Guild Wars 2

I do love the idea of profession ratings.

I think the lesson here is the HIGH cost to player morale of the average player, of resetting MMRs. Top players just romp and stomp.

That new ranking is paid for with a LOT of unhappiness, for hours and hours.

Darnis.4056:

Im watching Noscoc (top5) roll hotjoin scrubs in Unranked with a full premade 500 to 10.

Sounds like MMR should lock them out of playing a pug of people who just bought guildwars2 and went into the mists, but no.

Just think how many people these top teams have turned off to unranked play in the past two hours. Five per match, maybe 6 matches per hour. Say, 60 more disappointed people.

How many in a day from other top rated people doing this?

GhOst, did you read my post?

I meant reward me with information about my play. Not a physical reward.

Darnis.4056:

Laserbolt.6731:

Mogar.9216:

:) do the math, every match there are 5 winners and 5 losers. However the top 10% players will win almost every match they are in . If there are 1000 matches played every night the top 10% of the players most likely take up a large chunk of those wins by themselves. This means an avg player will have a losing record. In fact I would think unless you are in the top 20% you will most likely be below 50% on the win ratio just because the best players takes up so much of the wins.

So you are saying that the top people group together. And they destroy groups of 5 average players, over and over and over all night…eventually generating a large pool of the average players who are now dispirited and unhappy, and whose rating has plummeted from mostly losing.

I’m watching it right now, noscoc(Team que 5-10) has a full team in unranked just 500 to 0’ing pugs of scrubs.

They don’t que ranked because there’s no leaderboard.

You shouldn’t be able to que unranked with more then 4…
And just farm terribles who just installed gw2.

That’s sad. I am watching it now, too. Ugh.

It would be interesting if the top-ranked people would Q solo for a week and record their win/loss records.

Mogar.9216:

:) do the math, every match there are 5 winners and 5 losers. However the top 10% players will win almost every match they are in . If there are 1000 matches played every night the top 10% of the players most likely take up a large chunk of those wins by themselves. This means an avg player will have a losing record. In fact I would think unless you are in the top 20% you will most likely be below 50% on the win ratio just because the best players takes up so much of the wins.

So you are saying that the top people group together. And they destroy groups of 5 average players, over and over and over all night…eventually generating a large pool of the average players who are now dispirited and unhappy, and whose rating has plummeted from mostly losing.

“It’s not whether you win or lose, it’s how well you played the game.”

…that’s the official design strategy posted by the ArenaNet Team on how they see GW2 sPvP. They want ranking to be raised when you play better than expected. It’s all built around their Expected Win calculation for every match.

And it’s a great goal! But it’s different from every other MMO PvP out there. And that’s a hard sell for people expecting to “WIN, or be a scrub”.

I’d encourage the sPvP system to let you know immediately, at the end of the match, whether it thinks you played better than expected and how it has just changed your ranking.

That would be ideal, given the philosophy of GW2 PvP competition.

GW2 is proudly not gear-based. It’s team-based, and it’s you-based. So it would be great to get that feedback about you right after the match ended and the events are fresh in your mind.

And sure, have a ladder and brackets. But if the reward is about your relative performance, we should get that reward immediately.

It’ll make us hungry to try again and queue again.

Thoughts?

GhOst.4019:

manveruppd.7601:

I have played with people who, due to their experience with other competitive games, were absolute beasts in pvp only 2 weeks after buying the game. I’ve also played with people who’ve been pvping since launch and they’re still dead weight. You never know. That rank 1 noob could be a future champion, but if you rage at them in chat, throw your toys out of the pram, and go afk, you’ll just drive them away from the game (or they’ll put you on their blocklist and never want to team up with you).

Absolutely. Each player is different. I didn’t say I wanted newbs to bend over for me. They just need a little direction and not be yelled at. Have I yelled at players? Absolutely….and so has everyone else. When tying to help the unhelpable and they continue to free lance the roads…Get’s your blood pressure up. Most of the time they apologize for their lack of knowledge and thank you for the help and pointers. It will be a frustrating start to all this but a surrender option should never be there. You don’t get better at PvP by steam rolling newbs…you get better by getting steam rolled. You learn what went wrong, how you can counter it and move forward working at it until you get better. Yes, some players will never get better but that’s just life.

Sorry. I edited my last comment. I can see you aren’t asking for continued submission by weak players, just better players. But some really do want to “own noobs” for as long as possible, and I was directing my comments at them.

I am enjoying my power necro. I realize that its liability is no heals and low toughness and not a huge health pool. No stealth or escape mechanism. When you are in, you are in to the death it seems.

Seems high reward but high risk of death. No?

Then again, I just tested Dagger vs axe, and the dagger does kill faster, even with just the #1…and it hits 2 people as well.

Thanks for the replies so far.

Axe and well of suffering = huge damage from a power necro. No? And at 600 distance at that.

They are still adjusting the matchmaking to get relatively even matches and accurate match predictions so they can know how well a team performed relative to the toughness of the match for them.

Once that happens they can reward better-than-expected performances by raising your ladder ranking. And eventually, brackets so you can play with people at your own current level of skill.

