Implementing Alliances (Part 8)

The alliance system that I have proposed over the past week is a massive change from the existing server system. This begs the obvious question: how to move from servers to alliances? To tie this series up, here’s what I suggest.

NOTE: The following order presumes that the infrastructure and back-end to implement alliances has been built, and all that remains is to transition.

Step 1: Servers into Mercenaries

At the start of a week’s match, a switch is flipped that turns all existing server matches (tiers 1-8/9) into mercenary-based matches. For sake of understanding, each of these matches is named based on the old servers that were a part of it. Players are all considered mercenaries, and can freely join any match with the same restrictions as outlined in Part 4.


Having a “match selection” mechanic in place before alliance matches even happen removes a technical hurdle early on in the process, and gets players used to the new way of entering matches before the ladder activates.

Step 2: Open Alliance Registration

Ideally happening at the same time everyone becomes mercenaries, either an in-game interface (more convenient) or official website (could be faster to develop) launches, allowing guild leaders to register an alliance. This should have a few simple administration functions:

  1. Can start a new alliance, including name and description
  2. Can invite other guilds
  3. Can kick other guilds

The ability to kick can be implemented several ways, perhaps the alliance leader gets to do so, majority vote chooses, or the alliance leader can delegate kick authority.

The website should also have a list of registered alliances and their member guilds, allowing anyone to see who’s out there.


Alliances can’t work unless there’s someplace to sign up, as well as somewhere to see who’s organized together. Opening registration is a mandatory part of making that a reality.

Step 3: Start the Alliance Ladder

After a couple of weeks of signups, or an adequate number of alliances has signed up (I suggest 12 or more), alliances are pitted against each other based on raw membership count. For each alliance match that starts, one of the 8-9 mercenary matches deactivates until there are none. Mercenary matches after that point are based on the demand of WvW beyond the alliance matches (see again Part 4).

These initial ladder matchups should run on a fresh algorithm (similar to the Glicko resets that happened early in Guild Wars 2‘s release), with minor weighting given to ratings to account for population differences. This reduces the likelihood of a big fights-focused, 3,000-man alliance from losing to two other similar ones, and facing a roaming-focused, 200-man alliance the next week.

Perhaps to help with settling ratings out quicker, matches could be 2-3 days long each, rather than a single week. If so, 3 day matches should be on the weekends, with two 2-day matches during the week to reflect similar population conditions.


Population is a loose determiner of alliance strength, and with the flexibility of alliance size inherent in the setup (1-5000 people), can also be used by alliance members to “tilt the scale” toward a matchup they’re more interested in playing.

Step 4: The First Season

Once ratings have settled down (low volatility on the alliances), kick off the first season, pulling together groups of 6 and 9 alliances into leagues and organizing a format around them. This is virtually identical to how the previous 3 tournaments functioned, but takes into account the dynamic nature of alliances.

Step 5: Maintain

And that’s it. In four steps, WvW has fully transitioned from a server-based system to the new alliance-based system, with a consistent ladder and regular seasons.


Alliances are the best way to move forward with World vs. World. They’re not perfect, and they won’t satisfy everyone, but giving players the tools to self-organize and letting them compete on their terms best fits the competitive community that the mode has always fostered.

Further, by implementing alliances, mercenaries, and updated and prestigious rewards, ArenaNet can take a step back from WvW entirely. With minor course corrections and adjustments, an alliance system can be self-regulating and self-sustaining. It can reduce developer overhead (and player rage at WvW being ignored) tremendously.

On one level, alliances as I’ve proposed are a terrible idea. It relinquishes structural control over what players do, and if they destroy the game mode by their actions, the blame will still fall on the developer.

But any such destruction will be sudden and abrupt. WvW has been dying for years now, with activity dropping lower and lower, and its veterans getting more and more bitter over what they love being so utterly ignored.

