How does ANET Measure Population in relations to WvW?

[Sreoom.3690]

Are they looking at percentage of a server's population that play WvW 25% or more of their time?

or

Are they just measuring how many people are on a server regardless of their WvW players?

If it's the later they are looking at the wrong measure.

[Kate.3679]

They decide based on the amount of hours the server's WvW population plays

[Svarty.8019]

1 hour ago, Sreoom.3690 said:

How does ANET Measure Population in relations to WvW?

Simple. It doesn't.

[Xenesis.6389]

1 hour ago, Sreoom.3690 said:

Are they looking at percentage of a server's population that play WvW 25% or more of their time?

or

Are they just measuring how many people are on a server regardless of their WvW players?

If it's the later they are looking at the wrong measure.

They measure only by the people who step foot in wvw for the week, mostly by play time.

They use to measure by the amount of all players on a server, but that was changed like back in summer 2015.

[TheGrimm.5624]

1 hour ago, Sreoom.3690 said:

Are they looking at percentage of a server's population that play WvW 25% or more of their time?

or

Are they just measuring how many people are on a server regardless of their WvW players?

If it's the later they are looking at the wrong measure.

 

To Xen's point:

https://forum-en.gw2archive.eu/forum/game/wuv/World-Population-Changes-Are-Coming/page/1

https://forum-en.gw2archive.eu/forum/game/wuv/World-Population-Changes-Are-Coming/page/4#post5324934

https://forum-en.gw2archive.eu/forum/game/wuv/World-Population-Changes-Are-Cong/page/5#post5326517

I thought I had in past actually seen the algorithm in the wiki but not finding it currently and they might have removed it along the way. 

 

 

[Morden Kain.3489]

1 hour ago, TheGrimm.5624 said:

I thought I had in past actually seen the algorithm in the wiki but not finding it currently and they might have removed it along the way. 

Probably because they realized it was broke as <bleep>, or not working as intended. 

All I know is that the server populations listed are not what I see in WvW.  A "full" server should have at least 2-3 maps queued 2+ days out of the week, not just the weekends.  If their entire formula is based around what is happening from Friday @ reset into Sunday night 10p server time, they are completely off the mark for the rest of the week.  That particular time frame a person can go to Diamond Pip Chest, and have recycled Diamond 2+ times (depending on how dedicated they are).  However, that is not indicative of the entire week.

[Longpoke.8290]

4 hours ago, Sreoom.3690 said:

Are they looking at percentage of a server's population that play WvW 25% or more of their time?

or

Are they just measuring how many people are on a server regardless of their WvW players?

If it's the later they are looking at the wrong measure.

I've been saying the same thing. 

I believe that Gw2WvW should be a separated game. With servers that are 100% for wvw population, no pve content. 

[roederich.2716]

Y they take game hours.

the server with most hours is used as reference and gets status full and stays full until they change it.

all other below are assigned their population status is comparison to the reference then.

Dont know why they do this relinkin though every 8 weeks they simply could say „ay listen wvwers… from now on we dont do any relinks for you anymore, if you want balance transfer yourselves by the transfer system and shut it.“

Would actually be good for this wvw community which in terms of networking and organisation got lazy and thinks anet is responsible for the competition in a population based 24/7 war scenario…

Lots of crybabies who actually are overdue for a buttkick to get moving again.

Edited November 16, 2022 by roederich.2716

[Sahne.6950]

They are asking Mabi.black

[Karagee.6830]

15 hours ago, Svarty.8019 said:

Simple. It doesn't.

Ding ding ding we have a winner

[Karagee.6830]

13 hours ago, Morden Kain.3489 said:

Probably because they realized it was broke as <bleep>, or not working as intended. 

All I know is that the server populations listed are not what I see in WvW.  A "full" server should have at least 2-3 maps queued 2+ days out of the week, not just the weekends.  If their entire formula is based around what is happening from Friday @ reset into Sunday night 10p server time, they are completely off the mark for the rest of the week.  That particular time frame a person can go to Diamond Pip Chest, and have recycled Diamond 2+ times (depending on how dedicated they are).  However, that is not indicative of the entire week.

Allegedly Gandara is the reference for EU and so always closed. Amount of queues on Gandara? 1 map on Friday or Saturday when there is more than 1 tag on. If we're lucky. Generally, no queues ever. Even last relink with FR we almost had no queues.

During the beta week I had queues on 4 maps at prime time (including 60 on EB), 3 during the day at the weekend and 2 maps throughout the week. The overall amount of players on EU did not change, so it means that this situation must be normal for some other teams in the current system...and it's direct damning evidence that their algorithm is simply garbage.

And yes, Gandara has a disproportionally higher participation at the weekend and specifically Friday and Saturday evening. After that we're outnumbered even by servers that are tanking on purpose.

Edited November 16, 2022 by Karagee.6830

[Chaba.5410]

I don't understand why people get so confused about playhours and queues.  The "population" status of worlds is based on playhours, or number of hours a player spends in WvW.  Queues are based on actual players trying to get on a map and have little to do with playhours.

