31 replies to this topic

[Ryvucz]

The match making system is why we can't have fun tournaments.



Matches end up being insanely one sided.

I'd rather play versus an 8 man pre-made with the weight classes even, versus this silly magical mystery tour.

I understand there are going to be tweaks, but ELO is horrid for this kind of game, with 7 other people involved on your team, especially in a tournament setting.

I can appreciate the work done on the match maker thus far, but balanced weight class was the best thing for matches, in my opinion. (Especially Conquest)

[FrostCollar]

Sorry, I'm a bit distracted - but what are we supposed to be seeing in the picture?

[Ryvucz]

FrostCollar, on 27 April 2013 - 02:41 PM, said:

Sorry, I'm a bit distracted - but what are we supposed to be seeing in the picture?

Three lights on one side, one light on the other.

Conquest mode.

[Wintersdark]

The team with one light has two centurions. They're fast enough for conquest capping and strong enough that they should have no trouble with Jenners.

This is not a horribly lopsided battle, it was just poorly played.

[Ryvucz]

Wintersdark, on 27 April 2013 - 04:00 PM, said:

The team with one light has two centurions. They're fast enough for conquest capping and strong enough that they should have no trouble with Jenners.

This is not a horribly lopsided battle, it was just poorly played.

Given the terrain, I think you're horribly mistaken.

[Wintersdark]

On Frozen City Night? Why? It's a small map, Centurions can get around plenty fast enough. Hell, on FCN, they were fine even with slower mechs.

You lost because you were outplayed. Look at the damage done, the casualties. Team Blue lost, but certainly not because they didn't have enough lights. Team Red outright crushed you.

Even given that, it looks like it was pretty close to me, particularly given how you guys only had one person doing any real damage.

[Ryvucz]

Wintersdark, on 27 April 2013 - 05:12 PM, said:

On Frozen City Night? Why? It's a small map, Centurions can get around plenty fast enough. Hell, on FCN, they were fine even with slower mechs.

You lost because you were outplayed. Look at the damage done, the casualties. Team Blue lost, but certainly not because they didn't have enough lights. Team Red outright crushed you.

Even given that, it looks like it was pretty close to me, particularly given how you guys only had one person doing any real damage.

Alpine Peaks is the only real map that doesn't have terrain that can be drastically avoided by jumpjets at the moment.

I'm talking about the sheer cliff edges that any mech with jumpjets can clear for faster point cap.

This is not about damage done, this is about capping points with speed.

I could care less who did the most damage, what irks me is 3 lights with jumpjets on a map that maneuverability in a conquest mode plays a very strategic advantage.

[CheeseThief]

The real reason the OP lost was because his team had more Fedrats than the other. ELO hates that little golden sunburst with a passion and seems to go out of the way to make the team with more of them lose.

One fedrat is fine, individually they can do really well but the more you get in the same place the more likely you're going to lose. It maybe a close loss, or it maybe a complete thrashing, but your still going to lose.

[Ryvucz]

CheeseThief, on 27 April 2013 - 08:15 PM, said:

The real reason the OP lost was because his team had more Fedrats than the other. ELO hates that little golden sunburst with a passion and seems to go out of the way to make the team with more of them lose.

One fedrat is fine, individually they can do really well but the more you get in the same place the more likely you're going to lose. It maybe a close loss, or it maybe a complete thrashing, but your still going to lose.

You can't be serious. XD

[and zero]

Im just gonna quote myself here because I am sick of wasting time compiling feedback that pgi doesnt care about. I agree with ryvucz.

and zero, on 26 April 2013 - 03:18 PM, said:

This is absolutely false. Especially so since elo was implemented (poorly) and many of us now frequently get screwed with utterly incompetent teammates way below our skill level for some ungodly reason. If this was call of duty, your damn right I could and would carry my team. But this is not. If I do 900 damage with 5 kills, one other person does 500 damage, and everyone else does less than 100 before committing suicide with no other kills. We lose. And there is not anything I couldve done about it. Since elo, this happens to me a majority of games. I am not sure why, but it is ****.

Further, regardless of this specific factor, the tournament is still not really a measure of individual skill. It's more like 30% skill and 70% luck. Just like all the past ones. Because this game is so heavily team based and the matchmaking is still such crap (both elo/player skill levels and tonnage balance still being largely MIA). Even if they took out the win loss factor, it is hard to get more than one or two kills/assists when 5 people on your team do less than 60 damage before committing suicide one by one.

and zero, on 27 April 2013 - 10:39 PM, said:

Yea, this is sadly still true About 65-75% of games at least 4 of my team is pathetically bad. They run in and commit suicide doing 30 damage. They split up and run in random directions for no reason. I say "please stay together" they say "fack u nooob"...and then die 30 seconds later. Why the hell has it consistently been like this ever since elo was implemented? I do 700 damage, 6 kills. An atlas does 500 and the other 2 kills. 5 of the team tunnel rush on forest colony and die (as i informed them when I asked them not to do so) 3 doing less than 50 damage the other 2 less than 100. The last person runs around the tunnel in the open and dies. This happens a majority of my games, except usually I die only killing 2-3 when the odds are so bad. Whyyyy???? I am sick of this I want private matches.

[

Edited by and zero, 27 April 2013 - 11:45 PM.

[Wintersdark]

re: and zero -

I can't say I've ever had that problem.

I've never - not once, not since I started in November - seen people respond poorly to a "Stay together" request. Not even once. I've had poor teammates, but generally speaking since Elo went in they've been generally pretty decent.

The vast, vast majority of my matches see most players performing generally on par with each other. Obviously those killed earlier will do less damage, but I simply don't see matches where only one person did any damage at all and got lots of kills. I've seen the odd one where one fellow gets a lot of damage in a loss, but no kills, because the team was flat out rolled and he was in the back lobbing LRM's uninterrupted, but that sort of thing is definitely an outlier.

