Guild Wars 2

Oceanic time is fun.

LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL

Showing off your progess in the achivement tabs I see? :P

On topic: How do you think the losing team felt? The teams were similar in skill? Fix this or I’ll keep owning noobs 500/0 and make them quit your game…

I’m guessing they felt bad.

ya its cause its 2 late and you only have 2 warriors on each side.
Prime time every1 knows its 4 or 5 warriors on each side then its balanced

Hey that’s me Beefy Boy. Yeah, it’s sad that they stack one team. I bet one or two of those guys will never return to sPvP.

If people were ranked as follows

10, 9, 8, 7, 6, 5, 4, 3, 2, 1

It would be nice to see one team as
10, 8, 6, 4, 2 and the other team as 9, 7, 5, 3, 1

That way there seems to be an even match up even if the rank 1 and 2 guys are top 100 and the 9 and 10 guys are <85%

Would feel like a hyper-carry for the top rank players, but it’s soloqueue, every game should be challenging.

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

One thing to keep in mind is we perform matching based on a Glicko v2 rating, not player rank. So even though higher ranked players also tend to have a higher rating, it isn’t always the case.

@Cheri Wolf
I wrote a simulator to test several different ways of solving the problem, one of which is the way you suggested, so I’m glad we’re on the same page.

While that method improved things it also tends to weigh the match in favor of the first team. The current iteration we’re working on now is very similar, but greatly reduces the bias.

I agree that rank =! rating, although in this instance all of us on blue were in the 150-300 range. While I don’t know the rating of the enemy team, I seriously have my doubts it was anywhere reasonably close.

I attribute this to what the OP said, Oceanic times with 15 minute Q’s.
I really do appreciate the dev feedback on this issue though. Glad you are considering alternative methods!

Yeah wut he said.

I meant rated, not rank. so yeah exactly wut ODell was stating.

Ah, yeah I see the balance.

Maybe something like

10, 7, 6, 3, 2
9, 8, 5, 4, 1

Cool deal, I am very pleased to see this is being addressed.

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

One thing to keep in mind is we perform matching based on a Glicko v2 rating, not player rank. So even though higher ranked players also tend to have a higher rating, it isn’t always the case.

@Cheri Wolf
I wrote a simulator to test several different ways of solving the problem, one of which is the way you suggested, so I’m glad we’re on the same page.

While that method improved things it also tends to weigh the match in favor of the first team. The current iteration we’re working on now is very similar, but greatly reduces the bias.

How about making a system that matches people the same way it works with picks on mobas?
so, a 10 9 8 7 6 5 4 3 2 1 example would become:
Team 1: 10, 7, 6, 3, 2 (28)
Team 2: 9, 8, 5, 4, 1 (27)

I don’t think you can get more balanced than this, honestly.

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

One thing to keep in mind is we perform matching based on a Glicko v2 rating, not player rank. So even though higher ranked players also tend to have a higher rating, it isn’t always the case.

@Cheri Wolf
I wrote a simulator to test several different ways of solving the problem, one of which is the way you suggested, so I’m glad we’re on the same page.

While that method improved things it also tends to weigh the match in favor of the first team. The current iteration we’re working on now is very similar, but greatly reduces the bias.

I appreciate that this is being worked on, but can we have an explanation as to why this is just happening now? It’s hard to justify an argument with the “PvP is the red headed stepchild” crowd when you treat PvP like a redheaded stepchild.

doesnt matter how your matching system works, when there is only a few people queuing, you will not get desired results…

Justin ODell.9517:

evilapprentice.6379:

… can we have an explanation as to why this is just happening now?

It can happen for a lot of reasons, and I honestly don’t know exactly what happened here.

Speculating, it’s possible the 5th user that dropped was a very highly rated player. This can confuse the match maker because the average skill rating of the red team (adjusted for the time each team has been waiting in the queue) could be very near the average rating for the blue team.

Improving match making to include parameters beyond just skill rating is on our radar.

