Guild Wars 2

Currently CD is getting 110 glicko for this weeks win but that’s mostly because DH has died. Next week CD will not be getting anywhere near 100 glicko from the other 2 servers. Because of the volatility changes, they are losing the majority of there glicko this week and will only lose 10-20 each next week making it impossible at the current rate for CD to move up to T3. We are currently 300+ glicko behind SBI/HoD, so even if we continue getting 50 glicko a week from these 2 servers which is unlikely, it will probably be around 40 tops, it will still take 6+ weeks for us to get in range of T3 glicko and that is broken. To compare it to T3, JQ is beating there opponents quite convincingly PPT wise, similar to CD but are only getting 4 whole glicko for it. And Dragonbrand who is winning there Matchup in T2 by a decent amount is LOSING glicko. CD will soon be the same in a couple weeks, barely getting any glicko if this doesn’t change.

A simple solution would be to reset glicko for a week, that way the servers that own will move up to T1, and the servers that struggle will go down to t3/t4. In a couple more weeks it will work itself out and put servers where they belong glicko wise. Please Anet.

Yep, the strength of the four-server-alliance is showing everyone that a glicko reset is necessary to start new linkings. It will have the added bonus of letting servers fight new opponents. If ANet is doing a decent job balancing out the populations then there is no reason not to reset.

IDK about +/- 300 glicko, but that range was my concern also with the linking last time around and I remember some highly improbable matches occurring anyway. Hopefully the volatility and deviation reset is enough to get some changes.

I hope so too, but looking at the T2 and T3 matchups, where the winners are getting -2 and +4 glicko each, it doesnt look probable. Also pls dont delete or merge this thread anet, this is serious.

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current "faster change in glicko", it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

the way the linking works, the ‘host’ world gets all of the rating benefit of a win (or the loss from a loss). linked worlds don’t get any rating change at all. EBay for example hasn’t moved from 1528.6343 since 4/22, even though they are ranked #12, because they’ve been a linked world since then.

so a 4-server linking might be very very powerful, but only the host world will see any kind of rating change as a result — the other 3 won’t see any rating improvement at all. and when the links change in the future, we might see CD fall again as rapidly as they’re now rising.

-ken

Snowreap.5174:

so a 4-server linking might be very very powerful, but only the host world will see any kind of rating change as a result — the other 3 won’t see any rating improvement at all. and when the links change in the future, we might see CD fall again as rapidly as they’re now rising.

-ken

CD isn’t going anywhere with a 300 glicko wall between T3 and T4. By the time CD even get close to move up, the pairings will change.

The big deal is that the other 5 servers locked with the quadserver will have this “super-fun-and-happy” matchup for the duration of the pairings unless CD get manually moved or a T3 server dies off completely.

I went a-googling for some explanation for why WvW uses the ranking system it does, and I didn’t exactly find what I was looking for, but I did find this: https://www.reddit.com/r/Guildwars2/comments/18wb2i/when_the_wvw_ratings_break/

So it appears that this is a long-standing issue with the lowest tier, and not necessarily unique to this particular matchup.

Speaking as someone who spend the last linking-cycle in T3, the matchup on T3 became really stale with the same 3 servers 3 weeks out of the month with FA occasionally falling to replace maguuma or SBI.

HoD coming up would be a welcome change regardless of if we went down to T4 or stayed in T3, since it would be a different matchup.

T4 is currently more active than T1 from my observations across my 2 accounts. Glicko needs to be reset.

“More active than” doesn’t mean “better at winning than”, though.

People have been talking about several things:

• activity level
• population
• timezone coverage
• level of organization
• skill
• ability to win

which are all different things, albeit some interrelated.

Tris Apollumenon.6435:

“More active than” doesn’t mean “better at winning than”, though.

People have been talking about several things:

• activity level
• population
• timezone coverage
• level of organization
• skill
• ability to win

which are all different things, albeit some interrelated.

The fact that there are more kills going kinda stand up for activity… It means there are enemies engaging each other, even if one side is far larger than the others. Hell, all tier have more kills/deaths than T1 right now with the highest being T4 which is over 90k.

I wouldn’t count T4 being more active than T1 naturally. But YB trying to tank out of T1 doesn’t help to make it active.

I love how we have polls on tons of things but when it comes to resetting glicko it is a never asked question.

So many servers have imploded over the years to cause the glicko walls where they are at now.