I think they said the old leaderboards are history.

GhOst, I think it can be un-sportsmanlike to expect a very weak team to line up and bend over for you. In boxing, they stop matches when one guy is clearly beaten.

What some seem to want is for the weak, unskilled players to run up and spam their 1 button, or whatever they do, for a little longer and let you leap and cleave them 100 to 0 in 4 secs. Is that fun?

Now, a surrender system COULD be abused by people who bail whenever they are losing. Therefore, it should not be an option until the winning team is more than 250 points ahead, after 5 minutes or something reasonable.

Saiyan,

As I said, no user names or account information. It’s irrelevant anyway.

And we would not need to see all the gigabytes of stuff you mentioned.

I was talking about a data extract of match results and some factors about the players in the match.

No need to worry.

Saiyan.1704:

Laserbolt.6731:

The majority of people I see on the forums say they are being crushed regularly.

How can the majority be getting stomped? There have to be an equal amount of stompers.

I have always been sort of “in the middle” as a player. But in this new testing phase of the matcher, I have lost 70% of my matches.

SOMEBODY out there is winning most of their matches.

I guess they don’t talk about it because they feel great.

And the majority of the players don’t care enough to use the forums. I have no clue what my win % is but I’m actually winning quite a few matches queing solo.

You can find the stats on your Heart of The Mists page for ranked, unranked, by profession etc.

Crowd sourcing can often reap great insights.

If, for example, we could download an extract of the data for all the matches since the patch (without names, of course), perhaps those of us with analytical and software/database skills could find something in there that would be helpful in understanding what’s happening.

It would be a groundbreaking move to export the data in a CSV file or something.

The majority of people I see on the forums say they are being crushed regularly.

How can the majority be getting stomped? There have to be an equal amount of stompers.

I have always been sort of “in the middle” as a player. But in this new testing phase of the matcher, I have lost 70% of my matches.

SOMEBODY out there is winning most of their matches.

I guess they don’t talk about it because they feel great.

I hear people laughing at MMs as non-serious builds.

I really want to play a MM well and be effective against good players. Is that a reasonable expectation? Or is it unrealistic at the current time?

Any tips to playing a MM extremely well? Key issues?

What builds are my hard counters that I should not feel too bad losing a 1v1 to?

Who SHOULD I easily take on?

Thanks.

Thanks.

You said: “I have dual priorities, that do not conflict. 1) Improve match quality, 2) Ensure prediction is accurate.”

Isn’t #2 actually #1 as well?

Can you even have good match quality without an accurate prediction?

I assume you won’t be turning on the ladders until the prediction is very accurate, so you can award ladder points accurately.

But your MMR is still adjusted as a loss…even though you may have played well. But the person abandoned you and you lost.

Justin,

If the MMRs are not truly a measure of a player’s actual skill, then you would see blowouts, and the expected win prediction will usually be wrong.

I take it you guys are primarily turning the dials trying to FIRST make the Expected Win prediction truly accurate? Am I right?

Isn’t an accurate prediction the best proof that the MMRs are accurate?

butch,

Thanks.

That Wreckin Crew
Jade Quarry
Playstyle- sPvP
TIme Central US time

Narkodx.1472:

Indeed completely agree, all other things are conveniences and/or secondary to the most important aspect of matchmaking – MMR or Skill of the players

Which cannot be easily measured from their TEAM’S performance unless you look at hundreds of games. Many random factors. Your team members can afk…that tanks “your” MMR. You can be placed on a team where your build is a liability…again, not a measure of your skill, but your MMR is adjusted nonetheless.

I strongly suspect that the current MMR does not accurately measure skill.

That’s why matches are blowouts. The match was never even, but the faulty MMRs “added up” and the matcher algorithm said the teams were evenly matched.

But they were not. Hence, a blowout.

The key to proving this is to look at the “expected win” number for all the matches and see how many teams with a “50% chance” were actually shut out earning just 0 to 100 points.

Look…your individual MMR is derived from the GAMES you are in, from your entire team’s total performance against a specific, unique team…it’s not your particular contribution.

The theory is that over time, it will capture your actual skill. I kinda doubt it.

Or if it does, it would have to be 500 to 900 games to do it, since there are SO MANY random factors about who you are teamed with, and against.

Does unranked have a separate MMR? And is it by profession?

You’ll have to read both the ArenaNet posting about the new PvP, and search for developer posts in this forum…Many of these have been answered already

Backpack.8517:

My team qs lately… 500-0, 500-6, 500-28… cmon reseting MMR?? What is your reasoning behind doing so.

Yes, the MMR takes quite a long time to have enough games to be accurate for a given player. It is derived not from the player’s performance, but must be deduced from random TEAM performances the player is 1/5th of. That’s a very tough challenge that I am sure takes a 100 games or more to become accurate. Why start over?

Are you guys validating the “expected win algorithm” against actual games, then?

If there is not a very strong correlation, i.e. “80% of the predictions come true”, then there is a real problem.

Right?

I applaud the PvP developers for really trying hard to create a good matchup for everybody.

You can only measure what is truly accurately measurable.