It’s pretty much at the point of ArenaNet having nothing to lose by attempting something as radical as alliances, even if it fails. Players have been asking for it for years, guilds have left for other games over its lack, and it is far stronger than the server system ever can be. Take the chance, and make something amazing.

The Alliances Series

  1. Putting Guilds Back into Guild Wars: WvW Alliances
  2. Unbalanceable Problems: Why the Server System Cannot Work
  3. Heart and Soul: The Guilds of World vs. World
  4. The Case for WvW Alliances
  5. The Alliance Structure: Sidestepping the Problems
  6. The Pros and Cons of WvW Alliances
  7. Rewarding the Competition
  8. Implementing Alliances

Rewarding the Competition (Part 7)

World vs. World rewards suck. While changing elements of the game mode (like defense and extended siege) to incorporate rewards will help, rewards across the board need a serious overhaul. With the formalized structure of alliances, providing rewards that are both prestigious and useful should be much easier to accomplish than in the old server system.

This can be accomplished by providing different rewards for three sets of players: individual WvW players, alliance members, and tournament winners. In concert, WvW can be rewarding for many more than simply those who enjoy the experience alone.

Individual Rewards

Individual rewards are the trickiest ones to execute on, because loot rewards have to weigh exploitation against proper reward. Karma trains and other forms of farming are already common, so all of the rewards below function outside of the Gold Standard as much as possible.

Also important in the alliance system is keeping individual reward as similar as possible between mercenaries and alliance members. Picking an alliance (or not) is a choice, and individual rewards should not bias toward either side of that choice. The rewards below work for both mercenaries and alliance members.

Ditch WvW Rank-Up and Bonus Chests

These chests are terrible loot. While the small possibility of exotic and ascended drops exists, it functions as a fancy reminder of how much the Gold Standard makes the average loot complete garbage. It doesn’t help that they were added to bring more loot to WvW, and the mode still pales in comparison to a few minutes in PvE.

Adopt a system similar to Fractals, where some amount of tokens (relative to level attained) is dispensed in addition to account-bound valuable extras. Tokens can be accrued for acquiring WvW-exclusive skins, armor, infusions, etc., or there’s a small chance (player-based small, not community-based small) of getting any of those with the rank-up.

Skins in this case don’t necessarily have to be weapons and armor, but can be special siege skins, or reskinning the NPCs of an owned objective, or altering the looks of projectiles fired from manned siege. The key with skins is offering power-agnostic ways of letting players show off that they’ve been out on the fields of war.

Add Weekly Rewards

One of the persistent issues with rewards is the difficulty of measuring participation. In an alliance-based system, simply saying “they spent time in the current match” doesn’t work for participation. Mercenaries can be anywhere, contributing in a dozen different matches in small, but helpful ways.

Another problem with “they zoned into WvW” is that it requires no actual participation by a player*. As a result, rewards have to be mediocre or people doing practically nothing can net something awesome.

*This is based on my anecdotal evidence of times where I only logged in for maybe a few minutes, and still picked up a bonus chest at the end of the match. I looked on the wiki but didn’t see any mention of how much participation is actually required. Still, the bar is far too low.

Instead, implement weekly rewards similar to the existing dailies (dolyaks slapped, towers captured, players killed, etc.), with 3 tiers of each type. Achieving a tier gives a small reward, and capping a type out (all 3 tiers) gives a larger reward.

With a weekly, tiered reward system, participation can be readily measured. Someone with 3 total tiers achieved (of 21, to give an example) has minorly participated in WvW. Someone with 15 has heavily participated.

When a match ends, individuals can be given rewards that line up with their level of participation. Rather than having to go with the lowest common (unabuseable) denominator, players get an extra bonus that matches how much they dedicated.

Build Rewards into Achievement Tiers

While WvW achievements are getting a long-overdue overhaul with the launch of Heart of Thorns, they could also benefit from additional rewards when each tier is achieved. This gives the feeling of progress while working toward intentionally long-term goals beyond “more AP.”

