Guild Wars 2

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

While the Glicko system is a staple of chess rankings and many other ranking systems, I think it is worth saying that the implementation of a tool is more important than the tool itself. This is where the famous phrase “The right tool for the job” comes in. For example, the screwdriver and hammer are both time tested tools, both are proven to be successful, but each has a very specific application. If I was trying to torque a screw, I wouldn’t want to use a hammer. Get the picture Jon? Someone’s qualifications don’t indicate the success of a system and it’s implementation. I’m by no means saying that the Glicko system isn’t the right application, because, well, I don’t know whether it is or isn’t. It doesn’t take a statistician however to say that the current ranking system is entirely broken… the biggest problem is that the servers are not all on equal footing. Why not create a system that encourages players to spread out to various servers… for example, why not control the maximum amount of players in WvW for one server by applying a cap based on the standard deviation of the lowest server? This would allow for some level of control, but won’t entirely prevent one team from outmatching another (by more than a standard deviation).

thiagoperne.7340:

Here’s the difference:

• WULD System = Only Tier 1 RED vs Tier 3 GREEN = Rank 3 vs Rank 7 = So after the new matchup begins as rank 3 went down and rank 7 went up = Rank 4 vs Rank 6
• New System = Random Matchups = Possibility of Rank 1 vs Rank 9 or Rank 10 vs Rank 18 etc..

The new system isn’t totally random, and while it would be possible for a rank 1 vs rank 9 matchup to occur, it’s not likely to happen very often, if at all. If I get some time I’ll try to calculate exactly how unlikely it is. I already did the math to show that rank 1 vs rank 12 is impossible given the current ratings (or at least, the ratings from 2 weeks ago). rank 3 vs rank 12 is also impossible, but I didn’t check the likelihood of rank 1 vs rank 9 occuring.

Under WULD, rank 3 vs rank 7 blowouts will happen every other week.

-ken

This is real pointless the best way is 1st goes up 3rd goes down, simply as that…

Well at least this makes things “kinda” fresh and if it doesn’t we got that dude’s resume so we can call him and troll our hearts content~

Poor Ken, bringing logic into this discussion. Good luck, buddy.

This system addresses the current problems

1. Stale match-ups turns WvW into a pve /pvdoor contest
2. Reshuffling of Team Colors. Sick and tired of GREENS winning since the day
the game launch. They usually composed of the same servers over and over.
3. More importantly. It’s an attempt to fix the sinkhole problems due to glicko2
problems experienced by the Tiers 1 and Tier 8 & 9. Some middle servers
like Ebay experience this too. It’s better than a total reset and can probably take
2-3 weeks to get it to be a lot more stable. Less extreme than total ratings reset.
4. Prevents those annoying bandwagons always transferring from green server to green
server. When their servers stop winning they’d transfer to another green server.
5. Time for change. WvW became boring. Hope they bring new dynamics soon esp
around the borderland areas. Underdog buff up is encouraged in case score gaps
becomes too large.

Other Alternatives:
a. 17 international tiers so that gap between tiers are minimizes.
More servers to fight and change places with
b. ratings up ratings down
+100 – 1st place
+ 0 – 2nd place
- 100 – 3rd place

I genuinely don’t care to whom belongs the kitten I’m jamming my boot up, so long as it’s entertaining for me to do so. If my server steam rolls the lesser servers they won’t even bother showing up while we tick every borderland and battleground tower and keep with an off chance to flip a camp some try-hard PUG decided to take. Low tier servers, from what I can tell from forum posts, barely grasp the meta of the game mode which has more to do with sheer numbers and coverage than literally anything else…except maybe these OP arrow carts, I swear it feels like I’m cheating sometimes…

I distinctly recall BWE’S where my server owned literally everything and as such it was literally no fun at all. I think ANet has no idea the vast difference in how the game is played on the different tiers, and this may make some matchups so painfully one sided it’ll turn newcomers off to the game mode, and bore the veterans into playing a different game.

It’d be so much better if it were treated like, you know, a sport. Where every team has to play each other and eventually there’s a play off to determine the finalists, and a championship game between the best ones…oh if only there was some kind of “observer mode” that allowed one to watch recorded matchups and a streaming live event for the championship chalk full of ads from sponsors…but I think that kind of technology and concept is clearly to difficult for any development team to achieve…

…oh wait what?
…seriously?
…ANet did something similar to that in GW1?

…why do I even bother…

DB was never moving to T1

Atleast now i can finally get my long sought revenge.

so what if the person that did the calculation has a PHD. the problem is that the idea in the first place even before calculation is out. people who wvw regularly all know this. dont need a phd for that, just jump in wvw each week and move up and down tiers with ur server and u will realise each tier is a world apart.

cant just random matchup ‘oh lets see who will win?!’ it will be total annihilation and no fun for the winning and losing side.

ArtC.1827:

Poor Ken, bringing logic into this discussion. Good luck, buddy.

I think he’s doing a great job.

I am hugely looking forward to this system starting just so we can see how this plays out.

Introduce a weekly matchup victor’s reward chest of Laurels, but have the winner be determined by ratings change for the week. So, just winning the PPT coverage war doesn’t necessarily translate into a match victory, rather performance versus expected value will be the yardstick.

For those that think this will encourage total domination QQfests, it’s a bit naive to think that the brainless piling on wouldn’t happen anyways. This way, those getting dominated actually will have incentive to keep trying.

