Dec 072012
 

I wasn’t going to write more in this series but after a request for more information by Felladin, how could I say no. So here we go – Conditions in Guild Wars 2.

 

An overview

There are a total of 11 conditions that players can inflict in Guild Wars 2 and have a variety of effects and durations. When reapplied to a target with the same condition, some conditions will stack in duration or intensity and a couple don’t alter all. When a condition is removed by a skill, all of same condition is removed.

Bleeding – Does damage over time, where the damage applied per stack per second at level 80 can be calculated to be 42.5 + Condition Damage * 0.05. Multiple applications stack the condition up to a cap of 25. Your character may exclaim about bleeding when this is applied.

Burning – Does damage over time, where the damage applied per stack per second at level 80 can be calculated to be 328 + Condition Damage * 0.25. Stacks in duration. Your character may exclaim about being on fire when this is applied.

Blind – Causes target’s next attack to miss. Reapplication has no effect. Your character may exclaim that they can’t see when this is applied.

Chilled – Reduces movement speed and skill recharge of the target by 66%. Reapplication increases the duration of the condition. Your character may complain about being freezing when this condition is applied.

Crippling – Reduces the movement speed of the target by 50%. Stacks in duration. Your character may complain about their leg or feet when this is applied.

Confusion – Does damage when the target uses any skill (including auto attack), where the damage applied per stack at level 80 can be calculated to be 130 + Condition Damage * 0.15. Unlike all other conditions, damage from confusion is halved in sPvP. Multiple applications stack the condition up to a cap of 25. Your character may ask what is going on when this is applied.

Fear – Causes the target to move directly away from the caster or centre of the area causing the condition. Can’t be stacked.

Immobilized – The target can’t move or dodge. Stacks in duration. Some teleportation skills (such as the mesmer staff skill 2) still work.

Poison – Does damage over time, where the damage applied per stack per second at level 80 can be calculated to be 84 + Condition Damage * 0.1. Stacks in duration. Your character may exclaim about feeling unwell when this is applied.

Vulnerability – Increases the damage that the target takes by 1%. Multiple applications stack intensity by 1% per application up to a cap of 25. Your character may complain about feeling vulnerable when this is applied.

Weakness – Causes 50% of attacks to be glancing (which reduces the damage output of the attack) and reduced endurance regeneration by 50%. Stacks in duration.

 

 

Condition Damage

Condition Damage is one of the attributes that regularly appears on gear (both major and minor) and can also be increased through traits. However, it only affects 4 of the 11 conditions – Burning, Bleeding, Confusion and Poison. So if your preferred build doesn’t reliably generate lots of these particular conditions, stacking condition damage is going to be pointless for you.

For extra complications, every condition increases by a different amount in proportion to Condition Damage and level.

 

Condition Damage Bleeding Burning Poison Confusion Confusion sPvP

0

42.5

328

84

130

65

100

47.5

353

94

145

72.5

200

52.5

378

104

160

80

300

57.5

403

114

175

87.5

400

62.5

428

124

190

95

500

67.5

453

134

205

102.5

600

72.5

478

144

220

110

700

77.5

503

154

235

117.5

800

82.5

528

164

250

125

900

87.5

553

174

265

132.5

1000

92.5

578

184

280

140

1100

97.5

603

194

295

147.5

1200

102.5

628

204

310

155

1300

107.5

653

214

325

162.5

1400

112.5

678

224

340

170

This may seem really imbalanced, but the durations of the various conditions differs a lot.

 

Condition Duration

Condition duration is useful to everyone who can generate any type of condition, as it will apply to all of them. The game does not round the duration of conditions, but does round the number on the tooltip of the skill to the nearest 0.25s. Since the conditions that do damage per second (Burning, Bleeding and Poison) only do damage on the whole second mark there’s a possibility of being misled with how many pulses of damage you will get. Here’s a video I put together to explain.

(NB: In the video I say that the wiki is wrong. It isn’t. It’s just worded in a way that can be misleading)

So if the only conditions you have are ones that tick per second, there will be optimal steps of condition duration. Realistically there’s too many potential sources of conditions that may change at any time (due to game updates) to get into a meaningful discussion here, as it will be obsolete pretty quickly. What I have done however is note down the standard duration of each source of the big four by profession skills. A summary is below (click here for a condition duration by profession breakdown).