Add Exclusive Rewards to WvW Rank

Right now, any WvW Rank past 1,390, when all abilities are maxed out, is immaterial. With Heart of Thorns, the points required to max is going down. Allies can’t see ranks, and enemies end up using high-rank titles as an excuse to single out and remove specific players from the fight (typically the enemy commander).

While I’m sure the original intent with ranks was to avoid elitism (“you’re just an Assaulter, you don’t know squat”), making ranks purely a way for enemies to see more than “<Enemy Server> Invader” isn’t a good solution either.

Instead, give players the option to show their title, similar to the dozens of other ones PvE has been showered with through Living World. They can only show their currently attained rank, and not any ones below it.

Also, rank should unlock the ability to pick up different, special rewards. While a Soldier might only have the stock skins to pick from, a Silver General could pick up silver-plated siege. This gives prestige to the rank beyond simply the title of it, and allows players to show it off in ways that don’t affect power balance.

Alliance Rewards

Alliances are designed around competition and hardcore players pulling out all the stops to achieve victory. As such, all alliance rewards should focus on the prestige and glory of victory. And only victory.

For Alliance Members

Winning a match gives a small amount of a unique currency to alliance members. Similar to WvW Tournament Tokens, the rewards available can vary from simple, but unique trinkets, to exclusive siege skins, weapons, and armor. They could also work toward buying a Victor finisher as a capstone goal.

For Alliance Guilds

Guilds gain the ability to display a victory banner in their halls with their alliance’s first win. The banner gets more and more exquisite (or perhaps splits into multiple decorations) with each victory. Accruing enough wins could unlock new decoration sets and special guild siege skins based on their prestige.

Tournament Victory

Finally, seasons and the tournaments thereof need to have even higher prestige. Victors have consistently proven they’re better than their opponents, and should be recognized on that level.

For alliance members, the existing tournament token system is fairly close to ideal, and should be co-opted for the alliance system*.

*Though I’d argue that Hero’s and Mistforged Hero’s weapons should be converted to ladder victory rewards and new skins crafted for tournaments.

For the alliance itself, unique guild hall victory banners are crafted for each tournament, and awarded to the 1st, 2nd, and 3rd of each league, with a mention of which league they were in. Tournament placement can also apply a sizable amount of wins to each guild’s victory tally when it comes to additional decorations.


Combining all of these ideas creates a more rewarding system that can be tailored to match participation against rewards without having to fight with the Gold Standard. More importantly, it creates a place where prestige and straight-up bragging rights are achievable, without providing one iota of power advantage to the victors.

Now that I’ve addressed rewards, one thing remains: how could ArenaNet move from the existing server system to an alliance system? How can that transition be kept as smooth as possible? I’ll conclude the series by answering both of these questions.

The Alliances Series

  1. Putting Guilds Back into Guild Wars: WvW Alliances
  2. Unbalanceable Problems: Why the Server System Cannot Work
  3. Heart and Soul: The Guilds of World vs. World
  4. The Case for WvW Alliances
  5. The Alliance Structure: Sidestepping the Problems
  6. The Pros and Cons of WvW Alliances
  7. Rewarding the Competition
  8. Implementing Alliances

The Pros and Cons of WvW Alliances (Part 6)

An alliance system for World vs. World is a massive overhaul of the existing server system. But what it provides to the longevity and stability of the mode cannot be understated. For every disadvantage, far stronger advantages are present.

Advantages for the Playerbase

The alliance system strikes a balance between giving hardcore players something to fight for, amongst like-minded people, and allowing more casual players the flexibility to find what they enjoy.

Caters to the Hardcore

First and foremost, alliances provide a simple structure for consistent competition. Hardcore players want the excitement of facing off against a difficult foe, and proving victorious. While karma-training might provide lots of loot, it’s a pale shadow of the joy of beating an equal.

Second, they create asymmetric balance by reducing the amount of imbalance present in the existing, tier-constrained system. Add in mercenaries, and being provided with an equal amount of players to prosecute the war effort is much closer to a reality.

