You need to learn not to WOT, no one reads that ****. Your simpleton understanding of matchmaking also does not account for premades, or the fact that whatever values they use to assess "skill" can be manipulated easily or simply not actually be a good measure of skill, or the fact that between the 15 and 16th most skilled person could be a gap as wide as the distance from here and reality.

While i personally feel that this matchmaking isnt going to accomplish its objective i also havent been playing pts and thus really have no way of determining whether thats true or not.

This will be a bit of a nerf for premades in regs as its very unlikely any number of people greater than 4 will he together.

And that is exactly why we do not know our "rating" in regs. If we did - we would manipulate it.

I believe you and rafi are arguing philosophy at this point. He seems to feel any matchmaking is better than none and you seem to feel it doesn't matter either way.

You need to learn not to WOT, no one reads that ****. Your simpleton understanding of matchmaking also does not account for premades, or the fact that whatever values they use to assess "skill" can be manipulated easily or simply not actually be a good measure of skill, or the fact that between the 15 and 16th most skilled person could be a gap as wide as the distance from here and reality.

The matchmaking actually does account for premades (theoretically anyway). If you are in a premade, everyone's "rating" is equal to the highest person in the group's rating, so it is a slight penalty for being in a premade. So chances are pretty good that if there is 1 premade in queue, the top 4 solo queuers will most likely all be on the other team (it might actually be the top 8 depending on what everyone's rating is).

And if the skill between the 15th and 16th is a huge difference, than the teams could end up being 1-7 and 16 vs 8 to 15. I doubt it'd get that extreme, but 16 would have to be really really bad to make the team lose that match anyway.

While i personally feel that this matchmaking isnt going to accomplish its objective i also havent been playing pts and thus really have no way of determining whether thats true or not.

 

This will be a bit of a nerf for premades in regs as its very unlikely any number of people greater than 4 will he together.

 

And that is exactly why we do not know our "rating" in regs. If we did - we would manipulate it.

 

I believe you and rafi are arguing philosophy at this point. He seems to feel any matchmaking is better than none and you seem to feel it doesn't matter either way.

Having followed the PTS forum (though not installed it, I don't have enough play time to make it worth it nowadays) it seems like the matches themselves tend to be fair, there are just some complaints about still too many healers getting onto teams. But the matches themselves seem to be relatively evenly matched.

Having followed the PTS forum (though not installed it, I don't have enough play time to make it worth it nowadays) it seems like the matches themselves tend to be fair, there are just some complaints about still too many healers getting onto teams. But the matches themselves seem to be relatively evenly matched.

You are correct. When the healers have been limited, the matches are pretty balanced and have been some of the best I’ve had in 2 years.

The biggest issue at the moment is the healer cap is broken and too many people are queueing as healers. It makes it very hard to see if matchmaking has any problems we haven’t uncovered yet.

You need to learn not to WOT, no one reads that ****. Your simpleton understanding of matchmaking also does not account for premades, or the fact that whatever values they use to assess "skill" can be manipulated easily or simply not actually be a good measure of skill, or the fact that between the 15 and 16th most skilled person could be a gap as wide as the distance from here and reality.

I was simplifying because so far you seemed to fail to understand anything more complicated I say, so of course I didn't get into details. There you go:

Premades - Either everyone's rating is "boosted" to the rating of the top of the premade (which means they get the worse pugs) or for each member of the premade, the rating is increased by 10% for each player, which also means the premade eventually get greater noobs, but this time a premade of 3 idiots and one awesome player will not be over-handicapped (the reasoning behind the rating increase is "coordination bonus". It doesn't have to be exactly 10, of course]. I said the 1st option because I believe Eric said this is how it actually works, I personally favor the 2nd one much more...

"Skill manipulation" - If they release the exact way they compute the values, sure. Anyway, a win-loss counter can't be cheated without losing, and if you gotta lose in several games in order to get the advantage in other several games (only to win there and lose this advantage with just as many wins as losses), you might as well actually try to win in the first game. Good luck cheating that without hacking/trolling/afking, and if we are discussing those who hack/troll/afk, they are indeed a big problem, but not one related to matchmaking.

A win-loss counter is also quite the accurate assessment of your probability to win or lose, If you have +100 (which means you won 100 more games than you ever lost since the counter has begun counting) and another guy has -100 (which means he lost 100 more games than he ever won since the counter has begun counting), and both of you get the in the same team due to the matchmaking, then either you are good enough to carry him, you win, and he breaks his loss streak, or he is bad enough to make you lose, and he broke your streak.