 

Note that not every condition made as a result of a trait is affected by Condition Duration. It is not clear if this is a bug or intended.

Currently the only common sources of increasing overall condition duration besides food are in Traits and Runes. It would be possible to raise Condition Duration to 54% using both of these. After that, it’s possible to increase the duration of a particular type of condition through Traits and Sigils (although there are some bugs in this area). These numbers add up; ie. +30% Condition Duration +20% burning duration = 50% longer burns.

Let’s look at how these durations change

With respect to Bleeding, it is only worth increasing Condition Duration if it results in at least one extra whole second, giving an extra pulse of damage. It is highly unlikely that it’s worth trying to increase your condition duration if most of your condition skills apply durations of 1s. Similarly for 2s and 3s conditions, trying to make 50% and 34% condition duration may just be too much when considering the sacrifices required to other attributes such as Power or Toughness. At 4s, things become a lot easier. An increase of 25% duration is achievable by raising the first trait line to 25, and will result in one extra pulse of damage. Of course this is a balancing act if you’ve gone Precision.

The reason it’s different for Burning and Poison is that they stack in duration. If you apply a 3.5s burn and then apply another 3.5s burn before it ends, the durations will add up to a 7s burn with 7 pulses of damage (even if the burns come from different people). If you wait for the first burn to end before applying the second, you’ll only get 6 pulses of damage. If you have lots of sources of these types of conditions, you may be able to get more pulses than you should by applying them fairly quickly. Of course, if you only have one source, or it’s really short, you may be out of luck.

It’s not all about these 3 though. The other 8 conditions will all be of use to you if they last a little bit longer. An extra .25 duration on Vulnerability may allow you to get a good hit on an enemy, or with Crippled allow you to catch up to them to snare again. Confusion is an odd one because on the one hand it looks like you get no benefit until the person with it attacks. But if you force an enemy to not do anything for 5s because you put Confusion on them, that’s 5s you can pummel them to your heart’s content. And if they do fight back, they get a slap in the face!

 

What happens at 25 stacks?

Conditions that stack (Bleeds, Confusion and Vulnerability) will stack up to a maximum of 25 applications. Any more that are applied are ignored until one of the existing applications ends.

When you have conditions applied to you from multiple sources they go into the same stack dependent on type. The same thing happens on enemies. So if you and someone else in your team can both apply 25 stacks of bleeding and try to, one of you is wasting potential damage from your conditions. This is why it may not always be a great idea to try to build a bar that can push so many of one type out. Having a bit of variety helps you work better in a team. Bleeding is the usual one where people clash.

It will also help if condition removal is used as most only remove certain conditions or a certain number of types.

 

What happens when players with different amounts of condition damage attack the same target?

Short answer: pretty much what you expect.

If you and someone else with different amounts of Condition Damage both apply a condition that stacks in duration on the same target, the durations will add and the ticks from the first source will all fire before the ticks from the second source do. Where durations go over the top of the second (ie 6.5s + 6.5s = 13s = 13 pulses) the Condition Damage of the source that made it go over the top of the second counts (the second person in this example).

For conditions that stack, each application acts independently and times out as expected.

 

What about compared with other attributes?

Damage from conditions is an unstable source because it can be removed. Many condition removals will remove the entirety of one type of condition, so if you’re fighting someone who uses it just after you’ve applied all your bleeds you’ll lose that damage source until your skills recharge. You can help mitigate the effect of a condition removal skill by not focusing on one type of condition.

Of course, you’ve still got the use of your weapons and other attacks while the conditions are working their magic, so Power and Precision are useful even in a condition heavy bar!

 

So which should I prioritise between the two?

This is a rough and ready step through the decision making process. There will be exceptions to the steps below, but this should give you some guidance.

1. Count your condition types and their raw durations using this table below (remember to take off armor, traits etc. You may find it useful to go into the PvP lobby).

Duration (s) Bleeding, Burning, Confusion and Poison Other Conditions

1

2

3

4

5

6

7

8

9

10

11

12

20

30

2. If you have less than 3 sources of conditions, either tweak your build or stop here. Conditions won’t have enough of an effect on your build to be worth dedicating gear decisions to.

