I would still say they'd need a hard cap. For the sake of numbers lets say hard cap is 100 people. (It's not but in theory.) BG might be at 160 JQ 130 the other 'full' servers at 110. To have the ability to drop out and queue as the first dibs when a server finally drops to 99 to fill that one open spot - this would be the main incentive for people to leave their servers just to earn the first dibs options when it does open. Again it is in theory, but if 15 people from X wants to go to Z but X is 10 people above the cap... and those 15 drop out to claim their spot in queue... i think it might just be enough to make a dent in balancing. This previous statement assuming the people in full servers wanted spots in other full servers. The bigger problem is they'd eventually become hella long queues and people will likely sit in them for weeks/months... but it's still better than sitting here and seeing zero change at all.

    I dunno... maybe that is just me.

  • Strider Pj.2193Strider Pj.2193 Member ✭✭✭✭

    Yes, some hard adherence to a top end is necessary. At least until all the servers are close to that 'cap'.

    Now the 'fly in the ointment' (bear with me I've seen a few winters...) is the PvE population on each server that doesn't play WvW. If they ever are incentivized to join WvW, then those caps become fuzzy again. I think we will see this with people returning for PoF. Just not in the first week or two.

  • Very much so agree. This is what adds insult to injury to those of us who want to try to catch up with our guildies. If old players flood black and Anet doesn't remove their current stance... I can see many people just giving up on this entirely. No ones going to want to wait for a server with a full population, double over the cap. This would pretty much guarantee no good solution until WvW is gutted and changed by a new system entirely.

  • SkyShroud.2865SkyShroud.2865 Member ✭✭✭✭

    Thanks, I miss out on how population algorithm update behavior can be exploited.

    Founder & Leader of Equinox Solstice [TIME], a Singapore-Based International PvX Guild
    Henge of Denravi Server
    www.gw2time.com

    --

    Explanations of WvW Structures & Populations Issues

  • Just EOTM it and have it allow you to log into a randomized server with friends each time. Get rid of these entrenched loyalties and needlessly stubborn entities. People can still group up but one day you might be mag, another day you might be SBI etc etc. I can tell you with permanent home servers and expensive transferring you will never have the balance you seek to obtain, not by any stone's throw.

  • Sovereign.1093Sovereign.1093 Member ✭✭✭✭
    edited September 22, 2017

    just because some servers are full does not mean they are equivalent or the same because others give more effort than others.

    some use the best food. have full ascended infusions. use custom builds to break meta.composition.

    basically besided the population, each server.must strive to be better too.

    on that note, no to population cap =) so if servers are barren, remove them. if servers have q, eotm x more wvw maps. hehe

    why not have only 3 basic sides.

    login to server a b or c. then those three will duke it out against each other.
    population problem gg.

    and as for ques. make new maps or just copy alpines

    Slayers XD. Love wvw? Prove it. Join NSP.

  • Aeolus.3615Aeolus.3615 Member ✭✭✭✭
    edited September 22, 2017

    @XenesisII.1540 said:
    Wvw is an open join mode designed to hold large groups 24/7, there's always going to be an imbalance other than the first hour of reset. The only way you're ever going to get around that is by having smaller map caps, or controlled using instances like spvp to fill out the numbers, but even then it will still exist in some way. The system was also fully built around point scoring, which means there are players who will always want to be in the best situation for scoring, that means being on the biggest and best coverage server. It also means it created 2v1 for the wrong reasons, the 2 strongest beating on the weakest for points when it should be the other way around.

    Does anyone think BG would be as it is today if there wasn't a scoring system to motivate them to stack? All those free transfers did do quite a bit of damage to the system as well back then, although it could be put to good use these days to repopulate the lower servers, if anet could control it properly.

    that is what WvW was ment to do, 1st and 2nd server of each match ktraining the 3rd server map zones for rewards, now that ANet added the rewards tracks every one want to be on the winning side for easy farm.

    IMO WvW needs to be abolished how it is... for something simplier but better.

  • @Aeolus.3615 said:
    that is what WvW was ment to do, 1st and 2nd server of each match ktraining the 3rd server map zones for rewards, now that ANet added the rewards tracks every one want to be on the winning side for easy farm.

    Reward tracks didn't do that, the tournaments did (also human nature, but whatever). They, above all else, were the cause of the population imbalance in WvW.

  • SkyShroud.2865SkyShroud.2865 Member ✭✭✭✭
    edited October 30, 2017

    Updated analysis on possible population threshold algorithm frequency.
    Attached image, thanks to the forums image update.

    Founder & Leader of Equinox Solstice [TIME], a Singapore-Based International PvX Guild
    Henge of Denravi Server
    www.gw2time.com

    --

    Explanations of WvW Structures & Populations Issues

  • SkyShroud.2865SkyShroud.2865 Member ✭✭✭✭
    edited January 26, 2018

    Opps, quoted instead of edit.

    Updated insights on World Linking.

    Founder & Leader of Equinox Solstice [TIME], a Singapore-Based International PvX Guild
    Henge of Denravi Server
    www.gw2time.com

    --

    Explanations of WvW Structures & Populations Issues

  • Mysteriax.6049Mysteriax.6049 Member ✭✭✭
    edited January 26, 2018

    I like the idea of that forced 2v1. Think about big fights like in SM where the dominant T3 SM server has to hold off two groups instead of one! Also prevents second and first place focusing third place instead of second and third place putting pressure on first place. This 2v1 would happen in select situations and I'm sure it could be balanced.

  • Just read this, gave me some insight on a few things. +respect for putting loads of work into this!

    www.airguild.us Henge of Denravi

©2010–2018 ArenaNet, LLC. All rights reserved. Guild Wars, Guild Wars 2, Heart of Thorns, Guild Wars 2: Path of Fire, ArenaNet, NCSOFT, the Interlocking NC Logo, and all associated logos and designs are trademarks or registered trademarks of NCSOFT Corporation. All other trademarks are the property of their respective owners.