Allow me to preface this largely negative statement by thanking you for responding to me – regardless of how I feel about the game, I appreciate you taking the time to address my concerns, and post on the boards in general.

My point wasn’t exactly “Why is this happening now”, but rather, why was this a possibility at all? I understand the matchmaking system will never be perfect, and that there will always be outliers. The situation we’re looking at is not one comprised of outliers – it’s SOP – it’s how matchmaking works. I don’t understand how such an easily exploitable, obviously flawed matchmaking system made it past internal QA, and out into the open field – we aren’t playing a beta, we’re over a year into release.

If Anet needs to go back to the drawing board, that is disappointing, but we’d rather hear it straight up from the company themselves than infer it from the way things have been handled over the last 3 months. We’re sick of just assuming everything will be poorly handled, because precedence has taught us that is the safe bet.

What do you need? What do you need from us, the community, to improve the issues that have become commonplace in the PvP environment? I understand that there is ALOT of negative crap to wade through on the boards, but some of us are genuinly looking forward to when this game is in a solidy designed, enjoyable game to PvP in – it’s been over a year, and it still isn’t there. We’re interesting in helping, but the only word we’ve heard thus far is “keep the constructive criticism coming!”, and that’s hard to take seriously when it doesn’t appear as though that’s helping at all.

Justin ODell.9517:

evilapprentice.6379:

… can we have an explanation as to why this is just happening now?

It can happen for a lot of reasons, and I honestly don’t know exactly what happened here.

Speculating, it’s possible the 5th user that dropped was a very highly rated player. This can confuse the match maker because the average skill rating of the red team (adjusted for the time each team has been waiting in the queue) could be very near the average rating for the blue team.

Improving match making to include parameters beyond just skill rating is on our radar.

I could be wrong here, but I think he’s asking why is this fix or possible fixes only JUST NOW being looked into? This has been a huge issue since the matchmaking’s introduction. How are we to believe that SPVP is really on anyone’s radar when fundamental issues are overlooked for so long.

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

Does that refer to the following balance patch on the 15th of October or does this refer to some in house as of yet to be determined date?

Brutal Arts.6307:

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

Does that refer to the following balance patch on the 15th of October or does this refer to some in house as of yet to be determined date?

I was wondering the exact same thing.

@Justin ODell
Could you explain the process of how hard it would be to server program a PTR and test MMR?
And then weigh in having the community doing all the testing work.
How much time would devs had saved from testing if they had a PTR over a year ago if any?
p.s. I have no idea what Im talking about.

Justin ODell.9517:

evilapprentice.6379:

Why was this a possibility at all?

I understand where you are coming from, it’s frustrating, it bothers me as a player and bothers me even more as a developer. The upcoming changes should significantly improve things, but either way we will continue working to improve the experience.

Brutal Arts.6307:

Does that refer to the following balance patch on the 15th of October or does this refer to some in house as of yet to be determined date?

Sorry for the confusion, I was not referring to the skills and balance patch, but a drop some time after that. I do not know if we have a strong date for that or not.

Krayiss.4926:

Could you explain the process of how hard it would be to server program a PTR and test MMR?

A public test realm is something on our radar, and many of us would like to see it, but it’s far beyond the scope of the SPvP team.

So far, I really like you. I don’t know that I’ve seen you post on here before today, but your responses are a welcomed addition.

felivear.1536:

Justin ODell.9517:

evilapprentice.6379:

Why was this a possibility at all?

I understand where you are coming from, it’s frustrating, it bothers me as a player and bothers me even more as a developer. The upcoming changes should significantly improve things, but either way we will continue working to improve the experience.

Brutal Arts.6307:

Does that refer to the following balance patch on the 15th of October or does this refer to some in house as of yet to be determined date?

Sorry for the confusion, I was not referring to the skills and balance patch, but a drop some time after that. I do not know if we have a strong date for that or not.

Krayiss.4926:

Could you explain the process of how hard it would be to server program a PTR and test MMR?

A public test realm is something on our radar, and many of us would like to see it, but it’s far beyond the scope of the SPvP team.