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

I’ll bet Glickman is still paying student loans. Too bad Good Will Hunting hadn’t been released yet, he could have just paid some late charges at the library.

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

I love you.

Right, I’m going to have to point this out again.

This new system isn’t going to suddenly create uneven matchups, because those were already happening. The difference is that with this it will last for one week, instead of two months.

I’ll again point out the important part of the ratings that led to being locked in tiers. The system creates a prediction based on previous results, so if you got battered one week, the system expects the same to happen again. If you do roughly the same against the same servers (i.e. keep getting battered) your score doesn’t change, you don’t move tiers, and the same thing keeps happening over & over again. There may not be too many examples so far, but if it were allowed to continue then more examples would appear, and more servers would have completely unrealistic ratings and not be allowed to find their true level. ANet are saving you a lot of problems that you’re unable to envision because you’ll never experience them.

I’m don’t seeing any big difference between old system and new. I think anet trying to running around the problem and makes cosmetics changes with ~ 0 efficiency like NPC commander ….

Snowreap.5174:

thiagoperne.7340:

As much I hope this new system work out ok, I can’t see it happening. Unless they lower the spread matchups random math, even the chances are low, tier 1 servers can fight against tier 3 servers wich I don’t even need to say how bad that would be.

Why not a very basic ranking system(NO RATING)? Like this:

• FINISH MATCHUP 1ST PLACE = MOVE UP A TIER
• FINISH MATCHUP 2ND PLACE = REMAIN ON TIER
• FINISH MATCHUP 3RD PLACE = MOVE DOWN A TIER

This way a server will never ever face the same matchup in a row plus they will always be fighting against a server that you can compete as well as you are close each other(like an imaginary rating).

what you’ve described is usually called WULD (Winner Up Loser Down), and it turns out that WULD has even worse volatility.

WULD essentially guarantees that a T1 server will fight a T3 server every other week. Likewise, it will put a T2 server against a T4 server every other week, and so on.

To understand why, imagine three closely-matched servers in Tier 2. The winner will move up to tier 1, and the loser will go down to tier 3. meanwhile, the loser from tier 1 will come down to tier 2, and the winner from tier 3 will go up to tier 2.

The match will be a blowout. the T1 server will easily beat the T3 server (and probably the T2 as well), and the next week they’ll go back up and the T3 server will go back down, resulting in essentially the same matchups you started with. then a week later, the blowout will be repeated.

-ken

I don’t think that will be the result… you are correct in that a t1 server will face a t3 server… that’s not the issue. I think you will find that some of the lower servers are actually quite skilled or they are very fast to learn.

I think if you took away the numbers advantage… then a lot of lower servers are capable of beating the higher tiers.

For example, at one point we had ourselves (RiOT), a full VcY, VoTF, and XxX all on Ruins of Surmia server in tier 8. At that time, we could easily have beaten any of the tier 2 servers and it would only have been the numbers that would have prevented us beating the tier 1 servers.

Obviously, now that it is just RiOT and half of VcY left, it would be more challenging but it would not be a walk in the park for Gandara. We only lack the numbers of one decent size pvp guild during primetime to make the match more even. Similarly, while we have been here, the other servers around us have not been sat idle. GH now has FURY playing for them and I assume (forgive any arrogance here), that FURY and playing against ourselves has imparted a lot of knowledge to some of the other GH (and Dzag guilds) because they are now a lot more organised, more focused, and they have managed to harness much better the raw talent of their individual/small group players. The tier has improved immensely since we first arrived and I think it will be the same on a one up, one down system… what you will get is people learning from the tactics of other servers, adapting and eventually beating the dominant server in the group.

Incidently, any reasonable sized pvp guilds out there, would be more than welcome on RoS to replace those we recently lost… No lag (when wvw is working properly), no ques, and good competition from the other servers.

I guess this is all well and good, but I’ll miss my WvW family outside my server for sure. They were a good bunch even if they kept trying to kill me… that’s what family is for, right? Kick you when you’re down!

Personally I would have rather seen the Borderlands maps change so all three are unique or perhaps an entirely new map altogether. The matches for me weren’t becoming stale because the opponents were always the same; the matches sometimes aren’t fun simply because it’s the same old maps. Maybe this is in the works and I’ve missed it, but can this ever happen? Please?

I did like the suggestion about opening up free transfers but only to lower rank servers in an attempt to balance everyone out from the top down. It may or may not work, but it certainly would seem more appealing to make a move from “T1” to “T8” if it were free rather than a pile of gold. If you want to go from the bottom to the top, either pay up or fight!

A solution designed to placate whiners and solving nothing….

We WANT to fight the servers that are closest to us in power level and get the closest matchups the system can provide….the people claiming it is “boring” are just being silly.

If you want variation, you need to implement a handicapping system OR make a serious effort to equalize server populations. This will do nothing…

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

I would like to meet you and I would like to shake your hand.

I would prefer if you aligned servers in similar time zones together. Going to sleep, getting up and going to work, come home to find the battlegrounds has been taken over from another time zone server..gets old. And in my opinion has nothing to do with your formula you are using.

