Guild Wars 2

Come on, Kaineng! Let’s keep at ’em!

Please close free server transfers, this is a farce.

Can we get a sticky that explains these statistics and how they’re calculated? About the only thing I understand about these posts is the relative placement of the servers.

It’s utterly meaningless while free transfers are open anyway.

Big guilds are swapping Worlds based on the most recent ranking position, so whoever was at position 1 yesterday is irrelevant. This morning the Worlds all have totally different people on them, so the balance of power is changed and any automatic ranking system is completely broken.

Chamone.6890:

so whoever was at position 1 yesterday is irrelevant.

i dont know why but when i saw this line, i felt a sinking feeling.

It would be nice to know how these numbers were achieved and what they mean. Besides raw score, I could only guess what the others might be.

I’m guessing Deviation is how much has changed since the last match-up … maybe? And Volatility is how drastic the change was? The Volatility numbers seem very close to each other and I’m guessing it’s some kind of percentage value. Would abnormal shifts of population amongst servers (eg. “Bandwagonneers”) affect the volatility of certain servers?

They had a post earlier explaining it. I’m too lazy to search for it now though :P

I don’t know why they still post it anyway. There is a player driven reporting system to give up to date numbers. http://mos.millenium.org/matchups

That’s how large guilds know where to move in advance of the Friday re-match.

Arborstone use “flyhack”, this is pathetic….

Slowly turning into “Number of Oceanic Players in Descending Order”

How is volatility and deviation determined?

BakaOniiChan.9564:

How is volatility and deviation determined?

If you are able to endure math you can find all details in this document:

http://www.glicko.net/glicko/glicko2.pdf

It’s not particularly easy to explain Glicko2 without doing the math, but Mark Glickman’s gave some pretty catchy outline in his own words:

• Ratings Deviation: “The Glicko system [computes] not only a rating, which can be thought of as a best guess of one’s playing strength, but also a ratings deviation (RD) […] which measures the uncertainty in a rating (high RD’s correspond to unreliable ratings).”

• Volatility: “The volatility measure indicates the degree of expected fluctuation in a player’s rating. The volatility measure is high when a player has erratic performances (e.g., when the player has had exceptionally strong results after a period of stability), and the volatility measure is low when the player performs at a consistent level.”

~MRA

Mike, no delta chart this week?
Our guild found them really helpful.

Such as this one a few weeks back:
https://dviw3bl0enbyw.cloudfront.net/uploads/forum_attachment/file/10728/Na_Ratings_9-29_to_10-5.JPG

Let’s go SoR!

Go! Go! Go! Go!

Chamone.6890:

It’s utterly meaningless… whoever was at position 1 yesterday is irrelevant. This morning the Worlds all have totally different people on them, so the balance of power is changed and any automatic ranking system is completely broken.

TL;DR: the record does not support your claim.

I haven’t been studying the results, but I believe that Jade Quarry has been dominating from the start, and also that both Stormbluff Ise and Isle of Janthir have been top rated for quite a bit.

And haven’t everybody’s favorite underdogs (viz. Kaineng, Devona’s Rest, and Ferguson’s Crossing) have been consistently duking it out at the bottom?

My server, Sanctum of Rall, got thrashed some weeks, eeked out a narrow win a couple of weeks ago, and then crushed our opponents last week. This week we are in a tough fight with Sea of Sorrows; Crystal Desert is effective at the unit level, but apparently haven’t yet had the numbers to take a lead.

Don’t get me wrong — just on principle, I think that there should be a week or so delay before transfers from another server can play WvWvW. But I don’t think the rankings have been unduly influenced by server transfers.

Bogartan

Can we get a sticky that explains these statistics and how they’re calculated? About the only thing I understand about these posts is the relative placement of the servers.

Not stickied anymore but here was explained how they use the Glicko2 rating system.
https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/World-Rankings-as-of-9-28-New-Format/first
Had to look it up from my own posts because I remembered I posted a reply there So I agree that a sticky would be great.

Please DO NOT close free transfers. I have no interest paying to change servers when the one im on inevetibly dies because no one else wants to pay to move around and people get bored. Simply reduce it to once a week or two weeks.

CLOSE TRANSFERS! Smaller servers will rebuild.

Transfer to top 6/12 will cost you money, the rest is free. This way if big guilds are tired of queue, they can move to the lower tier to bring it up. With this more server will be able to give a fight, instead of just getting stomped whenever they move up a tier.

They should also make it so you can’t transferr to a server your currently facing in the system. That’s what is killing things is people going to whoever has zerging for faster xp.

Keep free server transfers!!

Just add in a debuff on transfer: you cant WvW until after that weeks WvW reset. (I’d prefer a 30 day block, but folks say thats a bit harsh) That will ensure there is no instant steamroll.

Dred

holska.4127:

So I agree that a sticky would be great.

Thanks Holska!
After looking at the info it appears the ratings with the deviation & volatility numbers leave a lot of room for error. Granted, all of the servers are subject to similar numbers in deviation and volatility, so server ratings relative to each other is probably close.

Taking my server, Sorrow’s Furnace, as an example, if I’m understanding this correctly, our score is 1309.328. According to Glickman’s document the deviation would be the score plus or minus twice the deviation, so we end up with a 97% chance of Sorrow’s Furnace rating being between 899.954 and 1718.702. That means we’re either fourth from the bottom or ninth from the top, with 13 possible positions between.

Not very exacting, really, but since all the servers are subject to the same math, the positions are statistically correct, if not the actual scores. Probably.