With the server linking I thought for sure they would reset the glicko since it won’t be accurate. This serve relink there is 4 servers together and still no reset.

Makes no sense but taking out crafting is a priority.

Anet do you realize that terrible matchups are part of the population issue that you are working on? Getting stuck in a bad matchup where everyone (except the matchmaking algorithm) can see a particular server shouldn’t be there is as frustrating as anything else in WvW. You’ve got to do something about this!

The glicko wall is awful.

Honestly the pairing for that match up should have been like this:

CD + BP + ET
DH + EB + K
SF + FC + GOM

Really was no reason to put 4 servers together. Glicko needs reset, I’ve been saying it since the first server pairings.

Snowreap.5174:

the way the linking works, the ‘host’ world gets all of the rating benefit of a win (or the loss from a loss). linked worlds don’t get any rating change at all. EBay for example hasn’t moved from 1528.6343 since 4/22, even though they are ranked #12, because they’ve been a linked world since then.

so a 4-server linking might be very very powerful, but only the host world will see any kind of rating change as a result — the other 3 won’t see any rating improvement at all. and when the links change in the future, we might see CD fall again as rapidly as they’re now rising.

-ken

Ding ding.

You’ll notice that BG is not nearly as effective at beating the other T1 opponents — PPT wise — which also indicates that ET is much much larger and should be much much higher in glicko rating.

I’m on ET and have been for all of last match up as well.

Regarding the Quad-server domination on kill-to-death ratio, it’s because during primetime (and I know it happens on other times too) its basically a non-stop zerg v zerg fight happening on EB – I’m not counting anymore cuz I’m over 2K+ rank but I know I’ve opened 10+ rank up chests in less than 3 hours a couple of times this week (that’s mostly from killing, not objective capture). Quad commanders with PUGs and WvW guilds in tow specifically bate — we go zerg hunting — typically its 2 commanders and lots of position/flanking/portal bombing. Mix in ownership of EMC and its typically quick to get back in it. Good stuff

Much much better than the T1 hit and roll zerg v zerg that happens once and then one of the teams runs to the nearest tower to hide/set up siege/go to an empty map and everyone moves to the next basically undefended objective.

Roaming is still pretty meh. Even if I get a 1v1 situation happening, someone always shows up (ally or enemy) and never ends without interference.

Artaz.3819:

Snowreap.5174:

the way the linking works, the ‘host’ world gets all of the rating benefit of a win (or the loss from a loss). linked worlds don’t get any rating change at all. EBay for example hasn’t moved from 1528.6343 since 4/22, even though they are ranked #12, because they’ve been a linked world since then.

so a 4-server linking might be very very powerful, but only the host world will see any kind of rating change as a result — the other 3 won’t see any rating improvement at all. and when the links change in the future, we might see CD fall again as rapidly as they’re now rising.

-ken

Ding ding.

You’ll notice that BG is not nearly as effective at beating the other T1 opponents — PPT wise — which also indicates that ET is much much larger and should be much much higher in glicko rating.

Actually, much of the past 2 months TC has not been pushing. TC is pulling crazy overtime this week, their groups seem to be running for 10+ hours at a time. At the same time BG is not showing up in force, probably sick of fighting TC and YB. I doubt anyone enjoys it much because TC get almost all their PPT during EU when BG is dead, and then BG gets almost all their PPT during SEA when TC is dead. Combined with YB actively avoiding fights and half their population wanting to throw so they can drop tiers, this week in t1 is kind of a case study in everything that is wrong in wvw.

It does make me wonder what ANet uses for their metrics to determine WvW population size, though. If it is average number of players on a map across a week, then TC is showing they are just as big as BG (their kills + deaths are higher, actually). If it is the number of player accounts that participate in a week, it seems to me BG is much higher (I see the same TC players / guilds in every day, where as plenty of BG players only seem to go in for their scheduled raids). Considering that BG is currently listed as full with TC very high, I would guess it is the second one.

How is the fantastic four supposed to move up when it will take the entire two months just to reach t3?

Since glicko doesn’t take into account the change in pairings, and doesn’t immediately take into account player transfers between servers, it’s not the optimal method of determining matchups, imho.