3. Look at the number and duration of sources of Bleeding, Burning, Confusion and Poison you have. If around 60% of them are under 4s in duration score a point for Condition Damage. If around 60% of them are over 4s in duration, score a point for Condition Duration. If you’re roughly equal, score a point for both.

4. If you have more than 3 sources of other conditions, score a point for Condition Duration.

5. You should now have a tally that will give you an idea if Condition Damage or Condition Duration is more important to you. If you’re tied, they’re both useful!

 

In conclusion

Conditions are a very useful part of any build and their proper use can increase the overall damage output of a team. To get the most out of them, you’ll need to give some thought to the sources you have in your build. Don’t always trust the tooltips!

 

Sources

To make this, I used the ever useful Guild Wars 2 Wiki, this reddit thread, A guide to Bleeding by Boons and Conditions and the game itself. There was lots of testing – for science!

Thank you to my guild mate fafq for being a guinea pig for some experiments, and to Lissahkins for proof reading this article.

 

Feedback

As always if you have any comments, please leave them down below. Please also let me know if you found this article useful!

Karma is a Bitch

 Posted by on August 31, 2012  Guild Wars 2, MMO Gaming, Oh shit maths  Comments Off on Karma is a Bitch
Aug 312012
 

If you’re a European player, chances are you probably missed the fireworks last night from ArenaNet’s latest status update that rolled into the early morning. Not only have the developers been busy coping with the popularity of their game (to the point that they’ve stopped selling digital copies on their own website) but they’ve also been dealing with account hackings, botting and exploits.

Most of the drama last night came from the permanent termination of around 3000 accounts that seriously engaged in exploiting a karma merchant who offered items for 1 thousanth the normal price. 1000 more will need to ride out a 72 hour ban for giving in to temptation but not taking it to extremes. There’s been a lot of debate on Reddit and Twitter about if this approach is right or wrong on an individual basis, but I’m not sure anyone has considered the big picture.

I had a quick total up the replies to all the people who posted their character names on Reddit asking how many times they exploited. The answer: 15,028 times from 55 people. If we take the average number of times someone exploited and multiply that up by the number of perma banned accounts, that’s 819709 items (estimate). That is a huge number of items especially considering only 1912 (estimate) should have been possible to purchase with that quantity of karma.

That large number of items would have had an impact on the economy. What’s more disturbing perhaps is that this represented 0.3% of the total playerbase. Imagine if the trading post and mail system hadn’t been down, or if people hadn’t spread the exploit so publically. This could have taken far longer to identify and the damage would have been done and spread. I don’t know about anyone else, but I’d love to be playing Guild Wars 2 in 4 years time without a messed up economy and with the developers being able to concentrate on making their game better rather than chasing those who try to break it.

I’m personally with ArenaNet in handing out the bans (even if they can now be revoked). It’s a clear message to anyone who exploits that they risk their account in the future. Just as the temporary bans for hate speech and bad language are having an effect on the game, so will this. Here’s to a friendlier, more honest player base.

Aug 062012
 

In a few weeks we’ll be stepping foot inside Tyria again for real. One important decision everyone will need to make is what server to call home and I’ve set myself the challenge of researching the communities on each of the Guild Wars 2 servers. So after the final beta weekend I asked members of the Guild Wars 2 community for the experiences of their servers. The response was more immense than I could of hoped for. I can only offer my heartfelt thanks to everyone who took the time to fill in the survey and help their fellow players (and myself) out!

 

Gathering information

Using the previous survey as a base and building on the areas that the survey fell short on last time, I built up a short set of questions covering information that included the number of guilds people were a member of, how satisfied they were with their choice of server, their experiences on their servers and the languages spoken on them. The existence of the survey was promoted to my followers on Twitter and retweeted several times by individuals and fansites within the Guild Wars 2 community. It was also promoted on Guild Wars 2 Guru, the Guild Wars 2 sub-Reddit and Guild Wars 2 Nederlandstalige Community.

As with the last survey, I have included every response I received even where I felt that response may have been deliberately misleading, and removed offensive words and identifying comments from the feedback. Again, the questions, original responses and some statistical analysis is available in a pdf for you to enjoy (if you enjoy that kind of thing). I’ll refer to some feedback below, and draw out the more interesting statistics.