Playhours reflects coverage.  If your server is full of grinders who PPT and Ktrain all day and inflate the playhours, that's not anyone else's problem.  If your server was overstacked with a PvE population prior to the algorithm change, you have more casuals than anyone else who can suddenly enter WvW and add playhours.  World restructuring will remove those casuals as servers are removed.

Edited November 18, 2022 by Chaba.5410

[Aeolus.3615]

They dont, if a server has ktrain during day and other server during night it is OK for Anet since both have similiar play time, but both never fight hard against each other.

 

Workign as intended, i suspect alliances will still be the same since Anet will never do checks on timezoes peaks.

[Xenesis.6389]

Anet probably wouldn't even need to check on timezones either if players manage to organize themselves into alliances with round the clock coverage.

 

It's funny how the very first alliance in the game all the way back in 2012 showed the flaw of the system, but also showed how to win the game, yet barely anyone has managed to replicate it in 10 years, it might come back after WR if the stakes are high enough for players to bother. But most players don't care for anything else other than their own time of play is stacked.

[Karagee.6830]

29 minutes ago, Chaba.5410 said:

I don't understand why people get so confused about playhours and queues.  The "population" status of worlds is based on playhours, or number of hours a player spends in WvW.  Queues are based on actual players trying to get on a map and have little to do with playhours.

Playhours reflects coverage.  If your server is full of grinders who PPT and Ktrain all day and inflate the playhours, that's not anyone else's problem.  If your server was overstacked with a PvE population prior to the algorithm change, you have more casuals than anyone else who can suddenly enter WvW and add playhours.  World restructuring will remove those casuals as servers are removed.

There is no confusion. There is the reality that Gandara that has no coverage morning and afternoon and we also are outnumbered 5 evening/nights a week. So our playhours, as you put it, cannot logically be higher than many other servers who do have queues AND have very strong PPT in the morning/early afternoon. Now, if they do not count exactly total playhours, but perhaps only specific time-slots or each time-slot is weighted (so prime time will have a higher weight) then it may be possible that we appear larger than we actually are. However as we have no queues at prime time, it just shows we don't hit the map cap and, whatever that is, it's a hard number that is the same for every team. I experienced 15x, 20x the amount of queues during the beta than I do during my normal WvW play and that may be an understatement.

[Karagee.6830]

18 minutes ago, Xenesis.6389 said:

Anet probably wouldn't even need to check on timezones either if players manage to organize themselves into alliances with round the clock coverage.

 

It's funny how the very first alliance in the game all the way back in 2012 showed the flaw of the system, but also showed how to win the game, yet barely anyone has managed to replicate it in 10 years, it might come back after WR if the stakes are high enough for players to bother. But most players don't care for anything else other than their own time of play is stacked.

You don't need coverage around the clock. You need to:

a) outnumber your opponents in the majority of time slots (so generally off-peak since peak has hard caps in place and it may be literally impossible)

b) crucially greatly outnumber the opponents before the start of the off-peak (so late night in the last populated time zone) so you can upgrade an insurmountable amount of objectives on multiple maps, while the opponents have a bunch of paper stuff and whoever is playing during off-peak has an easy time maintaining the status quo, defending T3 towers and keeps and flipping paper targets.

On EU an alliance with equal numbers and zero coverage at prime time would just easily win every match with the current scoring system.

Edited November 18, 2022 by Karagee.6830

[Xenesis.6389]

Oh so tiresome.

 

Quote

You don't need coverage around the clock

Quote

outnumber your opponents in the majority of time slots

 

That's what having coverage means.

 

 

 

[Morden Kain.3489]

1 hour ago, Chaba.5410 said:

Playhours reflects coverage.

This is not entirely accurate.  It is just "playhours"... as in time played.  Take the following as examples for a single map and a "single" server:

(50p x 24h x 3d) + (15p x 24h x 4d) = 5040 "playhours" (p = Players; h = Hours; d = days)

(30p x 24h x 7d) = 5040 "playhours"

In the above examples, the 30 players, rolling 24 hours a day, for 7 days has just as many "playhours", yet would be considered to have better coverage.  Therefore your "reflects coverage" is not wholly accurate.

I do agree that a queue does not mean playhours.  They are two different items of measurement.  However, a queue can be considered a reflection of playhours for the time slot it is happening (ie: Reset).

@Chaba.5410 Please explain a bit more in depth what you mean by "Playhours", or did I get the gist of your meaning in the above math?  Thank you.

[Kal Durak.2830]

WvW needs a big balance patch for population NOT based on PVE. 

[Riba.3271]

Biggest flaw with server system is how 2 servers together ruins whole population balance and transfer costs. You see full status is defined by highest population server, any server that has close to same population is marked full:  If 2 servers are linked, full status does very little. It still allows people to make Full + big server combinations every 2 months and beat everyone. In addition to this, most good main servers being full or expensive, it forces people and guilds to transfer to links. This puts the population cap lower and lower overtime and the system breaks since lot of servers have so many daily hunters that they can't fit any real WvW players anymore.

[Karagee.6830]

3 hours ago, Xenesis.6389 said:

Oh so tiresome.

 

 

That's what having coverage means.