I’d rather see a ladder-type system, where the team with the highest score of each tier moves up to the next tier on the next weekly reset, and the team with the lowest score of each tier moves down a tier, though obviously the lowest scorer of T4 and the highest scorer of T1 would just stay put. At the start of each 2-month cycle, when the new pairings are determined, the ladder is randomly populated, so if BG ended up in T4, they’d have to work their way up. It would keep things from getting too stale too fast, as it could take 3-4 weeks for each tier to kind of settle, but after that, you’d still have the highest and lowest teams of each tier swapping out, so the matchups would never be the same in any two consecutive weeks.

Josh XT.6053:

The glicko wall is awful.

Honestly the pairing for that match up should have been like this:

CD + BP + ET
DH + EB + K
SF + FC + GOM

Really was no reason to put 4 servers together. Glicko needs reset, I’ve been saying it since the first server pairings.

Why does Glicko need to be reset when its plain that the actual issue in this regard is putting the 4 servers together? As I’ve been saying it calls into question anets whole linking system and how they decide the linkages. It appears to be an absolutely abysmal decision. How was it made?

Aitadis.8269:

A simple solution would be to reset glicko for a week, that way the servers that own will move up to T1, and the servers that struggle will go down to t3/t4. In a couple more weeks it will work itself out and put servers where they belong glicko wise. Please Anet.

I just think Glicko parameters needs to be modified so it adapts faster rather than slower.

I believe there is something called ‘RD’ in Glicko-1 that can be modified to do this and something called volatility in Glicko-2 that can be modified for the same purpose.

With these new linking systems the convergence needs to be faster.

dzeRnumbrd.6129:

Aitadis.8269:

A simple solution would be to reset glicko for a week, that way the servers that own will move up to T1, and the servers that struggle will go down to t3/t4. In a couple more weeks it will work itself out and put servers where they belong glicko wise. Please Anet.

I just think Glicko parameters needs to be modified so it adapts faster rather than slower.

I believe there is something called ‘RD’ in Glicko-1 that can be modified to do this and something called volatility in Glicko-2 that can be modified for the same purpose.

With these new linking systems the convergence needs to be faster.

In this initial period volatility has apparently been made more sensitive which is why we get match ups like the current T3 match up.

I thought they made glicko more volatile when they did the linking. We were seeing a lot more +/-60… Then they changed the scoring a bit and everything is around +/- 4…except cd.

morrolan.9608:

In this initial period volatility has apparently been made more sensitive which is why we get match ups like the current T3 match up.

I think the volatility parameter and “random” parameter are two different elements.

Glicko determine the rankings but the “random” component is then used after the Glicko is determined to add variety to the matchups.

Right now JQ’s Glicko is too stable (not adapting fast enough to such a big thrashing) but the “random” component is what pushed JQ into T3.

Correct me if I’m wrong anet devs?

dzeRnumbrd.6129:

Right now JQ’s Glicko is too stable (not adapting fast enough to such a big thrashing) but the “random” component is what pushed JQ into T3.

Correct me if I’m wrong anet devs?

Yes, JQ is on T3 because of the random element after the glicko is calculated (we belive it’s somenthing around +/- 100). Their small gain due the fact that they are facing servers with a much lower glicko than their, and glicko expects that they already perform far better than the other servers.

CD is gaining 100+ rating because their rating is much lower than the other two servers… However by the time CD get’s to, let’s say 1600, the other servers will have 1300-1200 which will make extremely hard to CD gain any rating.

BAITness.1083:

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day. I think it was two nights ago that was the least active I had seen them but they still had a blob that was map hopping. Otherwise, they have one massive blob in EBG usually ranging from 50 to 70, a blob in their BL between 30 and 50 and smaller groups usually 20 or less hopping maps. This is every. Single. Day. Since Friday reset and this is 24 hours a day.

To be honest, I’m not horribly upset because it’s been relatively fun. There’s non stop action and I stopped caring about PPT a long, long time ago. Still, their numbers are totally absurd whether I care about the score or not. Tonight when we were attempting to take SMC for like the fifth time, I had to stand back for a second to appreciate how many people they had in that room. Our own zerg had two tags and we must have been 40 – 50 tops but CD just swallowed our zerg like it was nothing.

I’m going to go on record and say that I’m not at all mad about this match up or Anet’s poor server pairing but I will say that it’s pretty kitten ridiculous. I believe what I had said in the server pairing thread was “Good LORD, four servers?” And I was right to be concerned about it. That many communities mixed in to one means that one servers down time will be another servers prime time. Effectively making their zergs prime time numbers all hours of the day and night.

SpellOfIniquity.1780:

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day.