General comments

There was a far larger response to this survey than the previous two undertaken. Combined with the fewer number of servers provided for the final Beta Weekend event, the responses will give a better indication of what the communities and experiences of people on those surveys were actually like. However, caution must be undertaken when applying the findings of this survey. The ArenaNet blog post What’s New in the Final Guild Wars 2 Beta Weekend Event indicated that “hundreds of thousands of players” were expected to take part in the final event, and the 365 responses are a very small sample of that expected amount.

That said, only 2 servers gave no response at all and there were 13 servers who gave 10 or more responses. This makes this the most reliable of my surveys so far and I am hopeful that people may find the information contained within it useful when it comes to choosing their server for release. However the variety of comments from the same server means that the results must be taken with a large pinch of salt!

 

Game Formats

There are no large differences between the two regions when it comes to the main formats people were playing over the final beta weekend. Overall 83% of responders said they spent most of their time this weekend in PvE, compared with 12% in WvW and 5% in PvP.

When asked about their experiences in the different formats, some interesting differences emerged. There was some irregularity between those who answered “Did Not Play Format” between the 3 questions (for example, some who indicated they did not play a format in one question gave their experiences of that format in the other two). Overall nearly 98% did some form of pve, while only 35% did some form of sPvP. Those on European servers are more likely to do some form of PvP than American Servers – over 7% difference in both sPvP and WvW.

 

Guilds

There was a large difference between the number of people who chose not to be in a guild on the American and European servers, with 13% more Europeans being a member of one or more guild(s).

Across all servers, of those in a guild 79% of those who responded were in only 1 guild. This was slightly lower on American servers (75%), who had proportionally more people in 2 (17%) or 3 (5%) guilds than Europe (14% and 1%).

 

Countries & Languages

Perhaps unsurprisingly, there were 10 more countries represented on European servers (25) than on American servers (15). In addition there were 9 more languages seen on European servers (17) than on US servers (8). This diversity showed through in comments about server communities, with some responses expressing frustration at the number of different languages spoken on their chosen server. Some also commented about the overflow servers being grouped together, which added to linguistic frustration. In general these comments did not occur alongside reports of negative server satisfaction.

Even though there was a large variety of languages spoken on the various European servers, each server did seem to have a primary language. English was spoken to some extent on all servers.

 

Sense of Community Throughout the Formats

Trying to measure a sense of community is extremely tricky, but the responses from 3 questions broken by format were used to attempt to get an impression of moment to moment life on a server and in the two regions. The amount of text in the various formats gives an indication of communication and resurrection gives an indication of symbolic interaction while perceived co-ordination gives an indication of working towards a common goal.

By format overall PvE scored highest on the sense of community factors, with more positive responses to the quantity of text communication, resurrection and overall coordination. WvW was positive overall, but less so than PvE. However, PvP scored quite low on the text and resurrection factors, and less positively in cooperation. This may be in part due to the fast paced nature of PvP compared to the other two formats, along with its quick resurrection timers.

The differences in responses between the two regions were highest in PvP, with a less negative reaction overall in text and cooperation on US servers while EU servers were slightly ahead with resurrection. Overall the differences don’t give a clear indication of one region scoring higher in sense of community factors.

 

Community

Just as with the second Beta Weekend, the comments received on community are pretty hard to analyse objectively. Generally speaking the comments seem positive about the communities on all servers, with most containing some form of positive and negative experience. Few of the comments were entirely negative about their server. A number of people commented that they spent most of the weekend playing solo and some of these highlighted a lack of community outside a guild. Others said they didn’t notice any community on their servers at all, which was countered by rare stories of people going significantly out of their way to help their server mates.

Amongst the various comments, there were a few that spoke of the community in Guild Wars 2 being better than the community in other MMOs, but some of these also mentioned that they felt the community was worse than it had been for Beta Weekend 2. A few comments expressed frustration at being forced to move servers that they had been on for the last 2 weekends to the detriment of the community. Amongst European servers, there were comments of frustration about the number of different languages being used on overflow servers.

Compared to the community responses for Beta Weekend 2, there were more comments relating to trolls on the servers and complaints about being forced into overflow servers.