Areoh

I’m repeating what I’ve already said elsewhere, but this might be a better place to say it…

WvW Matchup Changes

Here’s what AN envisions… Using one of their examples, a possible matchup between rank 8 Kaineng going up against rank 1 Blackgate: Kaineng is inspired by this unexpected and otherwise impossible chance to test its mettle against #1 Blackgate, and by sheer force of will and mighty perseverance, and against all odds – they beat Blackgate! This in turn inspires others to win against overwhelming odds.

More likely scenario: Kaineng gets thoroughly stomped by Blackgate over the weekend after reset, are totally dispirited and give up on WvW for the rest of the week, with many people wondering when the next best pvp MMO will be released, so they’ll have something else to play during weeks like this.

Ok, I simulated 100,000,000 random matchups to see what’s likely to happen, and what’s not likely. I started with the ratings and volatility values from this page:
https://www.guildwars2.com/en/news/big-changes-coming-to-wvw-matchups/

Blackgate
65.6% chance of being matched against Jade Quarry
54.6% chance of being matched against Sanctum of Rall
29.8% chance of being matched against Dragonbrand
19.6% chance of being matched against Fort Aspenwood
18.4% chance of being matched against Tarnished Coast
4.5% chance of being matched against Maguuma
4.4% chance of being matched against Kaineng
2.1% chance of being matched against Yak’s Bend

Ehmry Bay
23.9% vs Anvil Rock
23.3% vs Borlis Pass
22.9% vs Stormbluff Isle
22.9% vs Crystal Desert
18.3% vs Darkhaven
17.3% vs Sea of Sorrows
14.0% vs Yak’s Bend
12.0% vs Isle of Janthir
11.1% vs Sorrow’s Furnace
10.4% vs Kaineng
10.0% vs Maguuma
3.8% vs Devona’s Rest
3.4% vs Gate of Madness
2.8% vs Northern Shiverpeaks

Eredon Terrace
57.8% vs Ferguson’s Crossing
49.1% vs Henge of Denravi
25.4% vs Northern Shiverpeaks
23.5% vs Gate of Madness
22.3% vs Devona’s Rest
7.9% vs Isle of Janthir
3.1% vs Darkhaven

I hope this helps people understand what matchups are likely under the new system, and what matchups are unlikely.

Note that the percentages add up to 200% because in each match you will face 2 opponents.

-ken

Snowreap.5174:

I hope this helps people understand what matchups are likely under the new system, and what matchups are unlikely.

Here’s the issue I see with this new system…

High ranked servers are going to hemorrhage rating points every week to lower ranked servers just because they’re in the same matchup. In the current system, every time a server drops to a new tier they bleed ratings points for a couple of weeks even if they win the match. Lower ranked servers will soak up these free rating points and every server moves more towards the middle of the curve.

However, Blackgate is still Blackgate and Eredon Terrace is still Eredon Terrace. These two servers should never be matched up but artificial gains and losses to server ratings could easily result in this matchup a few months down the road. All the server ratings are going to bunch up around the middle of the scale and the matchups will be decided by the random number alone.

thiagoperne.7340:

Snowreap.5174:

thiagoperne.7340:

As much I hope this new system work out ok, I can’t see it happening. Unless they lower the spread matchups random math, even the chances are low, tier 1 servers can fight against tier 3 servers wich I don’t even need to say how bad that would be.

Why not a very basic ranking system(NO RATING)? Like this:

• FINISH MATCHUP 1ST PLACE = MOVE UP A TIER
• FINISH MATCHUP 2ND PLACE = REMAIN ON TIER
• FINISH MATCHUP 3RD PLACE = MOVE DOWN A TIER

This way a server will never ever face the same matchup in a row plus they will always be fighting against a server that you can compete as well as you are close each other(like an imaginary rating).

what you’ve described is usually called WULD (Winner Up Loser Down), and it turns out that WULD has even worse volatility.

WULD essentially guarantees that a T1 server will fight a T3 server every other week. Likewise, it will put a T2 server against a T4 server every other week, and so on.

To understand why, imagine three closely-matched servers in Tier 2. The winner will move up to tier 1, and the loser will go down to tier 3. meanwhile, the loser from tier 1 will come down to tier 2, and the winner from tier 3 will go up to tier 2.

The match will be a blowout. the T1 server will easily beat the T3 server (and probably the T2 as well), and the next week they’ll go back up and the T3 server will go back down, resulting in essentially the same matchups you started with. then a week later, the blowout will be repeated.

-ken

Here’s the difference:

• WULD System = Only Tier 1 RED vs Tier 3 GREEN = Rank 3 vs Rank 7 = So after the new matchup begins as rank 3 went down and rank 7 went up = Rank 4 vs Rank 6
• New System = Random Matchups = Possibility of Rank 1 vs Rank 9 or Rank 10 vs Rank 18 etc..

My thread (go check it out if you want), shows that with the new system, if they make the random number range larger, then you won’t get (at least on a very very rare basis) a rank 1 vs a rank 9. If they lower the random number range to like -0.5 to 0.5, then the most likely server ranking difference is 2 with a probability of ~50%. Then a spread of within the next tier above/below is ~50%. So, there is only a very rare chance to get a match with servers 2 tiers away. So with this slight chance, the matches aren’t as stagnant, but you’re much less likely to get total blow outs.

thiagoperne.7340:

Here’s the difference:

• WULD System = Only Tier 1 RED vs Tier 3 GREEN = Rank 3 vs Rank 7 = So after the new matchup begins as rank 3 went down and rank 7 went up = Rank 4 vs Rank 6
• New System = Random Matchups = Possibility of Rank 1 vs Rank 9 or Rank 10 vs Rank 18 etc..