So far, I really like you. I don’t know that I’ve seen you post on here before today, but your responses are a welcomed addition.

I give ‘it’ a 50/50 so far lol we will see

It’s finally nice to have an explanation on this subject, and also that it’s being worked on.

Justin ODell.9517:

evilapprentice.6379:

Why was this a possibility at all?

I understand where you are coming from, it’s frustrating, it bothers me as a player and bothers me even more as a developer. The upcoming changes should significantly improve things, but either way we will continue working to improve the experience.

Brutal Arts.6307:

Does that refer to the following balance patch on the 15th of October or does this refer to some in house as of yet to be determined date?

Sorry for the confusion, I was not referring to the skills and balance patch, but a drop some time after that. I do not know if we have a strong date for that or not.

I appreciate the candor, but unfortunately “the future” isn’t a timeframe that sets my mind at ease. The world promised me flying cars and hoverboards by now, and we still haven’t seen them. As thrilled as I am to hear that the issues we care about are being worked on, a big part of the PvP communities gripe is the time frame in which PvP changes/upgrades are delivered. The fact that matchmaking is still dysfunctional is a point of contention, and as nice as it is to hear that it’s being worked on, the nebulous “in the future/when its ready” response has exhausted its good will and trustworthiness. This isn’t your fault, I’m just pointing it out.

Krayiss.4926:

Could you explain the process of how hard it would be to server program a PTR and test MMR?

A public test realm is something on our radar, and many of us would like to see it, but it’s far beyond the scope of the SPvP team.

It really shouldn’t be beyond the scope of the PvP team. I don’t mean to tell your higher ups how they should run things, but as far as I can tell, PvE doesn’t need a test realm – PvP does. PvP is constantly borked at patch time (maybe PvE is too? I don’t know, I don’t PvE), and PvP is sorely in need of more thorough QA and testing – that’s what a PTR is for. As I’ve said before, I appreciate and understand that Anet does not ask a monthly sub from me to play this game – if access to the PTR costs 5 bucks a month, I’ll gladly pay it, because it fulfills all my wildest fantasies as far as this game is concerned – it will help deliver more balanced, well tested patches (in a more timely manner, hopefully), and best of all, myself and other serious PvP players will be involved in the process – that is literally a dream come true, as far as I’m concerned.

I think Anet radar is so huge u can detect easily aliens or other stuff with it.

Justin don’t let us down!

I think the best solution would be to consider also skill variance other than average skill and to team up people to make them have average and variance of the skill level as close as possible. The only problem is the resource draining…

Having played several low population competitive games, here’s my usual suggestion for matchmaking.

The problem with on-demand matches is that at any given time players have no idea how many others are playing in matches already. This means if they experience anything negative – long waits, unbalanced matches – they will likely give up, reducing the player pool, and making things worse. It’s a snow ball effect.

For low population games, you need a way of pooling your players so that the match maker has the best chance of getting what it needs. Here’s one system:

Matches start at regular intervals. Eg. Every 15 min. This interval is dynamic based on current numbers of people playing. Time to next match is displayed prominently on the HUD if the player chooses to see it.
→ By having a known time, players can arrange their schedules a bit better (ie. Get a sandwich, surf the web, go to toilet etc). Having a set interval means there will be more players queuing at the same time, which means the matchmaker can put them all together in better games.
→ of course, there will be odd people out. (eg. 41 people are queuing, enough for 4 matches with 1 left out.) I suggest some sort of semi random system where if you miss out on 1 match, you are guaranteed to get in the next one. Otherwise it’s chosen at random. Perhaps dishonoured players have reduced priority too.

When match time comes, display a ready check message. Players that are ready are then placed in the matching pool. They are sorted by rating, and then placed into teams in an alternating way. (ie. Best player goes to red, 2nd best to blue, 3rd to red, etc) once a match is filled up, the next players start filling up the 2nd match, etc.

so if we have 40 players in this cycle, that means 4 matches will begin. The 10 best players will be in Match A, alternately divided into the two teams. The next 10 into Match B, etc.