Of course eventually, if you are a good player, you would rotate between +100 and +50, because at some point the bads in your group will not be bad enough to keep you from winning (some other good player will probably snatch some bads to the opposing side), and same goes for truly bad players. They will probably be around -50 and -100 because as they reach -50 they will not get good enough gods to carry them. So it is not like skill won't matter.

Every statistic has anomalies and peak points, of course, and some seriously good or bad player would be able to not be entirely "matchable", but it will mostly converge into something sensible, and mostly, it would be far more rare to see stomps as frequently as we can see them now.

P.S If you don't like to read math, you are welcome not to read it, but unless you fully read and understand it, don't dismiss it...

I was simplifying because so far you seemed to fail to understand anything more complicated I say, so of course I didn't get into details. There you go:

 

Premades - Either everyone's rating is "boosted" to the rating of the top of the premade (which means they get the worse pugs) or for each member of the premade, the rating is increased by 10% for each player, which also means the premade eventually get greater noobs, but this time a premade of 3 idiots and one awesome player will not be over-handicapped (the reasoning behind the rating increase is "coordination bonus". It doesn't have to be exactly 10, of course]. I said the 1st option because I believe Eric said this is how it actually works, I personally favor the 2nd one much more...

 

"Skill manipulation" - If they release the exact way they compute the values, sure. Anyway, a win-loss counter can't be cheated without losing, and if you gotta lose in several games in order to get the advantage in other several games (only to win there and lose this advantage with just as many wins as losses), you might as well actually try to win in the first game. Good luck cheating that without hacking/trolling/afking, and if we are discussing those who hack/troll/afk, they are indeed a big problem, but not one related to matchmaking.

A win-loss counter is also quite the accurate assessment of your probability to win or lose, If you have +100 (which means you won 100 more games than you ever lost since the counter has begun counting) and another guy has -100 (which means he lost 100 more games than he ever won since the counter has begun counting), and both of you get the in the same team due to the matchmaking, then either you are good enough to carry him, you win, and he breaks his loss streak, or he is bad enough to make you lose, and he broke your streak.

Of course eventually, if you are a good player, you would rotate between +100 and +50, because at some point the bads in your group will not be bad enough to keep you from winning (some other good player will probably snatch some bads to the opposing side), and same goes for truly bad players. They will probably be around -50 and -100 because as they reach -50 they will not get good enough gods to carry them. So it is not like skill won't matter.

 

Every statistic has anomalies and peak points, of course, and some seriously good or bad player would be able to not be entirely "matchable", but it will mostly converge into something sensible, and mostly, it would be far more rare to see stomps as frequently as we can see them now.

 

P.S If you don't like to read math, you are welcome not to read it, but unless you fully read and understand it, don't dismiss it...

No, wrong. You're still missing a massive flaw in your logic here.

No, wrong. You're still missing a massive flaw in your logic here.

Would you kindly enlighten us all what exactly would this logical flaw be, or are you still looking for it yourself (and you just "know" it must be somewhere there because you "know" you are right)?

You are correct. When the healers have been limited, the matches are pretty balanced and have been some of the best I’ve had in 2 years.

The biggest issue at the moment is the healer cap is broken and too many people are queueing as healers. It makes it very hard to see if matchmaking has any problems we haven’t uncovered yet.

We had the exact opposite today, for the "rishi arena test" session. I don't think I saw a single healer at all. The games still felt mostly fair to me. But I've started to come around to the opinion that it has more to do with the small pool of people on the PTS all being at about the same skill level. We don't really have any "newbs" on the PTS, and I don't think we got much for real superstars either (sorry to anyone who is a super star that showed up ).

Also, hardly anyone gives up on a match, and no one rants much when a loss happens. Which helps the overall attitude to be much more pleasant. I think if people are still trying, and no one is ranting and rage quitting, then the matches feel more fair even if you don't really have a chance.

We'll see, I guess, if the matchmaking is the holy grail, but I suspect we'll still have plenty of people that think they can see the loss coming, and will blame it on any and everything except themselves. And then the quitter-slippery-slope will happen just like it always has. I hope I'm wrong. But every time I dip back into the live servers lately I seem to be reminded just how delusional people are about their own skill, and usefulness to the team.