 

Satisfaction and Server Population

When it comes to server population, there are a few relatively minor differences between the two regions. Those on American servers expressed a higher preference for their servers to be busier than their European counterparts (35.75% reported they’d prefer the servers to be “A little busier” or “A lot busier” compared to 30.64% in Europe). Conversely 15.59% of those who responded from European servers would have preferred their servers to be “A little quieter” compared with 6.70% of those on American servers. Both regions had roughly the same percentage of responses that expressed “No strong feelings” about their server populations.

The overall results between the two regions when it comes to server satisfaction are roughly equal, with there being slightly more people who reported being unsatisfied with their server choice on US servers. 5% more of their counterparts on European servers reported being satisfied with their server choice.

 

Conclusion

When it comes to choosing a server for release, it’ll be up to you to consider what the most important factors are to your in game happiness. The overwhelming outcome of this survey is the variety of differences between all servers in terms what formats the majority of the population play, the language base and the sense of community on the server. Hopefully it’ll be of some use when people come to make their decisions.

It’s unfortunate that some contributors commented that the sense of community had deteriorated from Beta Weekend 2 to Beta Weekend 3 (the final beta weekend), but this is in line with expectations. Further analysis will need to be done to see if there is any statistical evidence to back up these reports. However there did seem to be a better quantity of servers for the final weekend, especially for US servers. Given the comments regarding the language problems on overflow servers in Europe, I think it may be a good idea for ArenaNet to investigate how they are grouping the servers together to try to provide the most harmonious output.

 

Your Feedback

If you’ve found this information useful to you, please do drop a note in the comments below. Currently I have no plans to do any more of these surveys, but if I get enough feedback from people saying it’s worthwhile doing, I will.  Likewise, please contact me if you have any feedback you’d like to give about the content of the survey itself.

Jun 292012
 

Escort the caravan. Rescue the princess. Kill ten rats. Cliches, everyone one. Aren’t you sick of them? Don’t you wish that developers would drop them and do something different?

This actually has nothing to do with Guild Wars 2 or any other particular MMO. It’s a general problem with all rpg games. In order to progress through the game you have to have your character do something, and there’s only so many different types of somethings you can do.

If you weren’t tasked with escorting an npc from one place to another, rescuing a damsel in distress, collecting a number of items, or taking out some targets, what would your character be doing in the game? And, if you did come up with something else to do that didn’t ultimately fall into one of those 4 categories, how long before it became as old and stale to you as those have?

Ultimately there’s only so many types of things that developers can do with a world, especially one that very closely mimics our own. We collect groceries from the supermarket, go to find our lost pets, go paint balling and take our less able family members (read: children and grandparents) out to the shops or zoo because they can’t go by themselves. That in game activities reflect our real life ones may be why there are frequent complaints about a certain quest type in a given rpg; people want to escape that world, rather than jump into another that copies it. But what are the alternatives?

Maybe there aren’t any and maybe we as a gamers should stop rolling our eyes when we see “yet another” escort quest and instead take a look at how those quest types are executed. Our next assignment may be to kill ten rats, but what if those rats are smart and are going after an npc vital to your character? Sure, not every task undertaken can be an epic battle for survival otherwise we’d grow unappreciative of the scale of the task, but execution is key. Make us care, determined to succeed and analyse the failures to rise another day.

Games may not be able to generate the same amount of emotional involvement as some movies because of the way they are played, but that doesn’t mean every session can’t be filled with a tug on the heart strings and be full of entertaining and memorable moments. In my opinion there’s nothing wrong with the notion of saving the damsel in distress – just make us give a damn first.

Jun 212012
 

After the success of the first community survey following the first beta weekend, it was time to do it all over again! Just like last time, I asked members of the Guild Wars 2 community for the experiences of their servers to gather data that would help people to pick their server and to generally capture the state of the in-game Guild Wars 2 community. And everyone really delivered (which is why this is a little later than I’d have liked)! Thank you to everyone who filled in the survey, encouraged a friend to or posted it somewhere for people to see. You are the backbone of this community.

Gathering information

Using the previous survey as a base, I built up a short set of questions covering information that included the number of guilds people were a member of, how satisfied they were with their choice of server, their experiences on their servers and the languages spoken on them. The existence of the survey was promoted to my followers on Twitter and retweeted several times by individuals and fansites within the Guild Wars 2 community. It was also promoted on Guild Wars 2 Guru and Guild Wars 2 Nederlandstalige Community.

