My server is empty - A solution to balance, and empty teams

[Spadassin.4076]

2 minutes ago, Chaba.5410 said:

It sounded like you were expecting this 2nd WR team formation to deliver on some supposed promise that was made.  I don't recall any such promise being made in that regard, only a promise that the algorithm is going to be "iterated" on over time.  A schedule was never promised upon.  Maybe you can point to where this promise was made, which was my original question?

I believe they have mentioned somewhere that they'll keep improving the algorithm. I think we're saying the same thing.

[Bloodstealer.5978]

12 hours ago, Cyninja.2954 said:

That's not what happened. Not even in the slightest.

Servers where deleted overall. There are no servers any more. There are shards, which are recreated every 4 weeks currently.

As far as Palawadan, this is the activity for that shard atm:

21.144

kills, 17.805

deaths, 1.19

K/D,  17.27 % PPK,

177 VP

https://gw2mists.com/teams/Palawadan

That is NOT a dead shard. We have sufficient access to data to disprove such claims.

As far as tiers, Eu is BACK to 5 tiers, after having been increased to 6 last restructure. That's adjusting to incorrectly increasing tiers from 5 to 6 at the beginning of the event.

There was a warning. There have been months and years of betas and information about the World Restructuring. If you missed all that, that is on you.

Total bullcrud.. there was no warning they were deleting 3 of the matchups.. even the wiki still states 18 teams .. they downed WvW and when it was put back up they had pushed pops from those empty teams into other empty teams in the hope it would create balance. Add to that they moved players off other matchup onto this low pop rubbish instead.

As for your bs data.. try playing on the matchup its dead 90% of the time for those on Palawadan and when there is activity for the odd hour we have barely enough players to cover 1 map and no chance on any of the others.

Be a fanboi all you want of this debacle but there are many of us that have had WvW as a gamemode ruined.

And yeah its just as bad for those that are on stacked teams when the reset has worn off and they have nothing else to do when the matchups dwindle away until that small window of activity happens again

 

EDIT - ooohh yeah and lol, ANET have today updated the wiki . only 5 days late to let us know what they were doing.. classy!

Edited July 30 by Bloodstealer.5978

[Cyninja.2954]

12 hours ago, Bloodstealer.5978 said:

Total bullcrud.. there was no warning they were deleting 3 of the matchups.. even the wiki still states 18 teams .. they downed WvW and when it was put back up they had pushed pops from those empty teams into other empty teams in the hope it would create balance. Add to that they moved players off other matchup onto this low pop rubbish instead.

I am unsure you know what your are talking about. Let's clear up some terminology while at it too.

1. No, 3 match-ups were NOT removed. 3 shards were removed, which means the loss of Tier 6 on EU. We had 5 tier before the WR event 6 weeks ago.

The wiki is player maintained and the latest restructure was last Friday evening EU time.

2. The removal of 3 shards was indeed done to spread players across less shards. This is literally a reversion of the increase in shards and tiers from before the WR event, which was likely done assuming , more activity would be happening (and thus more queues). We are now back to the same amount of tiers as during the server linking system.

Quote

As for your bs data.. try playing on the matchup its dead 90% of the time for those on Palawadan and when there is activity for the odd hour we have barely enough players to cover 1 map and no chance on any of the others.

We have the activity data literally on a per hour basis. That's data from the games api, taken from activity in game, fed to the website.

Quote

 

Be a fanboi all you want of this debacle but there are many of us that have had WvW as a gamemode ruined.

And yeah its just as bad for those that are on stacked teams when the reset has worn off and they have nothing else to do when the matchups dwindle away until that small window of activity happens again

 

I'm not being a fan boy. I am pointing out inconsistencies and simple lack of understanding on your part. Which in turn does not fill me with confidence in your ability to make correct assumptions on the activity of the shard you are on. 

Quote

EDIT - ooohh yeah and lol, ANET have today updated the wiki . only 5 days late to let us know what they were doing.. classy!