In SEA, OCX and EU times? I find that hard to believe unless its holdovers from the T1 linking still on the respective servers, in which case I again state why weren’t those numbers taken into account in determining the linkage?

Just got out of playing a little after midnight PST (3am EST) and here’s a pic of the map…yes, they have a decent late night crew:

http://mos.millenium.org/na/matchups/

So FA, maguuma, JQ, SoS, and SBI could all potentially be either T2 or T3 at reset, but the abyss between SBI and HoD is pretty wide.

Anet’s own chart…

https://leaderboards.guildwars2.com/en/na/wvw

Doesn’t make any sense, since half those servers shouldn’t have their own score.

SpellOfIniquity.1780:

BAITness.1083:

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day. I think it was two nights ago that was the least active I had seen them but they still had a blob that was map hopping. Otherwise, they have one massive blob in EBG usually ranging from 50 to 70, a blob in their BL between 30 and 50 and smaller groups usually 20 or less hopping maps. This is every. Single. Day. Since Friday reset and this is 24 hours a day.

To be honest, I’m not horribly upset because it’s been relatively fun. There’s non stop action and I stopped caring about PPT a long, long time ago. Still, their numbers are totally absurd whether I care about the score or not. Tonight when we were attempting to take SMC for like the fifth time, I had to stand back for a second to appreciate how many people they had in that room. Our own zerg had two tags and we must have been 40 – 50 tops but CD just swallowed our zerg like it was nothing.

I’m going to go on record and say that I’m not at all mad about this match up or Anet’s poor server pairing but I will say that it’s pretty kitten ridiculous. I believe what I had said in the server pairing thread was “Good LORD, four servers?” And I was right to be concerned about it. That many communities mixed in to one means that one servers down time will be another servers prime time. Effectively making their zergs prime time numbers all hours of the day and night.

Hmm, maybe we’re looking at this the wrong way. Maybe this is what we should be hoping for. (See my bolding of your post). This is what WvW should be.

Maybe we should be hoping Anet links even more servers together. Perhaps all we need is 3 tiers in NA.

Johje Holan.4607:

SpellOfIniquity.1780:

BAITness.1083:

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day. I think it was two nights ago that was the least active I had seen them but they still had a blob that was map hopping. Otherwise, they have one massive blob in EBG usually ranging from 50 to 70, a blob in their BL between 30 and 50 and smaller groups usually 20 or less hopping maps. This is every. Single. Day. Since Friday reset and this is 24 hours a day.

To be honest, I’m not horribly upset because it’s been relatively fun. There’s non stop action and I stopped caring about PPT a long, long time ago. Still, their numbers are totally absurd whether I care about the score or not. Tonight when we were attempting to take SMC for like the fifth time, I had to stand back for a second to appreciate how many people they had in that room. Our own zerg had two tags and we must have been 40 – 50 tops but CD just swallowed our zerg like it was nothing.

I’m going to go on record and say that I’m not at all mad about this match up or Anet’s poor server pairing but I will say that it’s pretty kitten ridiculous. I believe what I had said in the server pairing thread was “Good LORD, four servers?” And I was right to be concerned about it. That many communities mixed in to one means that one servers down time will be another servers prime time. Effectively making their zergs prime time numbers all hours of the day and night.

Hmm, maybe we’re looking at this the wrong way. Maybe this is what we should be hoping for. (See my bolding of your post). This is what WvW should be.

Maybe we should be hoping Anet links even more servers together. Perhaps all we need is 3 tiers in NA.

Just to note that for many, this is already feeling like EotM.

joneirikb.7506:

Johje Holan.4607:

SpellOfIniquity.1780:

BAITness.1083:

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day. I think it was two nights ago that was the least active I had seen them but they still had a blob that was map hopping. Otherwise, they have one massive blob in EBG usually ranging from 50 to 70, a blob in their BL between 30 and 50 and smaller groups usually 20 or less hopping maps. This is every. Single. Day. Since Friday reset and this is 24 hours a day.

To be honest, I’m not horribly upset because it’s been relatively fun. There’s non stop action and I stopped caring about PPT a long, long time ago. Still, their numbers are totally absurd whether I care about the score or not. Tonight when we were attempting to take SMC for like the fifth time, I had to stand back for a second to appreciate how many people they had in that room. Our own zerg had two tags and we must have been 40 – 50 tops but CD just swallowed our zerg like it was nothing.