I think this is a really tough problem because different factors may not combine in a smooth, linear fashion. There may be step functions and leaps when some things come together. 1 plus 1 = 4. Or 10!

A good pattern matching algorithm sifting the game data might be able to tease out some correlations that are not readily apparent. They may be non-intuitive once you see the correlations.

It may be that ANY group vs 5 solos has a much bigger advantage than we suspect. If so, then the expected win calculator MUST know that or the ladder rankings will be all “off”. Perhaps they may be essentially random if you are not careful.

The “expected win algorithm” is everything. It determines the rewards for ranking and it MUST be accurate in assessing a team.

Aereniel.7356:

Pandatrice.7183:

We also supposed to kill the two monsters at the very beginning, do we?

Killing Svanir and Chieftain at the start is the right call only very seldom, and even then it heavily depends on the enemy team’s composition. If the enemy team has highly mobile classes that can spike, such as (competent) thieves, the risk of a steal is very great.

In evenly-matched teamQ Forest games the mobs are often left alone completely. In soloQ however you can pull bosses with more security, given that people don’t coordinate and next to nobody realizes that 90% of the game is actually played on the mini map.

Please elaborate how the game is played on the minimap. You mean you watch for patterns and figure out the enemy team’s strategy and where you now need to be?

Nitesky.6427:

Patch has been deployed 2 days ago and things are not working out according to plan. This is indeed very scandalous, because in software engineering, everything works perfectly from the very first second something new gets released. I mean what are they thinking? It’s their job to work around the clock to appease spoiled teenagers sitting in their basement playing games all day.

Sorry for the antagonistic comment, but in our times it seems, people really need to relearn empathy and patience.

+10

Exactly right. For a non-subscription game at that.

Dunning-Kruger here, big time.

Remember, ladder rank is going to be awarded based on the total points your team scores compared with your expected chance of winning. (See https://www.guildwars2.com/en/news/finding-the-perfect-match/)

Here is the chart ArenaNet provided:

I suspect that: players who feel unreasonably entitled to have everything given for free still live with, or are supported by, their parents.

Seriously. Think about it. Look at yourself and ask if you expect too much as you criticize the developers and call them incompetent.

+10 I love the new UI.

No subscription, Robert. No subscription fee. You pay nothing per month. And you still get new content constantly.

You paid $60 bucks and you got up to two YEARS of the game so far, and there is no subscription fee.

Yet some of you have the entitlement attitude to complain about the PvP upgrade.

For which you have paid exactly ZERO dollars.

Just keep that in mind. You’ll feel better.

Mjgander.1905:

Yes, I do also think it’s unreasonable.

They do not know the proper settings yet. They have to see how the complex settings work on real matches. That is just the way it is with complicated software.

MjGander,

You don’t seem to understand what is happening right now and why it is happening.

I will explain it for you.

1) They are testing out the new matchmaking system
2) They said they are adjusting it and uneven, silly matches WILL occur in this testing phase.
3) You seem to think this is unreasonable and these matches should not be happening now.

So, you are complaining constantly in the forums. You think it is unreasonable that bad matches happen during the testing and adjustment period.

But the rest of us understand that they are working the bugs out. We get it.

If you play now you are going to get facerolled, maybe a lot. So stop playing if that bothers you.

WHY are you still playing?

MJgander, you can stop posting now. We get it. You want it to work correctly right now, and not in two weeks.

Mj, given how you feel, I would avoid playing until they get it adjusted. You’ll be happier.

I like the new system. A lot.

When the match is ready, you must GO RIGHT NOW or suffer dishonor. No time for dueling or doing anything you cannot abandon in 2 seconds. That’s just the way it is now.

I was wrong. Your MMR will be adjusted from Unranked play , but your ladder rank will never change from unranked play.

Here is what they said in the recent Ready Up:

“Unranked Arena will give players a safe place to play solo or with their friends in any sized group without having to worry about affecting their ladder standing. If you’re ready to move on from hot-join, or if you want to test out new builds, warm up a new profession, try out new teammates, or just chill out and have fun in a structured PvP environment, Unranked Arena is for you.

“Unranked Arena also takes advantage of all the new matchmaking changes, so you will tend to be grouped together with players of your skill level and group size.”

https://www.guildwars2.com/en/news/announcing-improvements-to-pvp-and-the-ladder-test-season/

MjGander, no use complaining on this until after the 14-days adjustment period.

They can see all the games and scores on what maps much better than you can. They are aware.

Since you won’t get many ladder rank points for winning on a match that your team is EXPECTED to win…

And you actually can gain a lot of ladder rank points for decisively winning a game your team was expected to LOSE…

It would be MOTIVATING if the final score showed which team was EXPECTED to win, and by how much.

If your team LOST and had been expected to lose…but you lost by LESS than expected you will be thrilled to know your rating is going to go up some. The sting of the loss is reduced, and next time you will play hard even if your team is losing.

Winning teams who coasted at the end and won by a smaller than expected margin will know they are not going to be getting the full rating increase, and THEY will try harder next time as well.

That window covers the screen for a reason. It does not want you do get started with something that will cause you to bail the match.

Personally, I think PvP is “Under New Management”. And I like it.