As with the last survey, I have included every response I received even where I felt that response may have been deliberately misleading, and removed offensive words and identifying comments from the feedback. Again, the questions, original responses and some statistical analysis is available in a pdf for you to enjoy (if you enjoy that kind of thing). I’ll refer to some feedback below, and draw out the more interesting statistics.

Results

In total the survey was completed by 220 people, 123 from European servers (56% )and 97 from American servers (44%). This was over 3 times the number of direct responses I received last survey, and nearly 100 more total responses. However while were significantly more direct responses to the survey this time around to last, there were also twice as many servers as last time as well. Of the 96 servers for the second beta weekend, I received a survey from one or more residents of 63 servers. Responses were not received from 14 servers from Europe and 19 servers from America.

3 servers had more than 10 responses US Eternal Grove, EU Far Shiverpeaks and EU Desolation. These servers with 11, 37 and 11 responses respectively had a significant effect on the mean number of responses per server which was 2.292. This is compared with a median of 1 and a mode of 0. If these results are excluded, the mean is 1.731 – much closer to the median result. Unfortunately this means that once again we can’t draw any concrete conclusions about the state of any one particular server off the back of these results – one person’s response can’t give a full and accurate account of the entire community on any server on the list.

The reason for these servers producing such large results are likely to be down to the platforms that the survey was promoted on and the populations that had chosen to inhabit those servers. For example, during the first beta weekend, US Eternal Grove was designated the server for a large number of independent fan site leader and bloggers, several of whom follow my twitter account and are more likely to participate in community projects. In addition, EU Far Shiverpeaks was designated the server for the Dutch and Scandinavian population in both beta weekends – since GWNL is a Dutch language fansite, it’s likely that Dutch people would be resident on this server and hear about the survey. The reasons behind the reply spike from EU Desolation is less clear, but information about communities and the servers they had chosen from the first beta weekend suggests that the gaming websites Rock Paper Shotgun and PCGamer were choosing that server along with a few guilds. However, it’s not clear why that particular server got so many replies when compared to EU Gandara (for example) which was the designated Guild Wars 2 Guru server.

Server populations and satisfaction

While  I’m hesitant to try to draw any conclusions from individual server, there are some interesting trends across the group. For example, 85% of replies were of the opinion that the server was appropriately populated or could have been a little bit busier. Less than 1% of replies felt their server could have been a lot quieter. There wasn’t a large amount of variation between European and American Servers. This may indicate that if more servers need to be added to support the expected population for the third beta weekend, then adding less than 48 may be appropriate.

As to server satisfaction, overall 62% of those who responded said they were either satisfied or very satisfied with the server they had chosen to be on. 16% responded that they were either unsatisfied or very unsatisfied with their choice of server. It was noted that several of those who responded that they were very unsatisfied had said they had been separated from guild members or friends because of servera being full, and that was the reason for their lack of satisfaction.  Proportionally, European servers had more responses indicating players were very satisfied than American servers while American servers had more responses indicating players were satisfied than European Servers. The reason for this difference is not clear.

Languages

Respondents were asked to list the languages that they saw being spoken on their servers. A language was counted every time it was mentioned by a different replier for a server. For example, if 5 replies mentioned English as being spoken then that language was given a score of 5, whereas if 3 of those mentioned German, then that language was given a score of 3 for that server. The result is a weighting of languages for the particular server. Some “languages” were disregarded for not being officially recognised languages (eg “World of Warcraftish”).  While analysing the results in this manner, it was noted that several people had trouble distinguishing with certainty between Danish, Swedish, Finnish and Norwegian, so the decision was taken to group them into a Scandinavian category.

Perhaps unsurprisingly given the language of the survey and the target audiences for Guild Wars 2, English got the highest overall score being mentioned a total of 209 times. After English, German, French and Dutch were the most frequently seen languages. In total 17 languages were noted.

Some responses expressed frustration at certain linguistic communities residing on a particular server and this information not being readily available to those choosing which servers to choose. In other words, there’s work to be done to increase the profile of lists that note which communities are choosing to reside on a particular server so they can join or avoid linguistic communities as appropriate.