I’m going to go on record and say that I’m not at all mad about this match up or Anet’s poor server pairing but I will say that it’s pretty kitten ridiculous. I believe what I had said in the server pairing thread was “Good LORD, four servers?” And I was right to be concerned about it. That many communities mixed in to one means that one servers down time will be another servers prime time. Effectively making their zergs prime time numbers all hours of the day and night.

Hmm, maybe we’re looking at this the wrong way. Maybe this is what we should be hoping for. (See my bolding of your post). This is what WvW should be.

Maybe we should be hoping Anet links even more servers together. Perhaps all we need is 3 tiers in NA.

Just to note that for many, this is already feeling like EotM.

Isn’t that because they’re too dominant and have no one to fight? If every server had 40+ people on 24/7 then there would be someone to fight.

Yep, I think anet should (maybe) merge more servers. 3 Tiers in EU and NA should do it, maybe increase the amount of maps instead.

I think they are looking too much at server active players anyways, if germans have lot of players in their server, they shouldnt be punished for that by getting unmerged, what is I heard happening. At least 2 tiers should be removed in EU and 1 in NA.

Anyways NA is in bad spot atm, the glicko difference in EU tiers is less than 100, in NA its up to 300, its just stupid and NA glicko should be reset or at least reset to something like
t1 1900-1950
t2 1800-1850
t3 t700-1750
t4 1600-1650

Do not focus on tiers anymore. Focus instead on fun fights and such-like. All these linkages mean that tiers are totally meaningless. Just when you start to move tiers they are going to change linkages anyway as populations move around.

Kaiser.9873:

Do not focus on tiers anymore. Focus instead on fun fights and such-like. All these linkages mean that tiers are totally meaningless. Just when you start to move tiers they are going to change linkages anyway as populations move around.

What some of us are saying though is that the current T4 matchup didn’t take into account populations moving even since the last linkage. ET, BP and Kain got stacked more in the last linkage so some could still play in T1/T2, and yet Anet linked those 3 with CD? Not only are tiers meaningless, so is population balancing. And yes, it was totally fun having CD’s full zerg take our T3 keep in EBG as if it was a paper tower…that fight lasted all of one quick death and not enough time to rush back to do anything else except to get pushed all the way back to spawn by the zerg.

I agree each linking should reset the base scores allowing servers to move up or down easily. Nothing wrong with linking so many servers but locking a more heavily populated server in a tier with substantially less populated servers simply isn’t fair.

Sylvyn.4750:

Kaiser.9873:

Do not focus on tiers anymore. Focus instead on fun fights and such-like. All these linkages mean that tiers are totally meaningless. Just when you start to move tiers they are going to change linkages anyway as populations move around.

What some of us are saying though is that the current T4 matchup didn’t take into account populations moving even since the last linkage. ET, BP and Kain got stacked more in the last linkage so some could still play in T1/T2, and yet Anet linked those 3 with CD? Not only are tiers meaningless, so is population balancing. And yes, it was totally fun having CD’s full zerg take our T3 keep in EBG as if it was a paper tower…that fight lasted all of one quick death and not enough time to rush back to do anything else except to get pushed all the way back to spawn by the zerg.

Anet stated that they do take into account populations though. This leads me to wonder if this is due to the power of the fairweather effect. Is the fairweather effect so powerful that no matter what Anet does we are doomed to have population imbalance?

And if that is the case then Glicko needs to be dealt with so that it changes rapidly. Every single person who looks at the NA T4 match can see that the 4-server pairing needs to move up. If the Glicko algorithm can’t see that, then it needs to change.

• [quote=6218545;McKenna Berdrow.2759:]
• Q: What factors go into determining world links?
  A: We take into consideration a lot of factors when making the links, but the primary factor we look at is the WvW activity level for each world, during peak and off-peak hours. The goal being to normalize world populations among unlinked worlds and groups of linked worlds. For EU we also filter potential links by language.
  [/quote]

Johje Holan.4607:

joneirikb.7506:

Johje Holan.4607:

SpellOfIniquity.1780:

BAITness.1083:

joneirikb.7506:

The linking system sure doesn’t seem to work very well with Glicko. Uncertain just how fast glicko reacts to changes these days, but it might be unavoidable to have Glicko reset, or at least a partial naturalizing of the glicko scores on new links ?

After all, the Glicko score represents the status the server used to have before linking or the previous link. Each link basically creates an entirely new situation.