Third, hardcore players get to pick their overall meta. Alliances are flexible, allowing players to organize into whatever size and philosophy they want. Then like-minded alliances end up fighting each other, creating competition at multiple levels, far more than the current “only in the high tiers” server system.

Does not exclude Casuals

Mercenaries, those unaffiliated with an alliance, get to pick their match each and every week. They even get to change matches within a week, dabbling in a siege-based meta one night, and a fight-focused one the next. And if the alliance-based matches are of no interest, they can queue into the free-for-all of a mercenary-only matchup.

And since mercenaries augment existing matches, while giving priority to alliance members, their presence is more often helpful than harmful. No more “randoms clogging the queues” or “the fairweathers went home” complaints on map chat*.

*And if an alliance has fairweathering issues, that can be solved by restructuring the guilds within. No more blaming people who have no stake in the matter.

Creates a Smooth Transition from one to the other

The major advantage of registered alliances is that all member guilds and alliances are known by the player base at large. This allows alliance members to know who the competition is, and gives mercenaries the ability to know who they are fighting with.

Any given mercenary can spend time over the course of a match alongside a guild, learning from them and seeing how they operate. And if said mercenary is so inclined, join that guild and enter the meta and community he prefers.

Mercenaries can also never commit to an alliance, reaping the reward of playing whatever meta interests them at the time, all while not getting in the way of those who know what they like. Mercenaries and alliance members alike get to spend time surrounded by those who enjoy the same things.

Advantages for Developers

Removes the overhead of servers

Named servers only exist, and only matter, because of WvW. Ever since April 2014 when megaserver rolled out to PvE and structured PvP, WvW has come off as the oddball. While elsewhere players can enjoy the game with whoever they want to, in WvW it matters where specifically you are from.

Alliances remove the last reason to keep named, specific servers around. Playing with people is either a simple matter of partying up and joining the same match, or joining the same guild they’re in. Rather than suddenly being unable to play with each other, WvW will act very much like the rest of the game.

Enables easy adjustments based on demand

The alliance system avoids the square peg that server tiers have been compared to the round hole of WvW player interest. Rather than having to weigh server mergers to create “the right amount” of people against a sudden influx of players, alliances make it dynamic.

However many alliances there are determines the amount of WvW going on. While mercenary matches could potentially be a large overhead over the number of alliances, the hardcore player base has always put in the lion’s share of time in WvW. The amount of activity they have is a solid barometer for how many servers should be apportioned for WvW*.

*Before you all skewer me with pitchforks, I mean hardcore in the time spent sense. Those who spend more time in WvW will be more affected by large changes than those who spend less, by definition. Also, the amount of effort and time that the hardcore want to spend on an alliance system will likely reflect the amount of time casual players will want to spend.

Stronger, More Useful Metrics

Alliances also create a much better situation for using metrics. Rather than having to extrapolate hardcore, casual, and other types of players based on overall population numbers, the alliance member/mercenary distinction creates much better data.

Also, the differing metas among matches can show the popularity of certain styles of play. Players can choose what they want to play, and the data will show that.

Simpler ground for experimentation

Alliances are known quantities. Rather than having to push a change or event to every server, whether or not it’s good for them, ArenaNet gets the flexibility to pick specific alliances for test-driving a change, or activating an event designed to switch up a given meta.

Taking it a step further, the developers can straight-up ask alliances “hey, we’re considering doing this, would you mind testing it in your next match?” Put it to a poll, using the guild membership rosters to limit who votes. Add in a minor reward as a “thank you” for testing, like additional wXP gain or a nice piece of gear.

Mercenary matches could also be used explicitly as test servers. One match could be a “beta” server that anyone could join, providing a test within the live environment without forcing those uninterested in testing to participate as well.

Finally, having changed matches occurring in tandem with unchanged matches provides terrific grounds for comparison. WvW changes based on a host of different factors, and seeing both match types at the same time controls for many of them, making the effect of the changes much clearer.