WULD is a terrible system.

It guarantees T1 vs. T3 every other week, and yet you berate the new system (which you don’t even seem to understand) for having the possibility to cause that to happen. You’re effectively stating that a 100% chance of a T1vst3 server match up every other week (WULD) is better than a statistically improbable chance of it happening (Chaotic Deviation).

Vincent Von Demetri.1450:

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

While the Glicko system is a staple of chess rankings and many other ranking systems, I think it is worth saying that the implementation of a tool is more important than the tool itself. This is where the famous phrase “The right tool for the job” comes in. For example, the screwdriver and hammer are both time tested tools, both are proven to be successful, but each has a very specific application. If I was trying to torque a screw, I wouldn’t want to use a hammer. Get the picture Jon? Someone’s qualifications don’t indicate the success of a system and it’s implementation. I’m by no means saying that the Glicko system isn’t the right application, because, well, I don’t know whether it is or isn’t. It doesn’t take a statistician however to say that the current ranking system is entirely broken… the biggest problem is that the servers are not all on equal footing. Why not create a system that encourages players to spread out to various servers… for example, why not control the maximum amount of players in WvW for one server by applying a cap based on the standard deviation of the lowest server? This would allow for some level of control, but won’t entirely prevent one team from outmatching another (by more than a standard deviation).

Even though this was the creator of the glicko system, are they using it for GW2? Do you not think Mark Glickman could of possibly create a system specifically for WvW.

Snowreap.5174:

Ok, I simulated 100,000,000 random matchups to see what’s likely to happen, and what’s not likely. I started with the ratings and volatility values from this page:
https://www.guildwars2.com/en/news/big-changes-coming-to-wvw-matchups/

Blackgate
65.6% chance of being matched against Jade Quarry
54.6% chance of being matched against Sanctum of Rall
29.8% chance of being matched against Dragonbrand
19.6% chance of being matched against Fort Aspenwood
18.4% chance of being matched against Tarnished Coast
4.5% chance of being matched against Maguuma
4.4% chance of being matched against Kaineng
2.1% chance of being matched against Yak’s Bend

Ehmry Bay
23.9% vs Anvil Rock
23.3% vs Borlis Pass
22.9% vs Stormbluff Isle
22.9% vs Crystal Desert
18.3% vs Darkhaven
17.3% vs Sea of Sorrows
14.0% vs Yak’s Bend
12.0% vs Isle of Janthir
11.1% vs Sorrow’s Furnace
10.4% vs Kaineng
10.0% vs Maguuma
3.8% vs Devona’s Rest
3.4% vs Gate of Madness
2.8% vs Northern Shiverpeaks

Eredon Terrace
57.8% vs Ferguson’s Crossing
49.1% vs Henge of Denravi
25.4% vs Northern Shiverpeaks
23.5% vs Gate of Madness
22.3% vs Devona’s Rest
7.9% vs Isle of Janthir
3.1% vs Darkhaven

I hope this helps people understand what matchups are likely under the new system, and what matchups are unlikely.

Note that the percentages add up to 200% because in each match you will face 2 opponents.

-ken

By far the most useful post I have seen on the forums in a while. Thanks for sharing!

A lot of over reacting going on.

I support these changes and applaud anet for them

JonPeters.5630:

timidobserver.7925:

I have a couple of serious questions.
1. Did the person that came up with the math for this graduate from high school.
2. To prove it, please upload a copy of your high school diploma.

How about just a copy of their resume. I don’t see his high school experience on there. Hopefully the other stuff can count as relevant experience for your assessment.

http://www.glicko.net/cv.html

lol snap right there!
i’m actually very happy about the change, because the matchups were getting stuck and boring. with this change we will slowly see changes and less mass transfers while other servers are dying. yes, there will be chaos, but i’d rather have the chance to meet new servers all the time and not the same server for over 10 weeks.the system seems interesting and i cant wait for fresh air finally!

Did they already give an exact time when they want to make these changes?
Will the current matchup last until tuesday or friday?

Mystery.7392:

Did they already give an exact time when they want to make these changes?
Will the current matchup last until tuesday or friday?

Match ups will change on the 31st, Friday.

If coverage is the main contributing factor to the overall score per interval, why not use it (or rather wvw activity) as weight for point gain? That way points are worth more if meaningful opposition was present when they were gained.

One of the weak elements of WvW is problems finding balanced, fair, fun fights and you are making this problem substantially worse. This is because WvW has always been to to a certain extent a numbers game or about “coverage”.

This new system will result in even more blowouts. PUGs on the weaker servers will refuse to log in as nobody wants to be annihilated. Then the stronger guilds will stop logging in as they can’t have fun being constantly outnumbered, especially without the PUG support.

Do you guys actually WvW – and I’m not talking a solo thief. Do you ever face overwhelming odds?

Do you have a clue about human psychology?

The servers are largely faceless entities. While they differ somewhat, things change so fast from week to week that the continuity is not really there in terms of opponents. You are overestimating the benefit of variety and underestimating the willingness of people to handle blowouts. When a server is winning, more peeple log on. When they start getting blown out, they stop logging in.

I can’t see this working. It’s not a 1 hour tennis match with a different friend. It’s Mini-me playing basketabll against LeBron for a bloody week.