 

 

 

It seems some guys playing on NA fail to communicate in English properly, but maybe you would feel more comfortable with Spanish and BB after all... Which goes to show, language considerations are largely overblown if we're struggling this way!

Having 'round the clock coverage' does not mean 'having coverage'. As I very clearly stated any server with equal numbers and no coverage at prime time would win every match with the current system. That is not round the clock coverage and it was as clear as it could possibly be.

Edited November 18, 2022 by Karagee.6830

[Karagee.6830]

22 minutes ago, Riba.3271 said:

Biggest flaw with server system is how 2 servers together ruins whole population balance and transfer costs. You see full status is defined by highest population server, any server that has close to same population is marked full:  If 2 servers are linked, full status does very little. It still allows people to make Full + big server combinations every 2 months and beat everyone. In addition to this, most good main servers being full or expensive, it forces people and guilds to transfer to links. This puts the population cap lower and lower overtime and the system breaks since lot of servers have so many daily hunters that they can't fit any real WvW players anymore.

Death to all PVErs! Oh wait, that includes myself also as I play all game modes except open world.

Oh yes I just remember that kill number 14 in my streak earlier this week was a guy who was idle after capping my spawn camp (I arrived late). Salty whispers ensued because I jumped him and didn't wait until he finished whatever he was doing.

Edited November 18, 2022 by Karagee.6830

[Karagee.6830]

2 hours ago, Morden Kain.3489 said:

This is not entirely accurate.  It is just "playhours"... as in time played.  Take the following as examples for a single map and a "single" server:

(50p x 24h x 3d) + (15p x 24h x 4d) = 5040 "playhours" (p = Players; h = Hours; d = days)

(30p x 24h x 7d) = 5040 "playhours"

In the above examples, the 30 players, rolling 24 hours a day, for 7 days has just as many "playhours", yet would be considered to have better coverage.  Therefore your "reflects coverage" is not wholly accurate.

I do agree that a queue does not mean playhours.  They are two different items of measurement.  However, a queue can be considered a reflection of playhours for the time slot it is happening (ie: Reset).

@Chaba.5410 Please explain a bit more in depth what you mean by "Playhours", or did I get the gist of your meaning in the above math?  Thank you.

Bingo. What Chaba disregarded in his analysis is that there are:

a) structural limits, ie you can't have 300 people on a map and queues are indicative of whether your team is playing at maximum capacity or not. During the beta we had times with 4 queues. 4 queues means you are clocking the maximum theoretical 'playhours' possible at that given time. No queues means you are below that threshold, but does not tell you how much. Still if others have queues and you are not they definitely outnumber you

b) there are other in-game indicators, like the horrible little red icon that says outnumbered, that your population at a given time on a map is significantly lower than the combined enemies. So if you are outnumbered on EB and your home border simultaneously, it's safe to assume you are greatly outnumbered overall and the playhours you are clocking are far below your opponents and I don't know how someone could argue the opposite

Edited November 18, 2022 by Karagee.6830

[Chaba.5410]

4 hours ago, Morden Kain.3489 said:

This is not entirely accurate.  It is just "playhours"... as in time played.  Take the following as examples for a single map and a "single" server:

(50p x 24h x 3d) + (15p x 24h x 4d) = 5040 "playhours" (p = Players; h = Hours; d = days)

(30p x 24h x 7d) = 5040 "playhours"

In the above examples, the 30 players, rolling 24 hours a day, for 7 days has just as many "playhours", yet would be considered to have better coverage.  Therefore your "reflects coverage" is not wholly accurate.

I do agree that a queue does not mean playhours.  They are two different items of measurement.  However, a queue can be considered a reflection of playhours for the time slot it is happening (ie: Reset).

@Chaba.5410 Please explain a bit more in depth what you mean by "Playhours", or did I get the gist of your meaning in the above math?  Thank you.

I wasn't meant to be accurate.  It was meant to highlight the fact that players who play longer hours provide greater coverage over a larger number of skirmishes than players who only play one or two skirmishes.

To adjust your example math, plug in different numbers for your hours variable.

80p x 2h x 7d = 1120 hours (1 map queue every day for 1 prime time skirmish)
1120 hours / (6h x 7d) = about 26 players playing 3 skirmishes every day

Twenty-six players is certainly less than 80 players, but because they play longer hours, they have an inflated value per individual in the world population calculation.  It's not a map queue, but so what?

The calculation is more complex though than simple total playhours.  It's a rolling average of a server's total playhours smoothed out over several weeks so that communities can't easily circumvent it by organizing a mass-log-out for a few days.  Anet also uses WXP gain as a secondary stat to sanity-check playhours.

Edited November 19, 2022 by Chaba.5410

[Chaba.5410]

5 hours ago, Karagee.6830 said:

On EU an alliance with equal numbers and zero coverage at prime time would just easily win every match with the current scoring system.

Exactly.  "Off-hours" players have a bigger impact on score because of the way the scoring system works.  That's why 26 players who play 3 skirmishes every day have an inflated value in the population algorithm over the 80 EU prime time players who play only 1 skirmish a day and queue their map .  It's to match their inflated value in the scoring system.