13 replies to this topic

[Skyraxx]

Hey I was wondering what you guys think about tweaking ELO? Right now from what I understand it looks at overall team ELO which I think is a bad way of doing things. I really think it needs to be looking at an average for players within a team and trying to match based off that.

It just seems like too often an occurrence where when I'm pugging my team gets ROFLstomped. My guess is some lower elo players get placed with me to average out the team ELO and end up getting screwed because the other team is comprised of higher average players.

I also think that maybe PGI should give extra weight to people who play in certain units. I cringe when I get into a pug match and there's maybe 3 people who play in units on my team and the other team is full of players with units that play competitively. It's simply not fair to the people who play only casually to be stuck playing against a large group of people who play in organized 12 man games regularly.

Edited by Skyraxx, 19 November 2014 - 04:56 PM.

[Sandpit]

Skyraxx, on 19 November 2014 - 04:54 PM, said:

Hey I was wondering what you guys think about tweaking ELO? Right now from what I understand it looks at overall team ELO which I think is a bad way of doing things. I really think it needs to be looking at an average for players within a team and trying to match based off that.

It just seems like too often an occurrence where when I'm pugging my team gets ROFLstomped. My guess is some lower elo players get placed with me to average out the team ELO and end up getting screwed because the other team is comprised of higher average players.

I also think that maybe PGI should give extra weight to people who play in certain units. I cringe when I get into a pug match and there's maybe 3 people who play in units on my team and the other team is full of players with units that play competitively. It's simply not fair to the people who play only casually to be stuck playing against a large group of people who play in organized 12 man games regularly.

Team Elo is the average of all members and then there's a multiplier that inflates it.

"Casual" group queue has been discussed to death. The main reason it's just not feasible is player population. The MM simply cannot handle a continuous split of the queues.

[Davers]

Skyraxx, on 19 November 2014 - 04:54 PM, said:

Hey I was wondering what you guys think about tweaking ELO? Right now from what I understand it looks at overall team ELO which I think is a bad way of doing things. I really think it needs to be looking at an average for players within a team and trying to match based off that.

It just seems like too often an occurrence where when I'm pugging my team gets ROFLstomped. My guess is some lower elo players get placed with me to average out the team ELO and end up getting screwed because the other team is comprised of higher average players.

I also think that maybe PGI should give extra weight to people who play in certain units. I cringe when I get into a pug match and there's maybe 3 people who play in units on my team and the other team is full of players with units that play competitively. It's simply not fair to the people who play only casually to be stuck playing against a large group of people who play in organized 12 man games regularly.

It doesn't create 2 teams with a similar average Elo. It creates 2 teams populated with players of as near Elo as it can. Everyone thinks they are 'not the best, but a good player'. Sorry, but most players are average players. And average players still make lots of basic mistakes that can cost them the match. Players who don't make those mistakes are 'good'. They have superior map awareness and are amazingly accurate, even at long range. If you see a lot of effective LRM boats in your games, guess what? You are not above average.

[MadPanda]

I was following you until you started talking about people in units being better players, ROFLMAOLOL!

[Monkey Lover]

Seeing Cw is around the corner i dont think they're going to do anything.

[Wintersdark]

MadPanda, on 19 November 2014 - 06:37 PM, said:

I was following you until you started talking about people in units being better players, ROFLMAOLOL!

This.

Throughout even stronger units, you've got great players, good players, and average players. What strong units have that makes them strong is teamwork, much more than player skill. A run of the mill average player who can do what he's told and follow directions is, in a unit v. unit game, much more valuable than a very skilled player who does not.

Once you get into the solo queue, though, where a lot of those order following and formation-play skills don't work as well? Then the less skilled players really suffer. Solo queue play is fundamentally different from unit play in the group queue - teamwork still wins matches every time, but it's much more difficult to manage.

So, no, players from various units should not have an Elo modifier. Elo will do it's thing with them as it is - if they play grouped primarily, they'll end up with a high Elo anyways (if anything, an unfairly high Elo when they get to the solo queue).

What happens though is that the few (and they are few - most players, as Mr. Panda says, are purely average) good players stump the matchmaker, as there simply are often not adequate players to match them against. So, they end up in lower-ranked games.