Creates the means to be “hands-off”

Above all, an alliance-based system, despite its upfront effort, drastically reduces the amount of upkeep ArenaNet has to do to keep WvW going. Rather than screaming for “new content” every single day, players will have the power to create new situations and new approaches with alliances.

“New content” has always been a bandage seeking to make up for the lack of balance and competition in WvW. Alliances patch up a lot of the balance and competition, reducing the need for something new*.

*Though no one is opposed to new content. It just doesn’t have to happen every single month for WvW players to be happy.

WvW has always been about communities competing against each other in a structured environment. Alliances allow communities to do that, on their own terms. No longer does ArenaNet have to spend time trying to change up all of WvW to fix issues that only happen in certain parts of it.

The Prime Disadvantage

Before I get to any other disadvantages, alliances are player-created, player-managed, player-driven. As such, it can go terribly, terribly wrong. Consider these two examples of alliances going poorly:

Too Few Alliances

If there are too few alliances for the styles of play, hardcore players continue to have unavoidable mentality conflicts. Far from allowing alliance members the flexibility to pick their own meta, it will replicate the existing problems in the middle tiers of some people wanting much different things than others.

Likewise, there won’t be enough alliance matches to allow mercenaries to augment. They could very well overwhelm the alliance itself, purely because there aren’t enough alliance members to set the tone.

Stacking Servers Alliances

And if players stack a single alliance, or very few, trying to create an unstoppable juggernaut, competition vanishes. The existing situation of the vast majority of hardcore players concentrated on a few servers, all battling with queues amongst each other, will return.

Mercenary players will constantly be added to massively underdog alliances, keeping them on the back foot each and every time they load in and lowering their interest. Nobody likes always losing, so they will stop queueing into alliance matches, perhaps favoring ultimately meaningless, but even, mercenary matches.


Removes the Ease of Servers

Servers are simple, almost foolproof, from a “loading in and fighting now” approach. A player, no matter what server he is on, can pull up the WvW panel, click a map, and be fighting instantly. It’s delightfully uncomplicated.

With alliances, either the player is loading into a specific match as before (happens to be in an alliance guild), or has to pick what match to be in. For a newcomer, that’s a fairly tall order and can stop people before they even start.

To alleviate this side of things, match selection for new players should be automatic for the first few times. Put them where most of their guildmates are, or where their friends or party is, with no prompting. This keeps things simple for newcomers, but retains the flexibility for those with more experience.

Separates Players into Two Explicit Classes

In the server system, players are in the system’s eyes perfectly equal. No one is more important than another; only server choice matters.

Alliances separate players into alliance members and mercenaries, giving specific advantages to each, but not treating them the same. The “bump” aspect of alliance matches, explicitly places alliance members above mercenaries.

This stings, especially when it’s a systemically required difference. For all the advantages of competition, dynamic balancing, and mercenary flexibility, knowing that an alliance member can give a mercenary the boot doesn’t feel good.

The “bump” could be completely removed in implementation. This keeps two classes, but once they’re in a match they’re equal. My concept with bringing in a “bump” aspect is to avoid the opposite of the advantage above: alliance members kept from their only match griping about having to wait for randoms.

Removes Server Culture

Even if server culture isn’t as strong as it could be, it still exists. Switching to an alliance-based system removes name-based culture in one fell swoop, forcing members of those servers to reorganize into an alliance to maintain the culture they have.

This is unavoidable, and while a change to alliances could be broadcasted well in advance, people will still be caught offguard, and very unhappy about losing their server that they’ve been a part of for a long time*.

*I would argue that to make an omelette you’ve gotta break a few eggs, but that doesn’t change the fact that someone’s egg can be very precious to them, unbroken.


Alliances cannot work without the conscious, direct, and intentional effort by the hardcore WvW player base to make it happen. It’s taking all of the tools of community creation and creating competition and giving them squarely to players. It’s a big challenge, and it can very well backfire and kill WvW forever, but if players rise to it, the result will be amazing.