Matchups outside of your comfort zone are necessary now and then in order for the server ratings to be meaningful. I think people fear that these outlier matches will be more common than they actually are.

I ran another set of random trials overnight (10 billion trials this time). Here’s the full set of data, which will hopefully make it easier to see how the distribution of likely and unlikely matches looks (for NA only, sorry, I don’t have volatility data for EU at hand).

(split into 4 posts because it won’t all fit in 1)

Blackgate
65.657217 Jade Quarry
54.560907 Sanctum of Rall
29.791629 Dragonbrand
19.581320 Fort Aspenwood
18.370786 Tarnished Coast
4.508521 Maguuma
4.363235 Kaineng
2.108594 Yak’s Bend
1.018578 Sea of Sorrows
0.024463 Crystal Desert
0.013282 Stormbluff Isle
0.001454 Ehmry Bay
0.000009 Anvil Rock
0.000005 Borlis Pass
0.000000 Darkhaven
0.000000 Isle of Janthir
0.000000 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Jade Quarry
65.657217 Blackgate
51.751898 Sanctum of Rall
28.765118 Dragonbrand
19.873674 Fort Aspenwood
18.797098 Tarnished Coast
5.522435 Maguuma
5.355067 Kaineng
2.812321 Yak’s Bend
1.386921 Sea of Sorrows
0.045210 Crystal Desert
0.028149 Stormbluff Isle
0.004847 Ehmry Bay
0.000030 Anvil Rock
0.000016 Borlis Pass
0.000000 Darkhaven
0.000000 Isle of Janthir
0.000000 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Sanctum of Rall
54.560907 Blackgate
51.751898 Jade Quarry
26.491010 Dragonbrand
21.357870 Fort Aspenwood
20.622216 Tarnished Coast
8.496510 Maguuma
8.266378 Kaineng
5.124415 Yak’s Bend
2.914704 Sea of Sorrows
0.216044 Crystal Desert
0.159978 Stormbluff Isle
0.037735 Ehmry Bay
0.000214 Anvil Rock
0.000117 Borlis Pass
0.000003 Darkhaven
0.000000 Isle of Janthir
0.000000 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Dragonbrand
29.791629 Blackgate
28.765118 Jade Quarry
26.491010 Sanctum of Rall
29.270935 Fort Aspenwood
28.819525 Tarnished Coast
16.060183 Maguuma
15.697157 Kaineng
11.941738 Yak’s Bend
8.883480 Sea of Sorrows
2.109081 Crystal Desert
1.680902 Stormbluff Isle
0.457437 Ehmry Bay
0.014877 Anvil Rock
0.012351 Borlis Pass
0.003892 Darkhaven
0.000428 Isle of Janthir
0.000258 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Fort Aspenwood
19.581320 Blackgate
19.873674 Jade Quarry
21.357870 Sanctum of Rall
29.270935 Dragonbrand
30.528809 Tarnished Coast
19.750503 Maguuma
19.370550 Kaineng
16.240203 Yak’s Bend
13.230901 Sea of Sorrows
4.940582 Crystal Desert
4.183725 Stormbluff Isle
1.406415 Ehmry Bay
0.121322 Anvil Rock
0.103734 Borlis Pass
0.033601 Darkhaven
0.003663 Isle of Janthir
0.002193 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Tarnished Coast
18.370786 Blackgate
18.797098 Jade Quarry
20.622216 Sanctum of Rall
28.819525 Dragonbrand
30.528809 Fort Aspenwood
20.202149 Maguuma
19.824850 Kaineng
16.847125 Yak’s Bend
13.866970 Sea of Sorrows
5.451373 Crystal Desert
4.663470 Stormbluff Isle
1.657292 Ehmry Bay
0.159725 Anvil Rock
0.136767 Borlis Pass
0.044181 Darkhaven
0.004795 Isle of Janthir
0.002870 Sorrow’s Furnace
0.000000 Devona’s Rest
0.000000 Gate of Madness
0.000000 Northern Shiverpeaks
0.000000 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Maguuma
4.508521 Blackgate
5.522435 Jade Quarry
8.496510 Sanctum of Rall
16.060183 Dragonbrand
19.750503 Fort Aspenwood
20.202149 Tarnished Coast
27.679361 Kaineng
26.728516 Yak’s Bend
23.977925 Sea of Sorrows
16.009919 Crystal Desert
15.057389 Stormbluff Isle
9.977743 Ehmry Bay
2.676683 Anvil Rock
2.316407 Borlis Pass
0.789715 Darkhaven
0.131959 Isle of Janthir
0.095280 Sorrow’s Furnace
0.007936 Devona’s Rest
0.006414 Gate of Madness
0.004436 Northern Shiverpeaks
0.000017 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Kaineng
4.363235 Blackgate
5.355067 Jade Quarry
8.266378 Sanctum of Rall
15.697157 Dragonbrand
19.370550 Fort Aspenwood
19.824850 Tarnished Coast
27.679361 Maguuma
26.553397 Yak’s Bend
23.931394 Sea of Sorrows
16.411510 Crystal Desert
15.480485 Stormbluff Isle
10.409980 Ehmry Bay
2.935280 Anvil Rock
2.545272 Borlis Pass
0.881178 Darkhaven
0.155988 Isle of Janthir
0.114535 Sorrow’s Furnace
0.010277 Devona’s Rest
0.008318 Gate of Madness
0.005765 Northern Shiverpeaks
0.000022 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Yak’s Bend
2.108594 Blackgate
2.812321 Jade Quarry
5.124415 Sanctum of Rall
11.941738 Dragonbrand
16.240203 Fort Aspenwood
16.847125 Tarnished Coast
26.728516 Maguuma
26.553397 Kaineng
25.290745 Sea of Sorrows
20.528387 Crystal Desert
19.624791 Stormbluff Isle
14.015474 Ehmry Bay
5.006062 Anvil Rock
4.472772 Borlis Pass
1.813313 Darkhaven
0.446275 Isle of Janthir
0.352316 Sorrow’s Furnace
0.039668 Devona’s Rest
0.031910 Gate of Madness
0.021895 Northern Shiverpeaks
0.000083 Henge of Denravi
0.000000 Ferguson’s Crossing
0.000000 Eredon Terrace