Arenanet does not update the wiki. Players do (though developers can be individuals that also update the wiki). The wiki is NOT part of the studios official responsibility.

EDIT:

and as an update, within about 24 hours, Palawadan went from the stats yesterday at 10:29 (21.144 kills, 17.805 deaths, 1.19 K/D,  17.27 % PPK, 177 VP) to today at 11:05:: 25.233 kills, 21.019 deaths, 1.2 K/D, 16.1 % PPK, 229 VP and leading the bracket. That's a net gain of 4.089 kills and 3.214 deaths over approximately a 24h period.

Looking once again on the activity chart: https://gw2mists.com/teams/Palawadan

The shards is mostly active during: 06:00 - 11:00 )[probably a morning commander), 18:00 - 23:59 (evening primer time).

The most inactivity is during: 11:00 - 17:00 (mid day) and 0:00 - 6:00 (night time)

This activity is pretty consistent for the last few days, with higher activity on the weekend.

Edited July 31 by Cyninja.2954

[One more for the road.8950]

57 minutes ago, Cyninja.2954 said:

Which in turn does not fill me with confidence in your ability to make correct assumptions on the activity of the shard you are on.

Already admitted to not really playing, just spewing.

On 7/29/2024 at 7:23 PM, Spadassin.4076 said:

Perhaps I was on another team. Honestly I have barely played WvW since alliances came up

 

[SweetPotato.7456]

On 7/29/2024 at 7:27 AM, Spadassin.4076 said:

Looks like you've been drinking too much kool-aid. Can you all stop changing topic though? Why are so solo players, and small roaming guilds forced to join large alliances to not end up in a vacuum?

If you want to be solo, what's wrong with ending up in a vacuum, solo is being in a vacuum. you want people around but you don't want to join an alliance, complain it is vacuum yet wanted to be solo.... 

[fantom.8427]

6 hours ago, Spadassin.4076 said:

I believe they have mentioned somewhere that they'll keep improving the algorithm. I think we're saying the same thing.

I hope it gets better.  Right now 3 shards on NA have k/d under 0.7.  It is not a lot of fun being on the receiving end of 15v40 for hours on end.  Right now in T4 the green team is just bouncing back and forth between red side and blue side of EBG with a 50 man zerg fighting 10 to 15 over-matched defenders.  They are also taking over blue and red maps at the same time.  As of this post, blue has held their ebg keep for 5 minutes, red for 1 hour, and green for 94.2 hours, which if my math is correct means that they have held it since reset. 

[Spadassin.4076]

9 hours ago, One more for the road.8950 said:

Already admitted to not really playing, just spewing.

 

Why do you quote me as an answer to two other people's discussion? Stop spewing nonsense.

[Spadassin.4076]

8 hours ago, fantom.8427 said:

I hope it gets better.  Right now 3 shards on NA have k/d under 0.7.  It is not a lot of fun being on the receiving end of 15v40 for hours on end.  Right now in T4 the green team is just bouncing back and forth between red side and blue side of EBG with a 50 man zerg fighting 10 to 15 over-matched defenders.  They are also taking over blue and red maps at the same time.  As of this post, blue has held their ebg keep for 5 minutes, red for 1 hour, and green for 94.2 hours, which if my math is correct means that they have held it since reset. 

I don't believe it can get much better as it is. The only viable alternative is PvE like megaservers. Even T1 is very boring during off-hours.

[latlat.4516]

On 7/28/2024 at 11:47 PM, Spadassin.4076 said:

There are probably less than 500 players both capable and active in the EU. The size of an alliance is currently 500.

2500 players made at least 25 kills per day this week only on EU, and I'm basing this on mists stats which are incomplete since you need to be registered. 

 

Just throwing the info here, do whatever you want with it. 

