Average wait time: Less than 5m.

[Iscah]

Average wait time only shows the time of the person who has been longest in the queue and if you are alone in the queue then you will make it climb to 30 minutes. Average time of 5 minutes means that its very alive or that literally no one is in queue

This. It doesn't seem to actually be an average at all.

If you're the only person in the queue, it will feed your own queue time back to you, updated at three-minute intervals.

It may keep a genuine average if there are multiple people completing the queue and getting into the duty, but if so then the calculation falls over when the traffic flow ceases.

[Claustrum]

This. It doesn't seem to actually be an average at all.

If you're the only person in the queue, it will feed your own queue time back to you, updated at three-minute intervals.

It may keep a genuine average if there are multiple people completing the queue and getting into the duty, but if so then the calculation falls over when the traffic flow ceases.

So what purpose does this estimated wait time serve if it says I'll have to wait less than 5 minutes on average, but it actually takes longer than 8 hours?

[Iscah]

So what purpose does this estimated wait time serve if it says I'll have to wait less than 5 minutes on average, but it actually takes longer than 8 hours?

It isn't serving its purpose in that case. It is temporarily broken due to lack of data. The calculation doesn't seem to be designed for the scenario where less than one instance-worth of players have successfully queued in whatever span of time it keeps track of.

Knowing it does that makes the difference between it being "broken because it's broken" and broken in specific circumstances. It works well for things like roulettes or busy duties, not so much for low traffic areas.

I assume it was designed in the early days of the game where people were less spread out, or (as with many things in this game) they simply failed to consider the possibility that people might not run it constantly.


To be clear, I'm not making excuses for how it functions, just explaining why I think you are getting those results from it.

[Claustrum]

It isn't serving its purpose in that case. The calculation doesn't seem to be designed for the scenario where less than one instance-worth of players have successfully queued in whatever span of time it keeps track of.

OK!

So why does the game tell me it'll take me less than 5 minutes to find a group to do an activity, yet it takes more than 8 hours to find that group?

[Iscah]

OK!

So why does the game tell me it'll take me less than 5 minutes to find a group to do an activity, yet it takes more than 8 hours to find that group?

You may not have seen my longer edited post above.

Basically it is broken, but predictably broken in some circumstances, while working correctly in others.

I am not justifying this, only observing it.

[Claustrum]

You may not have seen my longer edited post above.

Basically it is broken, but predictably broken in some circumstances, while working correctly in others.

I am not justifying this, only observing it.

But you're defending the current system which gives average wait times which don't reflect the actual wait times of players. It kinda sounds like you're justifying it.

[Iscah]

But you're defending the current system which gives average wait times which don't reflect the actual wait times of players. It kinda sounds like you're justifying it.

Which part of what I've written is defending it, rather than simply explaining how it currently works (or fails to work)?

If they want to fix it, great.

[Hasrat]

OK!

So why does the game tell me it'll take me less than 5 minutes to find a group to do an activity, yet it takes more than 8 hours to find that group?

I'd say here lies your own misinterpretation.

The system isn't telling you how long you will wait to find a group.

The system is telling you how long you will wait in queue. There's no guarantee it will find a group.

[LilimoLimomo]

So why does the game tell me it'll take me less than 5 minutes to find a group to do an activity, yet it takes more than 8 hours to find that group?

I believe this is the core issue: you have misinterpreted what "Average Wait Time: Less than 5 minutes" means. It does not mean "you will have a group in less than 5 minutes."

Because there's nothing that the devs can do to tell you how long you will be in the queue. To do that they would have to predict the future. Specifically, they'd have to know how many people would join the queue, and when, and whether they would abandon or not, etc. There's simply no way to know this.

So they try to give you the next best thing: the average wait time. But remember, the average wait time is not an estimate of how long your queue will take. If when you join the queue there's only 1 other person queued, and they've been waiting for 4 minutes, then that means the average wait time is less than 5 minutes. I hope you can see why that in no way indicates that you'll get into a game promptly. And that's why over time as you stay in that queue without getting a game, the average wait time will generally increase, because your wait time is a meaningful part of that average.

[UkcsAlias]

It isn't serving its purpose in that case. It is temporarily broken due to lack of data.

Thats not temporarily though, its conditionaly broken (as the issue is always there). In most cases it works well because the longest wait time is generaly a good indication (as the longest waiting person is often prioritized, its unlikely to exceed that value by a lot), but if there isnt anyone, it shouldnt use that measurement anymore since its plain incorrect. In such case it should either resolve to the last time the duty was started, or indicate it simply cannot estimate (forming party is normaly my indication for that, as otherwise you would get a waiting position).

I think it should in these cases resolve to the previous run and check how long the queue was there. For day/night it might still vary a lot. But if it then already states 2h, if it ends up being 3h it at least makes sense (and the inverse then also works). Even if the queue isnt full, if it states it takes generaly 2h to fill, either i will be willing to wait (not), or i will wait until its closer the estimated time it normaly would start, and then queue in the hopes for more people. But yeah, this very issue is the reason why im not going to queue for any alliance raid outside of peak hours (which then can still easily take 30-60minutes of queue, while it states 5min average...).

If they realy want to take it further, they could make it so you can double queue until it reaches a remaining estimated wait time of for example 30mins (normal dungeons generaly wont take this long). It then will at least be able to indicate a better number of people that are waiting, and since there is a 30min treshold, if the duty would start, it doesnt matter as much that its 15min delayed because someone was already in a duty. 15mins in an already 2h long wait is only a small amount. And this would at least promote more people to queue, which is helpful in general.

Queues are one of the biggest problems in MMO's, so any frustration removed here would be a major QoL improvement.