Sea of Sorrows
1.018578 Blackgate
1.386921 Jade Quarry
2.914704 Sanctum of Rall
8.883480 Dragonbrand
13.230901 Fort Aspenwood
13.866970 Tarnished Coast
23.977925 Maguuma
23.931394 Kaineng
25.290745 Yak’s Bend
24.527165 Crystal Desert
23.603751 Stormbluff Isle
17.326625 Ehmry Bay
7.438645 Anvil Rock
6.828150 Borlis Pass
3.490654 Darkhaven
1.113734 Isle of Janthir
0.911899 Sorrow’s Furnace
0.109596 Devona’s Rest
0.087915 Gate of Madness
0.059958 Northern Shiverpeaks
0.000269 Henge of Denravi
0.000015 Ferguson’s Crossing
0.000007 Eredon Terrace

Crystal Desert
0.024463 Blackgate
0.045210 Jade Quarry
0.216044 Sanctum of Rall
2.109081 Dragonbrand
4.940582 Fort Aspenwood
5.451373 Tarnished Coast
16.009919 Maguuma
16.411510 Kaineng
20.528387 Yak’s Bend
24.527165 Sea of Sorrows
29.029871 Stormbluff Isle
22.901400 Ehmry Bay
16.123572 Anvil Rock
15.555905 Borlis Pass
11.461474 Darkhaven
6.461165 Isle of Janthir
5.715404 Sorrow’s Furnace
0.995722 Devona’s Rest
0.829001 Gate of Madness
0.610853 Northern Shiverpeaks
0.035701 Henge of Denravi
0.011466 Ferguson’s Crossing
0.004736 Eredon Terrace

Stormbluff Isle
0.013282 Blackgate
0.028149 Jade Quarry
0.159978 Sanctum of Rall
1.680902 Dragonbrand
4.183725 Fort Aspenwood
4.663470 Tarnished Coast
15.057389 Maguuma
15.480485 Kaineng
19.624791 Yak’s Bend
23.603751 Sea of Sorrows
29.029871 Crystal Desert
22.921988 Ehmry Bay
17.243594 Anvil Rock
16.690345 Borlis Pass
12.532646 Darkhaven
7.321464 Isle of Janthir
6.533374 Sorrow’s Furnace
1.274371 Devona’s Rest
1.071772 Gate of Madness
0.804372 Northern Shiverpeaks
0.055201 Henge of Denravi
0.017789 Ferguson’s Crossing
0.007290 Eredon Terrace

Ehmry Bay
0.001454 Blackgate
0.004847 Jade Quarry
0.037735 Sanctum of Rall
0.457437 Dragonbrand
1.406415 Fort Aspenwood
1.657292 Tarnished Coast
9.977743 Maguuma
10.409980 Kaineng
14.015474 Yak’s Bend
17.326625 Sea of Sorrows
22.901400 Crystal Desert
22.921988 Stormbluff Isle
23.891634 Anvil Rock
23.297344 Borlis Pass
18.283167 Darkhaven
12.031103 Isle of Janthir
11.078690 Sorrow’s Furnace
3.763294 Devona’s Rest
3.353993 Gate of Madness
2.753965 Northern Shiverpeaks
0.292330 Henge of Denravi
0.095918 Ferguson’s Crossing
0.040173 Eredon Terrace

Anvil Rock
0.000009 Blackgate
0.000030 Jade Quarry
0.000214 Sanctum of Rall
0.014877 Dragonbrand
0.121322 Fort Aspenwood
0.159725 Tarnished Coast
2.676683 Maguuma
2.935280 Kaineng
5.006062 Yak’s Bend
7.438645 Sea of Sorrows
16.123572 Crystal Desert
17.243594 Stormbluff Isle
23.891634 Ehmry Bay
28.158492 Borlis Pass
23.922513 Darkhaven
19.067790 Isle of Janthir
18.254133 Sorrow’s Furnace
10.652433 Devona’s Rest
10.135871 Gate of Madness
9.325549 Northern Shiverpeaks
2.745221 Henge of Denravi
1.329529 Ferguson’s Crossing
0.796821 Eredon Terrace