I don’t know the number-crunching of Glicko, so don’t know how it would work with the current “faster change in glicko”, it might just end up making the winnder of T4 shoot into t1, get totally smashed, and get sent down to bottom of t4 again etc. So be careful what you wish for. :p

This is very true, and a real concern, but this linking thing is showing us some very surprising numbers. Right now the four-server-alliance is the most active WvW participant – by far. Comparing to BG, which is widely considered the most active server, the Quad server is about 40% more active (using kills + deaths as measure of activity). Part of this may be BG having an off-week and YB being absolute pros at avoiding fights at all costs, but the quad server is keeping a ridiculously high level of activity. It is rolling at a monstrous 1.28 kdr. The only server with a better KDR right now is JQ+AR, and they are not nearly as active.

What I am trying to say is, from the limited info we have, it looks like the quad server could be a real contender in any tier, and maybe even be the best server for the duration of the pairing. It depends 100% on whether or not their activity level stays this high.

But unless we get glicko resets, we may never find out.

Since Friday reset of last week, CD has had at least one zerg of 40+ all 24 hours of every day. I think it was two nights ago that was the least active I had seen them but they still had a blob that was map hopping. Otherwise, they have one massive blob in EBG usually ranging from 50 to 70, a blob in their BL between 30 and 50 and smaller groups usually 20 or less hopping maps. This is every. Single. Day. Since Friday reset and this is 24 hours a day.

To be honest, I’m not horribly upset because it’s been relatively fun. There’s non stop action and I stopped caring about PPT a long, long time ago. Still, their numbers are totally absurd whether I care about the score or not. Tonight when we were attempting to take SMC for like the fifth time, I had to stand back for a second to appreciate how many people they had in that room. Our own zerg had two tags and we must have been 40 – 50 tops but CD just swallowed our zerg like it was nothing.

I’m going to go on record and say that I’m not at all mad about this match up or Anet’s poor server pairing but I will say that it’s pretty kitten ridiculous. I believe what I had said in the server pairing thread was “Good LORD, four servers?” And I was right to be concerned about it. That many communities mixed in to one means that one servers down time will be another servers prime time. Effectively making their zergs prime time numbers all hours of the day and night.

Hmm, maybe we’re looking at this the wrong way. Maybe this is what we should be hoping for. (See my bolding of your post). This is what WvW should be.

Maybe we should be hoping Anet links even more servers together. Perhaps all we need is 3 tiers in NA.

Just to note that for many, this is already feeling like EotM.

Isn’t that because they’re too dominant and have no one to fight? If every server had 40+ people on 24/7 then there would be someone to fight.

Personally noticed it with the TC+Kain link, where I entered a WvW map and found lots of people and guilds I had no idea who where, what where doing, who was running private or open, and a bunch of other things. It felt just like going into EotM, and just watching a swarm of people you don’t know.

And then knowing that in 2-3 months, when you finally got to know some of them, things would change again, and you would meet new people.

Same thing is exemplified quite well in the T4 match-up, lots of servers mashed together, not only don’t I know most of the people, I don’t even know what server they are from, etc. 4 Servers trying to organize a team-speak and failing, and some wonder if it is really worth the trouble to later sort out all the accounts in 2 months again.

Simply put, the feeling of “community” is washed away, and you stick to your guilds instead. At least that is how I’ve reacted to this. This is exactly how I also feel when entering EotM.

I’m not against Linking, I think it has many good qualities and uses. But the more I experience it, I wonder why they don’t just mash us together into an EotM system and call it done (I know some servers are large enough that they don’t need to be linked, and thus won’t feel the same etc). Guess I’m just saying that the difference between being linked 2-3-4 servers and EotM is very small, you saying: “Perhaps this is how WvW should be?” and I think: “Like EotM?”.

Just food for thought, trying to highlight the perception of linking vs EotM, and how small the difference really is.

PS: Almost every time I’ve been in EotM I’ve been red, and gotten run over :p so nope!

Looks like things have changed and now both JQ and DB are losing glicko for there wins this week. Hope SF and DH packed the lube, because it’s gonna be a long 7 more weeks of this matchup if anet doesn’t intervene.

No intervention, we’re heving this again.

/cheers

We can already have a glimpse of the result of this matchup now.

Before someone comes with the argument “it still saturday”, I want to let you know that glicko works with the % difference between the score and not the points itself. 40k/20k/20k is the same as 500k/250k/250k for glicko.