We'll see, I guess, if the matchmaking is the holy grail, but I suspect we'll still have plenty of people that think they can see the loss coming, and will blame it on any and everything except themselves. And then the quitter-slippery-slope will happen just like it always has. I hope I'm wrong. But every time I dip back into the live servers lately I seem to be reminded just how delusional people are about their own skill, and usefulness to the team.

Who is this typing on Banderal's account? The eternally optimistic Banderal, this surely isn't him.

What I am reminded every day when I enter WZs is how lopsided teams can be constructed. It's utterly ridiculous sometimes. To use your words, all it takes is 2-3 "super stars" versus a really weak side and the match completely folds inside out.

I notice some well-geared people quit matches before they start when the team has a lot of 230 or less geared players. It's no guarantee they are bads, but I can see why as often times these lesser geared players not only wear crap gears but they play poorly too.

I don't know what matchmaking will accomplish if much at all if it only goes off of class specs. If they somehow could manage a matchmaking system from some sort of ELO then it might be even better than class spec matchmaking only because skill level seems to trump all.

Didn't Eric or someone mention some sort of hidden ELO rating all players have that they could use to possibly match up players according to skill? I don't know, maybe the population aint big enough for that anyway.

I guess I am just frustrated as of lately.

Who is this typing on Banderal's account? The eternally optimistic Banderal, this surely isn't him.

haha. Sorry, I guess one too many matches where I'm solo guarding west, and the ENTIRE rest of my team is standing in front of south while it gets taken, and then someone starts up in chat with how worthless everyone is because *THEY* let the node get taken. I just sat quietly, because why feed the flames, but I'm thinking pretty loudly at that point... "***? YOU were right there too".

I get that people tunnel, make mistakes, etc. But own up to it. Don't start blaming everyone else when you right there making the same mistake. I've been one of those 7 when the node gets taken. At that point I figure it's 1/7th my fault, and 7/7th my responsibility to get it back.

Ok, now back to your regularly scheduled program.

Didn't Eric or someone mention some sort of hidden ELO rating all players have that they could use to possibly match up players according to skill? I don't know, maybe the population aint big enough for that anyway.

 

I guess I am just frustrated as of lately.

Yes, match making is supposed to take a hidden "skill" factor into account. I don't think we've gotten any actual info, at all, about how that factor is computed. If it's by player, by toon, by legacy or what. He also said they were going to take class/spec into account beyond just healer/tank/dps. And also gear. At least that's the way I read it. Eric's 2nd post in this very thread...

Actually, since SWTOR launched players have always had a "rating" behind the scenes. I am hesitant to say rating since this isn't the same thing as your Ranked score, but it's the easiest way to explain it. This is more of a way that we approximate player skill solely for the purposes of matchmaking. This is also the reason we don't expose this information, since it is not really a rating (like Elo, etc).

 

In 5.9.2 we are making optimizations in how we use this information along with things like role, spec, gear, and more to try to create the most balanced teams we can.

 

-eric

Who is this typing on Banderal's account? The eternally optimistic Banderal, this surely isn't him.

 

What I am reminded every day when I enter WZs is how lopsided teams can be constructed. It's utterly ridiculous sometimes. To use your words, all it takes is 2-3 "super stars" versus a really weak side and the match completely folds inside out.

 

I notice some well-geared people quit matches before they start when the team has a lot of 230 or less geared players. It's no guarantee they are bads, but I can see why as often times these lesser geared players not only wear crap gears but they play poorly too.

 

I don't know what matchmaking will accomplish if much at all if it only goes off of class specs. If they somehow could manage a matchmaking system from some sort of ELO then it might be even better than class spec matchmaking only because skill level seems to trump all.

 

Didn't Eric or someone mention some sort of hidden ELO rating all players have that they could use to possibly match up players according to skill? I don't know, maybe the population aint big enough for that anyway.

 

I guess I am just frustrated as of lately.

They are combining all that together to make sure the teams are even:

Actually, since SWTOR launched players have always had a "rating" behind the scenes. I am hesitant to say rating since this isn't the same thing as your Ranked score, but it's the easiest way to explain it. This is more of a way that we approximate player skill solely for the purposes of matchmaking. This is also the reason we don't expose this information, since it is not really a rating (like Elo, etc).

 

In 5.9.2 we are making optimizations in how we use this information along with things like role, spec, gear, and more to try to create the most balanced teams we can.

 

-eric