Borlis Pass
0.000005 Blackgate
0.000016 Jade Quarry
0.000117 Sanctum of Rall
0.012351 Dragonbrand
0.103734 Fort Aspenwood
0.136767 Tarnished Coast
2.316407 Maguuma
2.545272 Kaineng
4.472772 Yak’s Bend
6.828150 Sea of Sorrows
15.555905 Crystal Desert
16.690345 Stormbluff Isle
23.297344 Ehmry Bay
28.158492 Anvil Rock
23.986152 Darkhaven
19.529510 Isle of Janthir
18.750156 Sorrow’s Furnace
11.242174 Devona’s Rest
10.726757 Gate of Madness
9.915894 Northern Shiverpeaks
3.154464 Henge of Denravi
1.600135 Ferguson’s Crossing
0.977082 Eredon Terrace

Darkhaven
0.000000 Blackgate
0.000000 Jade Quarry
0.000003 Sanctum of Rall
0.003892 Dragonbrand
0.033601 Fort Aspenwood
0.044181 Tarnished Coast
0.789715 Maguuma
0.881178 Kaineng
1.813313 Yak’s Bend
3.490654 Sea of Sorrows
11.461474 Crystal Desert
12.532646 Stormbluff Isle
18.283167 Ehmry Bay
23.922513 Anvil Rock
23.986152 Borlis Pass
22.669795 Isle of Janthir
22.114733 Sorrow’s Furnace
15.276686 Devona’s Rest
14.780849 Gate of Madness
13.998679 Northern Shiverpeaks
6.485715 Henge of Denravi
4.304137 Ferguson’s Crossing
3.126920 Eredon Terrace

Isle of Janthir
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000428 Dragonbrand
0.003663 Fort Aspenwood
0.004795 Tarnished Coast
0.131959 Maguuma
0.155988 Kaineng
0.446275 Yak’s Bend
1.113734 Sea of Sorrows
6.461165 Crystal Desert
7.321464 Stormbluff Isle
12.031103 Ehmry Bay
19.067790 Anvil Rock
19.529510 Borlis Pass
22.669795 Darkhaven
24.518666 Sorrow’s Furnace
19.824457 Devona’s Rest
19.430264 Gate of Madness
18.800056 Northern Shiverpeaks
11.410986 Henge of Denravi
9.169243 Ferguson’s Crossing
7.908661 Eredon Terrace

Sorrow’s Furnace
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000258 Dragonbrand
0.002193 Fort Aspenwood
0.002870 Tarnished Coast
0.095280 Maguuma
0.114535 Kaineng
0.352316 Yak’s Bend
0.911899 Sea of Sorrows
5.715404 Crystal Desert
6.533374 Stormbluff Isle
11.078690 Ehmry Bay
18.254133 Anvil Rock
18.750156 Borlis Pass
22.114733 Darkhaven
24.518666 Isle of Janthir
20.468579 Devona’s Rest
20.108875 Gate of Madness
19.524489 Northern Shiverpeaks
12.322618 Henge of Denravi
10.165482 Ferguson’s Crossing
8.965450 Eredon Terrace

Devona’s Rest
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.007936 Maguuma
0.010277 Kaineng
0.039668 Yak’s Bend
0.109596 Sea of Sorrows
0.995722 Crystal Desert
1.274371 Stormbluff Isle
3.763294 Ehmry Bay
10.652433 Anvil Rock
11.242174 Borlis Pass
15.276686 Darkhaven
19.824457 Isle of Janthir
20.468579 Sorrow’s Furnace
24.948011 Gate of Madness
25.001552 Northern Shiverpeaks
22.004490 Henge of Denravi
22.053395 Ferguson’s Crossing
22.327361 Eredon Terrace

Gate of Madness
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.006414 Maguuma
0.008318 Kaineng
0.031910 Yak’s Bend
0.087915 Sea of Sorrows
0.829001 Crystal Desert
1.071772 Stormbluff Isle
3.353993 Ehmry Bay
10.135871 Anvil Rock
10.726757 Borlis Pass
14.780849 Darkhaven
19.430264 Isle of Janthir
20.108875 Sorrow’s Furnace
24.948011 Devona’s Rest
25.191816 Northern Shiverpeaks
22.744338 Henge of Denravi
23.055667 Ferguson’s Crossing
23.488228 Eredon Terrace

Northern Shiverpeaks
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.004436 Maguuma
0.005765 Kaineng
0.021895 Yak’s Bend
0.059958 Sea of Sorrows
0.610853 Crystal Desert
0.804372 Stormbluff Isle
2.753965 Ehmry Bay
9.325549 Anvil Rock
9.915894 Borlis Pass
13.998679 Darkhaven
18.800056 Isle of Janthir
19.524489 Sorrow’s Furnace
25.001552 Devona’s Rest
25.191816 Gate of Madness
23.928621 Henge of Denravi
24.680521 Ferguson’s Crossing
25.371579 Eredon Terrace

Henge of Denravi
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.000017 Maguuma
0.000022 Kaineng
0.000083 Yak’s Bend
0.000269 Sea of Sorrows
0.035701 Crystal Desert
0.055201 Stormbluff Isle
0.292330 Ehmry Bay
2.745221 Anvil Rock
3.154464 Borlis Pass
6.485715 Darkhaven
11.410986 Isle of Janthir
12.322618 Sorrow’s Furnace
22.004490 Devona’s Rest
22.744338 Gate of Madness
23.928621 Northern Shiverpeaks
45.675469 Ferguson’s Crossing
49.144457 Eredon Terrace