Guilds

The survey asked people to disclose the number of guilds they were a member of during the second beta weekend, ranging from 0 to 6 or more. All but 4 of the respondents replied they were a member of 3 guilds or less, and the remaining 4 replied with 6 or more. The lack of replies for 4 or 5 guilds raised suspicion for the validity of those results for me, especially since 3 of those didn’t leave valid contact details. The one I did contact said they were playing around with the guild system as there were no penalties to doing so.

Overall, 58% of people who responded are a member of 1 guild, with 22% electing not to be a member of any guild. Only 18% took advantage of the ability to be a member of more than one guild. This suggests that in general, people will only be a member of one guild when given the choice. Interestingly, there is a difference between those on European and American servers. On European servers 63% of people are a member of one guild, while on America servers 52% of people are in one guild. Conversely, 16% of respondents on American servers are in 2 guilds compared with 12% on European servers, and 25% of those on American servers are not in a guild, compared to 19% in European servers. The number of people in 3 or more guilds did not significantly differ between server groups.

Overall 75% of those questioned chose to be in a guild were only a member of 1 guild.

Communication

A large amount of feedback from the first survey contained comments about how infrequently players communicated by text to others nearby. Some of the frustrations may have been down to technical problems with the game itself, but this time around I wanted to try to capture any difference in experience with text communication. The responses were reviewed and categorized into a high quantity of text, some text (or an unclear result) and none.

66% of those who responded to the survey said that their server had a high level of text communication. 20% responded that they either didn’t pay attention to text communication, registered a moderate level or encountered differing amounts of text communication in various areas of the game (such as PvE vs WvW). Many commented that there was more communication in WvW than other areas of the game, and that in all areas of the game people tended to thank others that resurrected them. Others commented that there seemed to be more communication via text than the last beta. There wasn’t a lot of difference between responses from both sides of the Atlantic.

Coordination

Across all servers, 44% of people responded that they considered players on their servers to be coordinated with each other, however there were a large number of comments about a difference between WvW (more coordinated) and PvE (less coordinated).  There was around 10% difference in the amount of responses that counted coordination as being high between European and American servers, with European servers rating higher. American servers recorded around 40% with no clear view compared to 37% for European servers.

There is a great deal of ambiguity around what counts as coordination though and what it actually means in different areas of gameplay. This is a question I’ll be looking to improve for the survey following the third beta weekend.

Resurrection

Perhaps the most encouraging of all the data collected, 95% of respondents said that they felt that they and others were resurrecting players all, or nearly all of the time. Less than 1% felt no one on their server was being resurrected. There was no significant difference between European and American servers. As this can be a considered an example of symbolic exchange between players, this is quite encouraging to helping foster a positive sense of community across all servers.

Community experiences

These seemed to be a really mixed bag of responses and when it came to analyse them into some form of statistical information, I found it hard to do it in any way that would be really meaningful. As a result this will be a question I’ll be looking to improve for the survey following the third beta weekend.

While reading the replies though, a few things stuck out for me. One was the number of people who said they had really positive experiences of the community on their servers, saying they were better than either they expected, or other mmos they had played. Some commented on the lack of rudeness, swearing or similar in general chat, and the eagerness of others to answer questions. Others however said that a few trolls or similar had had an effect on their experiences, but regardless they enjoyed themselves. Some complained of a lack of population.

Comparison to the first beta survey

In the first beta survey there were lots of comments about the overflow server separating friends from one another. The overflow servers were not mentioned once in the 220 responses received. However, some did mention problems with joining friends on other servers and that they got separated from players they would have liked to be with, which was something that was mentioned last time.

Conclusions & further research

As with the results of the first beta weekend survey, there are signs that on the whole that Guild Wars 2 in game experiences are generally positive and some were really impressed by their experiences of their server community. There are no definitive signs that one server is substantially worse than any other. There is more work to do on behalf of the community to inform players of unofficial language servers to help people make educated choices about their server. ArenaNet will need to closely monitor the quantity of servers needed to balance having a lag free playing experience with a well populated world.

I have already received some feedback on the questions in this survey and what others would like to see included in the survey for the third beta weekend, including player orientation (PvE, PvP, WvW or PvX) and country. I’d also like to break down some questions so they’re easier to analyse, while still allowing room for detailed expression of experiences.

If you have any feedback you’d like to provide about the questions that are used in this survey, or would just like to let me know that you’ve found this useful (it does help motivate me to do this again), please drop me a note. Thank you.