[Reztek.7805]

11 hours ago, Bloodstealer.5978 said:

Total bullcrud.. there was no warning they were deleting 3 of the matchups.. even the wiki still states 18 teams .. they downed WvW and when it was put back up they had pushed pops from those empty teams into other empty teams in the hope it would create balance. Add to that they moved players off other matchup onto this low pop rubbish instead.

As for your bs data.. try playing on the matchup its dead 90% of the time for those on Palawadan and when there is activity for the odd hour we have barely enough players to cover 1 map and no chance on any of the others.

Be a fanboi all you want of this debacle but there are many of us that have had WvW as a gamemode ruined.

And yeah its just as bad for those that are on stacked teams when the reset has worn off and they have nothing else to do when the matchups dwindle away until that small window of activity happens again

 

EDIT - ooohh yeah and lol, ANET have today updated the wiki . only 5 days late to let us know what they were doing.. classy!

While the 2nd "reshuffle" is much worse than Anets 1st attempt, Palawadan is far from dead. (EU T5, green)

https://www.gw2matchup.com/home?region=eu

https://www.gw2matchup.com/matchup?id=2-5

The team that could complain is blue, Domain of Anguish.

[Spadassin.4076]

1 hour ago, latlat.4516 said:

2500 players made at least 25 kills per day this week only on EU, and I'm basing this on mists stats which are incomplete since you need to be registered. 

 

Just throwing the info here, do whatever you want with it. 

What's 25 kills? Like one blob fight?

How many of these do you believe are actually good at the game and not just tagging bags?

500 sounds about right - and there is a power law within these people as well. I have been in enough squads to know that it's always the same people that are top DPS, cleanse, strip.

[latlat.4516]

20 hours ago, Spadassin.4076 said:

What's 25 kills? Like one blob fight?

Mists don't really show the intensity of kills, but if you don't play on zergs 25/day it's an honest number for anyone that would play the game 3-6hours a week, and it's the majority of the player base.

 

20 hours ago, Spadassin.4076 said:

How many of these do you believe are actually good at the game and not just tagging bags?

My opinion on the matter is that you are conflating the difference between "good at the game", "active" and "being the top", and overweighting the importance "being good" should have within the matchmaking algorithm. Trying to express "being good at the game" is not easy as any single or compound of metrics you would use can be read in different ways, and depend on environmental settings. 

The problem with your statistics is that it sounds "about right' for an arbitrary limit you set yourself. Obviously, if you compare [BiG] players against [here] players you won't get the same level of investment in the game and the result will show. But again, if you put [here] against [PYRE] for example, you get another result. And that's leaving KT alliances out of the picture. The question is at what point does this specific difference start to count, and what point does it not ? 

As a side note, the metrics I used to count active players is also arbitrary (25 kill/day does not mean anything by itself). Yet it has the added benefit of using a method (I went to mists and looked at all players that had >100 kills on tuesday), which your "500 good players max" does not have except "it sounds right", and I have difficulty buying it. 

 

How did you come up with your "500" exactly ? Wet finger ? Common sense ? Feelings off your experience during the last month ? You made an add-hoc analysis based off the fact that only 3 worlds seemed to be active and competitive during the June/July link and used an hypothesis that only about 10% of the population of these worlds is playing well ? You asked opinions from people in your guild ? Surely you didn't just made up this number because you are pissed off at the fact that people can group up to 500 players together and that it allows very active guilds to hold a firm grip on their matches, right ? 

And on another hand, even if it is true. In the future day when gw2 will be in gw1 mode, will you ask for guilds to be 10 people max because only 50 people are good at the game ? Is it even a relevant metric to evaluate the degree of allowed bandwagoning ? 

[KrHome.1920]

Basing the algorythm exclusive on K/D would probably be better than anything ANet has done up till now.

In my matchups the score correlates pretty well with K/D. This week:

• Blue: K/D: 19.000 / 26.000 => 45.000 in total // Score: 252
• Red: K/D: 29.000 / 23.000 => 52.000 in total // Score: 282
• Green: K/D: 32.000 / 23.000 = 55.000 in total // Score: 282

Red currently being equal to green is just a result of nightcapping. Green will inevitably win this matchup (they are leading the current skirrmish as well).