The game should also announce how many matches have begun in each round. So if you find your self in a very mismatched fight, at least you know why when the game says only 1 match was created this cycle (ie. Less than 20 players were queuing)

Anyways, this system provides an effective and transparent way to do match making for a low population game.

Please also fix teaming up people with large difference in ladderranking.

I once asked a teammate what ladderposition he was at after winning a particularly low ranked(overall) match and he told me he checked it up and he was sitting around 62% (to my top100) at that point in time.

He might have lied but had no reason to do so, so i believe him.

Games really just should not start under these circumstances.

this could be a problem of base rating… what rating new player get? average? if yes, we all know where is the problem was the same in WoW when new teams got 1500 rating, they changed it to 0 and you should change it too…

I’ve done a chunk of solo queue tPvP – and I have been thrown into completely unbalanced experience matches quite a few times (along with the all to common 4v5s). No player swapping or drops – it’s just a team of 0-20 vs 30-50s. My only guess is it’s like matching the party, but when matching to the competing party, it is lax on matching or just doesn’t match at all. It is more common for me to see mis-matched ladder fights than to see a wide spread of levels in teams. As I am still low rank myself, I’m often on the team that is just learning PvP against these players that have done it a thousand times with nothing to do but try my hardest to end as top player in my team with a sub-50 score. But then I’ve also been placed on a team of 20-30s against 0-10s and end the match with 30 kills and the opposing team not breaking 100. The matching algorithm definitely needs work.

So what’s best? Be stricter on matchups but possibly cause much longer wait times? Create team spreads after 10 are like-levels are found? Regular intervals as Rieselle suggested?

Whatever the case, I admit I’ve avoided solo tPvP a bit lately due to these frustratingly unbalanced matchups – even fights are a blast. Being blasted or just blasting is not.

Side issue is how to handle drop-outs. 5vLess is often an imbalance that you can’t counter unless you’re team is better (in which case you usually don’t have a drop anyways), so it becomes a loss. Personally I think drop-outs should be pretty heavily punished (maybe a loss of glory?) with the short man team provided a vote if to forfeit or not. Within the first 60s, no win/loss is counted for either team. 60-120, win for the full team, no loss on the short team. After is standard count. Allowing quick joins like in sPvP could be an option, but it sucks joining a losing game so they usually just drop anyways and just becomes a revolving door.

Whatever the case, I hope to see better matchups in the future!

Justin ODell.9517:

We’re working on this right now. Hopefully it will be available in the next PvP drop.

I wrote a simulator to test several different ways of solving the problem, one of which is the way you suggested, so I’m glad we’re on the same page.

#mancrush

You sir, are a gem. Thank you for giving a direct answer and at least a hopeful patch timeline.

The simulator bit though… Perhaps someone should have thought of that before launching the original match making algorithm? A sample size of <100 would’ve shown you the fail. I suppose hindsight is 20/20, and at least you are addressing it now, so thank you.

personally i think people under a certain rank should not be able to do solo que. especially rank 1. your supposed to figure out your skills in hotjoin and than apply those skills in solo que.

Magische Boek.2530:

personally i think people under a certain rank should not be able to do solo que. especially rank 1. your supposed to figure out your skills in hotjoin and than apply those skills in solo que.

That was pretty much how it worked in Guild Wars 1.
Random Arenas was the lowest tier of competition. Once you get 5 consecutive wins in that gamemode, so either you reached an adequate skill level or get carried really hard (which is almost impossible on a 4-man gamemode), you gain access to Team Arena.

Then, to Team Arena, you had to get 5 consecutive wins there to have access to Heroes’ Ascent, so it is quite hard that you find players with no knowledge whatsoever of the game mechanics.

Also, in Guild Wars 1, rank actually meant something and each gamemode had a different rank.

People may say that I’m looking back with pick coloured glasses, but it’s really hard to believe that Guild Wars 2 has done better than Guild Wars 1 in those aspects of the game.