This is not a correctable problem. If there are insufficient players of a given skill level available when the match is being made, then that is all there is to it.

Edited by Wintersdark, 19 November 2014 - 07:43 PM.

[Kain Demos]

I wish it would just match people of similar skill against each other rather than put 2 or 3 decent players on a team full of tards and hoping it averages itself out.

[Bhael Fire]

The public group queue will level out once the CW queue is in place, because many of the hardcore 12-mans and other large competitive groups will mainly stay there.

As for getting stomped in the solo queue, it's possible to get stuck in what is known as "Elo Hell" if you get an unusual streak of lucky wins against teams that you were predicted to lose against. This can artificially pump up your Elo score making the Matchmaker think it needs to "challenge" you by putting you up against even harder teams, often times erroneously thinking that you are good enough to carry lower Elo players to victory against these foes.

This in turn, creates a series of really nasty ROFLSTOMPS that can last for days...until the Matchmaker realizes that you "ain't all that" and puts you back with similar Elo players again.

[Skyraxx]

Someone who plays in a unit is going to have a lot more practice and better understanding of strategy and builds in this game then someone who drops solo all the time on average. You stack one side with people who are all in units and another who are just solo droppers, the side with tags are going to win.

[Wintersdark]

Skyraxx, on 22 November 2014 - 08:34 AM, said:

Someone who plays in a unit is going to have a lot more practice and better understanding of strategy and builds in this game then someone who drops solo all the time on average. You stack one side with people who are all in units and another who are just solo droppers, the side with tags are going to win.

No.

Strategy and builds for unit play are different from strategy and builds solo, to begin with. Group play and solo play are very, very different animals.

And it's definitely been my experience over the years that an equal percentage of unit-based players are totally ignorant as to how the game works, even among successful units. They work out fine because they do what their told which is what's critically important for unit success. Dropping in the solo queue, though, they don't have that leadership, don't understand solo queue play, and have no idea how to make a good build for it. Even if they're using a decent build someone gave them, their personal lack of skill is a huge disadvantage for them there.

Primarily group players and primarily solo players are no different. There are good players and bad players, in equal measure, in both cases.


Now, with that said. If you get into a match in the solo queue with a lot of players on one side in the same unit due to fortuitous (intentional or not) sync dropping, then yeah, their side has a noteworthy advantage. But simple presence of a unit tag does not even remotely imply any amount of skill.

[Skyraxx]

Yeah, I agree strategy and builds for unit play can be different. But people who play in units are going to have more experience and a better idea of how to craft a mech. Not everyone is going to be playing a meta mech when they solo drop but that doesn't mean they haven't taken what they've learned and are going to apply it to the solo queue.

Not to mention people in units worry about their reputation and are less likely to do something stupid to get themselves killed and make their whole team hate them.

[nehebkau]

ELO needs to account for drop weight.... two players with the same ELO, 1 in a 30t mech and the other in 100T mech CAN NOT BE EXPECTED TO CARRY THE SAME AMOUNT!

They already set ELO based on mech weight, but do not account for drop weight which should reduce ones 'EFFECTIVE' ELO.

Edited by nehebkau, 22 November 2014 - 09:56 AM.

[Elizander]

I would prefer they set Elo per mech. Like in Street Fighter IV you are ranked by character and not your account as a whole. Taking out my Dual Gauss Dual ERLL 6xML Master Direwolf with all the modules is not the same as me picking up a fresh Stalker or a Trial Victor.

The game will say "Oh this guy won a bunch so let's get some lower Elo folk to balance out the team" but you won't perform as the game expects because you are bringing a mech that isn't maxed out or a mech you are not as good at in playing.

I think the game is probably expecting a high Elo Light or Medium pilot to carry as hard as a high Elo heavy or assault pilot. It is also troublesome for support mechs since the game might not give you the proper players to support. This is the problem with mixing high and low. Sometimes you get a team that knows what they are doing and sometimes you have someone doing their first 25 matches.

[FupDup]

It would be nice if a team's elo used the median instead of the average, because the former is less vulnerable to outliers than the latter...