[Arya Whitefire.8423]

2 hours ago, latlat.4516 said:

And on another hand, even if it is true. In the future day when gw2 will be in gw1 mode, will you ask for guilds to be 10 people max because only 50 people are good at the game ? Is it even a relevant metric to evaluate the degree of allowed bandwagoning ? 

I think 10 player guild caps is a great idea.  Make it so!  (assuming you actually want some sort of competitive game mode, I'm not sure I even want anything beyond a sandbox at this point)

[Dawdler.8521]

10 minutes ago, Arya Whitefire.8423 said:

I think 10 player guild caps is a great idea.  Make it so!  (assuming you actually want some sort of competitive game mode, I'm not sure I even want anything beyond a sandbox at this point)

With a 1 player guild cap no one could argue it anymore.

[One more for the road.8950]

1 hour ago, Dawdler.8521 said:

With a 1 player guild cap no one could argue it anymore.

Can you imagine all that bank space for 6 one man guilds? Freeeeeeeeeeeeeedom

[Spadassin.4076]

13 hours ago, latlat.4516 said:

Mists don't really show the intensity of kills, but if you don't play on zergs 25/day it's an honest number for anyone that would play the game 3-6hours a week, and it's the majority of the player base.

 

My opinion on the matter is that you are conflating the difference between "good at the game", "active" and "being the top", and overweighting the importance "being good" should have within the matchmaking algorithm. Trying to express "being good at the game" is not easy as any single or compound of metrics you would use can be read in different ways, and depend on environmental settings. 

The problem with your statistics is that it sounds "about right' for an arbitrary limit you set yourself. Obviously, if you compare [BiG] players against [here] players you won't get the same level of investment in the game and the result will show. But again, if you put [here] against [PYRE] for example, you get another result. And that's leaving KT alliances out of the picture. The question is at what point does this specific difference start to count, and what point does it not ? 

As a side note, the metrics I used to count active players is also arbitrary (25 kill/day does not mean anything by itself). Yet it has the added benefit of using a method (I went to mists and looked at all players that had >100 kills on tuesday), which your "500 good players max" does not have except "it sounds right", and I have difficulty buying it. 

 

How did you come up with your "500" exactly ? Wet finger ? Common sense ? Feelings off your experience during the last month ? You made an add-hoc analysis based off the fact that only 3 worlds seemed to be active and competitive during the June/July link and used an hypothesis that only about 10% of the population of these worlds is playing well ? You asked opinions from people in your guild ? Surely you didn't just made up this number because you are pissed off at the fact that people can group up to 500 players together and that it allows very active guilds to hold a firm grip on their matches, right ? 

And on another hand, even if it is true. In the future day when gw2 will be in gw1 mode, will you ask for guilds to be 10 people max because only 50 people are good at the game ? Is it even a relevant metric to evaluate the degree of allowed bandwagoning ? 

The real number is closer to 500 than let's say 2000. How do I know this? Experience...

I play WvW/GvG since launch, and I have seen it all. The top servers are carried by a small group of 100ish people. I have seen the same names for years. I have followed the "bandwagon" - started in jade sea, then went to vabbi, ros, wsr, ranik.

However I don't think you have read my post as I am NOT advocating for reducing guild size. As you pointed out it's a losing battle.

The servers I mentioned before have been top 1 without contest for years even when WvW had more players. That's why it's both a number and quality problem. Anet doesn't know how to measure quality - you can spam Necro marks from the backline and get a high kdr, as well as high activity but still be utterly useless (hello Mighty). 

What I am advocating for is PvE like megaservers in WvW. I haven't seen an argument against it yet in the entire thread.

PS: make 5v5 roaming great again.