Ferguson’s Crossing
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.000000 Maguuma
0.000000 Kaineng
0.000000 Yak’s Bend
0.000015 Sea of Sorrows
0.011466 Crystal Desert
0.017789 Stormbluff Isle
0.095918 Ehmry Bay
1.329529 Anvil Rock
1.600135 Borlis Pass
4.304137 Darkhaven
9.169243 Isle of Janthir
10.165482 Sorrow’s Furnace
22.053395 Devona’s Rest
23.055667 Gate of Madness
24.680521 Northern Shiverpeaks
45.675469 Henge of Denravi
57.841235 Eredon Terrace

Eredon Terrace
0.000000 Blackgate
0.000000 Jade Quarry
0.000000 Sanctum of Rall
0.000000 Dragonbrand
0.000000 Fort Aspenwood
0.000000 Tarnished Coast
0.000000 Maguuma
0.000000 Kaineng
0.000000 Yak’s Bend
0.000007 Sea of Sorrows
0.004736 Crystal Desert
0.007290 Stormbluff Isle
0.040173 Ehmry Bay
0.796821 Anvil Rock
0.977082 Borlis Pass
3.126920 Darkhaven
7.908661 Isle of Janthir
8.965450 Sorrow’s Furnace
22.327361 Devona’s Rest
23.488228 Gate of Madness
25.371579 Northern Shiverpeaks
49.144457 Henge of Denravi
57.841235 Ferguson’s Crossing

Nice work.
I’d love to see that for EU.
You can get the data here I guess? – https://leaderboards.guildwars2.com/en/eu/wvw

Sunspots.9861:

I find it amusing that so much of the support of this is coming from the lower tier servers. Oh well, you wanted it, you got it.

I know a lot of people that are excited about this and I don’t think it’s gonna be a bad thing either. We’re aware of the tier differences and have a long history of being completely blown out of a match, so I do think we know what we’re in for =)

Mystery.7392:

I’d love to see that for EU.
You can get the data here I guess? – https://leaderboards.guildwars2.com/en/eu/wvw

thanks for the link. check back in 6 hours or so, I’m running the numbers for EU now…

-ken

why not mix EU servers with US servers?

BioHaZarD.7904:

why not mix EU servers with US servers?

Different reset times for one. EU reset is mid-day server time.

BioHaZarD.7904:

why not mix EU servers with US servers?

EU sleeps while NA is playing, NA is at work while EU is playing. What should such a match be good for?

I applaud Snowreap work on this and showing that there is little chance that at the start the lower rank servers (if they are indeed deserving of said rank not to say anything bad about those servers). Have little to no chance that they will be put up against the server that is out of there league.

It is important to Keep in mind that Snowreap calculations are based upon the ratings of the servers represent the correct strength of those servers. This may or may not be necessarily the case.

https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/Statistics-of-the-new-ranking-system/first#post2082624

the current ratings are not correct; this is one of the reasons this change is being made — in order to allow some matches to occur outside of your normal tier, so that the ratings can be corrected.

see this post, from page 2 of this thread:
https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/New-matchup-system-official-info/2082413

-ken

Snowreap.5174:

the current ratings are not correct; this is one of the reasons this change is being made — in order to allow some matches to occur outside of your normal tier, so that the ratings can be corrected.

see this post, from page 2 of this thread:
https://forum-en.gw2archive.eu/forum/pvp/wuvwuv/New-matchup-system-official-info/2082413

-ken

Snowreap, thanks for showing those match ups! In my thread, where I showed similar statistics, I just showed the odds for matches with rank differences, so it was nice to see that, for a given server, what the exact odds for every server were (something I admit I was a bit lazy to do). If you want to add those to my thread so we can get all the statistics on one big page, that’d be amazing, and I’d be very very grateful!

The problem isn’t the first week of matchups. The problem will be a few weeks down the line when higher ranked servers have been bled of ratings points by being matched up with lower ranked servers. At that point it’s very possible for really ugly matches to occur.

What will happen with Vabbi [27th]? They are like 600 points behind the 25th rank.

style.6173:

The more I think about this, the more I like it. Those who need to plan out their reset nights will hate it of course, but it is also exciting. Until the reset happens, you have no idea who you are playing and you have no idea what color you will be. Should be really fun.

Doesn’t matter, all tier 1 servers have organized guilds to cover each of the maps on reset, good luck to those facing them, they will get rolled and most probably picked on for easy points.

From what I see in the article, the change will only make things worse if you look at the potential matchup listing they gave in that article.

JQ would steamroll its opposition during asian/oceanic time while fairing well against DB’s NA and good luck to FA in that matchup.

The 2nd matchup, SoR BG & Kaining, i think would end up in a 1 on 1 between SoR and BG with Kaining getting the short end of the stick. Kaining would get wiped off maps the moment they show an outmanned buff and have a hard time ticking above 200 ppt even during their primetime. SoR has the strongest NA and a large one at that while BG has good numbers around the clock.

And beyond that, all top tier servers will start bleeding rating points because it is almost impossible to dominate with a difference higher than 100k points even in an extremely unbalanced matchup, causing these top servers to be matched agains even weaker servers than the ones they will destroy…

I know with the old tier system they had an issue making it impossible for servers to change tiers as weeks went by, but at least matchups were pretty even or at least servers were matched up with those that fit their ability best. I feel this new change is horrible, all they needed to do was to tweak their current rating system to actually make it possible for a server to go up or drop down from a tier…