And while alliances do have their downsides, what they provide to both players and ArenaNet is much better than what the existing server system can provide.

As stated earlier, switching to an alliance-based system provides an ideal opportunity to update and enhance rewards. Rather than matches being meaningless, tomorrow I will lay out a possible way to adjust rewards based on the new system.

The Alliances Series

  1. Putting Guilds Back into Guild Wars: WvW Alliances
  2. Unbalanceable Problems: Why the Server System Cannot Work
  3. Heart and Soul: The Guilds of World vs. World
  4. The Case for WvW Alliances
  5. The Alliance Structure: Sidestepping the Problems
  6. The Pros and Cons of WvW Alliances
  7. Rewarding the Competition
  8. Implementing Alliances

The Alliance Structure: Sidestepping the Problems (Part 5)

Alliances neatly sidestep the problems that have plagued World vs. World since its inception, either eliminating them or reducing their effects. Consider the following problems highlighted in parts two and three.

Population Shifts

In an alliance-based system, drastic population shifts are less catastrophic due to alliances and players being locked to the match they started the week within. Any shifts have to wait until the match is over.

Fairweathering is by nature less likely because many alliances are the hardcore backbone that wants to play WvW. Alliance size determines initial place within the ladder, with roughly equal alliances facing each other*.

*More on this in Part 8.

There is still the possibility of alliances of equal power that don’t evenly divide by 3. This is where mercenaries come in. While they can possibly imbalance, or “overload” a given matchup, decent load balancing will funnel extra players to the alliance(s) at a disadvantage. Population will be dynamically shifted into a balance, even if alliances themselves aren’t.

Within seasons, designated victor situations will be harder to create. For one, no one can bandwagon the frontrunner after the season starts due to match lock. For two, alliances can only have 10 guilds. There is an upper bound limit for how much dogpiling people can do, and shuffling players and guilds around makes bandwagoning much more difficult.

Finally, the transfer meta disappears over time. Alliances provide players the choice of who they fight alongside, and the resulting style of play. Creating the “ideal team” is within a group’s grasp from day one. And of course the inevitable spike of activity in WvW due to the large overhaul that an alliance system would create.

Differences in Mentality

As stated above, alliances pick their style of play. Those who want to play “for the fights” can self-organize into alliances designed to beat the snot out of each other week in and week out. Likewise, PPTers stack the strength of their strategy up against each other.

However, an alliance-based system does not address the scoring system weakness of nightcapping. While alliances and mercenaries in concert can greatly reduce imbalances, players online during off-peak time remain more important to the score than those online during primetime.

Pride Issues

Alliances replace servers, in their entirety. Thus, the disconnect among “server pride,” “guild pride,” and rewards vanishes. Anyone within an alliance knows exactly who they fight with. While guilds and guild members might shift, alliances are deliberately formed as a team that servers only occasionally pulled off.

And while this system’s structure doesn’t address the meaninglessness of victory and defending, the sea change it ushers in provides an ideal opportunity to shift the rewards scheme completely*.

*I’ll go into more detail on this in part 7.

The Unimportance of Guilds

Solved. Alliances are explicitly made up of guilds, and a fixed number of them. The core structure of WvW is based around what was its heart and soul in the first place. Mercenaries may come and go night by night, but alliances and the guilds within them stand at the very center.


But alliances provide much more than simply sidestepping the old server system’s problems. While there is the possibility of things not working out as well, the advantages far outweigh the disadvantages. Which I’ll delve into in the next post.

The Alliances Series

  1. Putting Guilds Back into Guild Wars: WvW Alliances
  2. Unbalanceable Problems: Why the Server System Cannot Work
  3. Heart and Soul: The Guilds of World vs. World
  4. The Case for WvW Alliances
  5. The Alliance Structure: Sidestepping the Problems
  6. The Pros and Cons of WvW Alliances
  7. Rewarding the Competition
  8. Implementing Alliances

The Case for WvW Alliances (Part 4)

World vs. World needs major changes, and while the recent announcement of WvW upgrade changes (and associated guild upgrades) could be a part of that, it doesn’t address the structural issues of WvW as a whole.