Right now, the quadserver is winning by double over the other two servers, which is far above the %s of the last match that ended at 467k/329k/270k, and is only gaining 55 rating points, oposed to the 110 gained last week.

Let’s say that the quadserver maintain this for the whole week and get the 50 rating now. They will now be 200 above the two competitors for the next match oposed to the 100 rating difference of the current match. Glicko will expect that the Quadserver wins even harder to get any significant ammount of rating (let’s say 150% above 2nd place), and it will expect to have the quadserver to keep performing even harder as the matchups go on since their rating will be far above the other two contenders. They will be squeezing water out of a rock at some point.

There was a similar issue 3 years ago when Sorrow’s Furnace was stuck in T8 and couldn’t move out simply because Eredon Terrace and Fergusson’s Crossing rating were far lower than theirs, at to some point that despite SF winning the matchup by almost 100% over the other two server they would still lose rating. Arena Net manually gave 100 rating to both ET and FC which allowed SF to finally gain enough rating to move out of the tier and make their way up, even beating T7 servers. We had a T6 server stuck in T8 for several months just because glicko wouldn’t adjust a way for them move up.

Imho, if Arena Net don’t want to manually move the quadserver, they should at least do the same they did back in that time to allow CD gain enough rating to get out of T4.

While the ideal would actually just reset everyone’s rating and let the servers adapt by themselves every time the pairings change, feeding artificial rating will help to reduce the huge rating walls that formed between the tiers, which would give us at least different matchups if we get extremely loopsided ones like the one we’re having right now.

While it would be nice for the quadserver to move up, Anet left the other two server pairings with so little population that they have no chance against any team that came down from T3 either. I guess the best we can hope for is a slightly better matchup than the roflstomp we have now.

Aitadis.8269:

Currently CD is getting 110 glicko for this weeks win but that’s mostly because DH has died. Next week CD will not be getting anywhere near 100 glicko from the other 2 servers. Because of the volatility changes, they are losing the majority of there glicko this week and will only lose 10-20 each next week making it impossible at the current rate for CD to move up to T3. We are currently 300+ glicko behind SBI/HoD, so even if we continue getting 50 glicko a week from these 2 servers which is unlikely, it will probably be around 40 tops, it will still take 6+ weeks for us to get in range of T3 glicko and that is broken. To compare it to T3, JQ is beating there opponents quite convincingly PPT wise, similar to CD but are only getting 4 whole glicko for it. And Dragonbrand who is winning there Matchup in T2 by a decent amount is LOSING glicko. CD will soon be the same in a couple weeks, barely getting any glicko if this doesn’t change.

A simple solution would be to reset glicko for a week, that way the servers that own will move up to T1, and the servers that struggle will go down to t3/t4. In a couple more weeks it will work itself out and put servers where they belong glicko wise. Please Anet.

I agree, I wish you guys could pop up to T3 and replace any T1 or T2 we get dropped on us but then what would drop to T4 for the 3rd spot ? General problem in the current link is that T3 seems to have only 2 servrs (links) SOS + NSP and SBI + HOD.

There is also the large novelty factor about facing the “4 headed hydra” that would probably draw out everyone and their mother to come play against you just out of sheer curiosity, so there is no telling if T3 would wake up and you guys would be faced with a very unbalanced situation again. But overall I think it’d be fun for a week or so.

It seems that no one at Anet even understands there own glicko system, maybe thats why it hasn’t been reset. It’s ridiculous that they can’t see the glaring problem with setting a 4server merge with only 1300 glicko to start and 400 glicko behind t3. A high school dropout could of seen this coming, were currently sitting at +34 glicko for the week, my predition was 40 tops so I was right there, next week we will get an even smaller amount of glicko making it impossible to move to t3, effectively killing 5 servers because they won’t wanna get blobbed down for 6 more weeks. Or maybe this was Anets plan all along, make money from people xfering OUT of t4, thats why it has more servers merged than any other tier, I certainly wouldn’t put it past them.

Oh boy I wonder who CD will get this week >.>

Aitadis.8269:

Oh boy I wonder who CD will get this week >.>

Same as before, not even colors changed.

Same Bat time, same Bat channel.

same stale matches as last week. thx anet

i’ve addressed this in another thread and many others before me. that the glicko has been stuffed for a while. with the double reset in one week to other varing factors. the glicko doesn’t show accurate information. and if it did, the glicko doesn’t know what to do with it