Edited August 1 by Spadassin.4076

[One more for the road.8950]

3 hours ago, Spadassin.4076 said:

started in jade sea, then went to vabbi, ros, wsr, ranik.

The servers I mentioned before have been top 1 without contest for years even when WvW had more players.

What do you mean, they have been completely dead at times.

[KrHome.1920]

7 hours ago, Spadassin.4076 said:

The servers I mentioned before have been top 1 without contest for years even when WvW had more players. That's why it's both a number and quality problem. Anet doesn't know how to measure quality

What outcome do you expect, when the game allows population shifts every few weeks? In a working matchups system good players move to the top and stay there till the end of time to face other good players. You can not balance server matchups when you allow players to transfer servers at the same time so that every player can look for easy wins. That's trival, but ANet did it nevertheless. The alliance system is technically the same just under a different name.

[Dawdler.8521]

2 hours ago, KrHome.1920 said:

In a working matchups system good players move to the top and stay there till the end of time to face other good players.

Doesn’t most matchup systems have seasons? Like take the NFL. Does the first and second ranked teams from the previous seasons only play against each other and no one else?

[Dayra.7405]

Now it's friday, last weekend the opener complained about the imbalance of EU-T1. That's really funny it's turned out from scoring and won/lost matches that EU-T1 is currently the most balanced match in whole EU. https://www.gw2matchup.com/home?region=eu

T1-Leader has 322 points in all other EU matches the leader has more than that (325-379). T1-last has most points of all last in EU 292 vs 252-290.

Generally I would rate EU quite balanced this week, except T4.

Looks like try cry of imbalance wasn't justified by results. Not only draws are balanced 🙂 

Edited August 2 by Dayra.7405

[KrHome.1920]

2 hours ago, Dawdler.8521 said:

Doesn’t most matchup systems have seasons? Like take the NFL. Does the first and second ranked teams from the previous seasons only play against each other and no one else?

Seasons in GW2 looked like this (reminder: I answered to a post that talked about the past):

Server relink => bandwagoning => 8 (later 4) weeks of imbalance => server relink => bandwagoning => ...

I am not into NFL. But I guess it is not allowed to bandwagon to the dominating team midseason, but if it is, then I think such things like queues so every bandwagoner has a chance to join each match do not exist in NFL. 🤪

Alliances:

If the alliance algorithm can identify specific player skill level and matchup accordingly, then all god, but that's not what I am seeing currently in my matchups. We have still dominating sides vs. frustrated sides.

[Dawdler.8521]

13 minutes ago, KrHome.1920 said:

 

Alliances:

If the alliance algorithm can identify specific player skill level and matchup accordingly, then all god, but that's not what I am seeing currently in my matchups. We have still dominating sides vs. frustrated sides.

Well yeah but won’t they move to the top then?

[Dayra.7405]

22 minutes ago, Dawdler.8521 said:

Well yeah but won’t they move to the top then?

Well if they got placed in T5 they not even reach T1 within the 4 weeks linking period

[Spadassin.4076]

2 hours ago, Dayra.7405 said:

Now it's friday, last weekend the opener complained about the imbalance of EU-T1. That's really funny it's turned out from scoring and won/lost matches that EU-T1 is currently the most balanced match in whole EU. https://www.gw2matchup.com/home?region=eu

T1-Leader has 322 points in all other EU matches the leader has more than that (325-379). T1-last has most points of all last in EU 292 vs 252-290.

Generally I would rate EU quite balanced this week, except T4.

Looks like try cry of imbalance wasn't justified by results. Not only draws are balanced 🙂 

The kdr from your own link says otherwise. Points are a poor measurement because no one cares about it, and it doesn't show active population. I wouldn't call it balanced when one team has half the kills of the two others - almost all matchups are like that.

Even if you put the low population teams against each other's (which Anet always fails at because they don't know how to measure player quality), you might call the matchup balanced but it will still be boring - megaservers solve that, prove me wrong.

Edited August 2 by Spadassin.4076