To recap on why, consider the unbalanceable problems already present in the mode. They’re the same problems that have existed for years, and no number of upgrades, maps, or similar feature changes are going to fix them.

Ultimately, the problems are present due to the focus on servers, a fluid collection of guilds and players, rather than the actual core of the mode, guilds themselves. For all the advantages guilds provide, including WvW’s very lifeblood, they have far too little focus.

Enter: The Alliance

Alliances, as I will present them, aim to reduce (or outright eliminate) the structural problems that WvW suffers from, and provide a system that caters to both the hardcore and the casual player of WvW.

Alliances have three facets: a maximum of 10 guilds, a consistent ladder, and regular seasons.

Maximum of 10 Guilds

Alliances can be of any size, up to a limit of 10 guilds. Guild membership limits functionally set the maximum size of an alliance. In other words, an alliance can have anywhere between 1 and 5000 people in it. While on the face of it, this sounds hopelessly imbalanced, hold that thought.

Every alliance needs to register, with any member guild starting the alliance and inviting the other guilds into it. This could either be done on a website (easier to implement), or in game (easier to utilize).

Guilds not within a registered alliance are considered mercenary guilds. More on that later.

Consistent Ladder

The existing server-based system shows a consistent ladder, but limited to only 8 (or 9, in EU) tiers. Simply put, matches will always pit 3 alliances against each other. However, the number of matches dynamically changes based on the number of alliances present. For example, if there are only 9 alliances, only 3 matches are available, but if there are 42 alliances, 14 matches will happen.

This dynamic expansion allows for alliances to create their own experiences. Smaller alliances that want to do small-group tactics, roaming, and the like, will be able to exist, without being pressured by the fixed number of matches into zerg tactics. And of course, Tier 1-type matches with masses of players capping out every map will also exist.

If there isn’t an evenly divisible number of alliances, the 1 or 2 “extra” are considered mercenaries as well. “Extra” is determined by age of alliance; the newest alliances are mercenaries.

The ladder itself can retain a Glicko-based rating system, with specific adjustments made for the possibility of alliances collapsing, or several alliances forming from match to match.

Once a match starts, an alliance and all its member guilds are locked to that match until it ends. Even if a guild moves to another alliance, it won’t move to the new alliance until the next match begins. This is critical for balance and reducing population-jumping exploit scenarios.

Likewise, the players in each guild are match-locked. If they move guilds during the week, they don’t change who they fight for until the next match.

EDIT: Multiple Memberships and Alliances

Since players can have as many as 5 different guild memberships, there is the possibility of being a member of multiple alliance-affiliated guilds. For balance, a player can only participate in one alliance match. As such, a player with multiple guild memberships has to note which one they fight for week to week.

Similar to the match lock, if a player changes their fighting guild during the week, it doesn’t change their match until the start of the next one.

This could be implemented as either the guild they are representing at the time of reset, or a specific checkbox added to the guild UI.

Regular Seasons

Seasons are a great idea at their core, but they have always stung due to population imbalances that trump competition. Overall, the point of having regular seasons (perhaps, twice a year in spring and fall) is to create competition beyond the typical ladder matches. League size should adjust based on the number of alliances present, with overall 6-9 alliances being in each league.

Season-based rewards should be a mix of goodies for all players of WvW, and prestige-based rewards exclusively available to alliances and their members. To steal an example from the previous seasons, Mistborn Hero’s weapons for all players, and finishers and guild hall decorations for alliances.

Once a season starts, all alliance guilds are locked to that particular alliance (and their matches) for the duration. Members, however, retain a match-based lock.

All three of these aspects create a solid environment for hardcore, organized WvW play. Further, it’s play not just for maximum-population alliances, but also deliberately smaller ones.

But what about people who like dabbling in WvW, or don’t want the commitment of signing up for an alliance?

Mercenaries: The Fighting Reserve

Mercenaries augment alliance matches. They also provide a system for non-hardcore players to enjoy WvW without forcing them into Edge of the Mists. Anyone not in an alliance at the beginning of the match is considered a mercenary*.

*As noted above, guilds not within alliances, and alliances that are “extra” are also considered mercenaries.

Whenever a mercenary player queues into WvW, they are given an option of which match to play. Upon picking a match, the server compares population numbers for each alliance and slots that player into the least-populated alliance. Once selected, that player stays with that alliance for the remainder of the match.

This sounds restrictive, but a mercenary has the choice of every match being played. Potentially, this can include “free for all” sorts of matches similar to Edge of the Mists where mercenaries are fighting against each other*.

*One aspect to consider is how mercenary match locking will work with guilds and parties. One possibility is having the first person in a guild to queue in to a match locking all other guildmates, whether or not they queue into WvW. Another is less ideal, but maybe unavoidable: each person gets slotted by the balancing algorithm, regardless of affiliation. This leads to players getting separated from their friends, but is simpler to balance.

However, alliance members hold priority to queue into their matches. This means that if a map is capped and there are mercenaries within that map, an alliance member queueing in “bumps” a mercenary out.

Bumping should not be instantaneous. Rather, all mercenaries not in combat with enemy players when a “bump” signal is sent will get an option to switch map voluntarily for 30 seconds (perhaps with a minor wXP bonus for doing so). If no one volunteers, one is selected at random and given a 5 minute timer before they are removed from the map. If the selected mercenary dies, they are kept from reviving or waypointing, and told to switch.*

*Also, if more than one volunteers, they all get the bonus.

This gives priority to alliance members to decide their own match, but prevents mercenaries actively engaged in fighting from getting an unwelcome zone back to Lion’s Arch.

Mercenary Matches

Mercenary matches function similarly to the existing Edge of the Mists: a free-for-all set of maps that have no bearing on an alliance’s score. The difference is that while Edge of the Mists can create (and destroy) infinite copies, mercenary matches should be restricted to a small percentage of active alliance matches and persist throughout the week. This provides an incentive to seek alliance matches, as well as prevents ghost maps from happening.

Aside: Population Balancing Data Analysis

Mercenaries present a unique challenge and opportunity. Ideally, they function as population balancers, filling in minor gaps in coverage or capability, while still leaving the onus on each alliance to show up and win over their opponents.

But to do that well requires capturing and analyzing data of both alliance and mercenary activity. Which mercenaries go to which alliance is a very important question, because done poorly it uses data that bolsters imbalances, rather than smooths them out. At minimum, I think the following data should be used to decide where mercenaries go:

  1. Average alliance population per hour, per day (establishes the baseline of possible imbalance)
  2. Average mercenary population per hour, per day (establishes how much is available to balance from)
  3. Historical scores and PPT per hour, per day (counterbalances, using cases where fewer, more skilled players outplay more, less skilled players)

Using these three factors, with statistical analysis and data over time, should provide a solid basis for balancing mercenaries into alliance matches.


Alliances, in concert with the mercenary system, provide opportunities for both the hardcore and the non-hardcore. It gives those who know exactly what they want out of WvW the freedom to do just that, while offering those who aren’t sure yet the ability to dabble.

Furthermore, as I’ll explore tomorrow, alliances address all of the problems that the server system has suffered from, either eliminating or reducing each and every one.

The Alliances Series

  1. Putting Guilds Back into Guild Wars: WvW Alliances
  2. Unbalanceable Problems: Why the Server System Cannot Work
  3. Heart and Soul: The Guilds of World vs. World
  4. The Case for WvW Alliances
  5. The Alliance Structure: Sidestepping the Problems
  6. The Pros and Cons of WvW Alliances
  7. Rewarding the Competition
  8. Implementing Alliances

Sometimes what you love just needs some TLC