Ryvucz, on 27 April 2013 - 05:32 PM, said:

Alpine Peaks is the only real map that doesn't have terrain that can be drastically avoided by jumpjets at the moment.

I'm talking about the sheer cliff edges that any mech with jumpjets can clear for faster point cap.

This is not about damage done, this is about capping points with speed.

I could care less who did the most damage, what irks me is 3 lights with jumpjets on a map that maneuverability in a conquest mode plays a very strategic advantage.

But even if they didn't have that strategic advantage, even if they didn't cap win, they still would have won. Yes, JJ's and speed confer a strategic advantage in Conquest matches. And? Speed and JJ's come at a cost of firepower and/or defense.

The matchmaker doesn't take Jump Jets and speed into account, only weight class. And even then, it loosens the bounds to get you matches faster (hence how you got two mediums instead of lights). The faster/more maneuverable team does have somewhat of a strategic advantage, but you've got one in outright combat. So? Matches are not and can not be 100% fair. This one wasn't terribly unfair either, though. You didn't lose because they had a strategic advantage, you lost because they just plain beat your teams *** down. Accept it, move on, do better next time. Now, I'm not saying you personally played badly, but their team outplayed yours overall.

Really, you're expecting more out of the matchmaker than is really reasonable.

[Purplefluffybunny]

Ryvucz, on 27 April 2013 - 01:08 PM, said:

[...] I understand there are going to be tweaks, but ELO is horrid for this kind of game, with 7 other people involved on your team, especially in a tournament setting.

I agree OP with your general sentiment. On the bit I cite, everything seems like it has been bought 'off the shelf' rather than developed from the ground up.

[Muaziz]

I don't think ELO can be that efficient with current user base. Matches must be created constantly and if there are no enough players then teams become unbalanced. I think PGI have two options: 1) create games quickly and allow unbalanced games or 2) try to set up balanced games (using ELO) but players must wait long times.

[and zero]

Wintersdark, on 28 April 2013 - 10:36 AM, said:

re: and zero -

I can't say I've ever had that problem.

I've never - not once, not since I started in November - seen people respond poorly to a "Stay together" request. Not even once. I've had poor teammates, but generally speaking since Elo went in they've been generally pretty decent.

The vast, vast majority of my matches see most players performing generally on par with each other. Obviously those killed earlier will do less damage, but I simply don't see matches where only one person did any damage at all and got lots of kills. I've seen the odd one where one fellow gets a lot of damage in a loss, but no kills, because the team was flat out rolled and he was in the back lobbing LRM's uninterrupted, but that sort of thing is definitely an outlier.

Maybe I am stuck in some sort of "elo hell'?

Back before myself and most of my friends largely stopped playing, when elo was implemented and we would play in groups what I described would happen to an obscene degree. The other 4 on our team would do like 50 damage while we would all do 500+ with a few kills. It seemed as though the matchmaker was taking 8 average players and putting them against 4 very good players (my team) and 4 terrible players to average out the elo ratings. Of course we didn't know our elo scores, but we all had high k/d ratios and very good win loss ratios.

Maybe you are just more fortunate than I.

[Wintersdark]

and zero, on 28 April 2013 - 02:19 PM, said:

Maybe I am stuck in some sort of "elo hell'?

Back before myself and most of my friends largely stopped playing, when elo was implemented and we would play in groups what I described would happen to an obscene degree. The other 4 on our team would do like 50 damage while we would all do 500+ with a few kills. It seemed as though the matchmaker was taking 8 average players and putting them against 4 very good players (my team) and 4 terrible players to average out the elo ratings. Of course we didn't know our elo scores, but we all had high k/d ratios and very good win loss ratios.

Maybe you are just more fortunate than I.

I solo drop.

If you group drop, it does funny things with your Elo. In a 4 man group, you'll be much more successful that you'd be with 4 equally skilled but ungrouped/no comms random people. So, this pushes your Elo bracket much higher than it "should" be. Before this patches fix, your Elo score would increase much too quickly, too, which made the problem much, much worse. If you're already a very good player who'd have a reasonably high Elo regardless, things get funky. The Matchmaker tries first to get as many people from your Elo bracket together as possible, but it'll also try to create a balanced match. With four guys with very high Elo scores from group dropping, if you're playing at less populated times it's difficult to get sufficient people at that play level all clicking LAUNCH within a couple seconds of each other. My theory, then, is that it's tossing in random terribads to try to "balance" the match so it's not just a pugstomp.

Early Elo was, of course, very buggy and needed more seeding - lots of people don't play a lot (myself included) so even with the month headstart, it needed some time to level out.

To see, try running some random solo drops, or just grouped with one other fellow, during a busy time. If my theory is correct, you should get much better matches.

But still... like I said, I don't ever see what you talked about. I play roughly 6-8 matches a day, and since Elo was fixed in the recent patch, virtually all my matches have been closely fought, awesome games, where pretty much everyone is at least decent. Most have been 8-4 or closer.

[and zero]

Wintersdark, on 28 April 2013 - 02:47 PM, said:

I solo drop.

If you group drop, it does funny things with your Elo. In a 4 man group, you'll be much more successful that you'd be with 4 equally skilled but ungrouped/no comms random people. So, this pushes your Elo bracket much higher than it "should" be. Before this patches fix, your Elo score would increase much too quickly, too, which made the problem much, much worse. If you're already a very good player who'd have a reasonably high Elo regardless, things get funky. The Matchmaker tries first to get as many people from your Elo bracket together as possible, but it'll also try to create a balanced match. With four guys with very high Elo scores from group dropping, if you're playing at less populated times it's difficult to get sufficient people at that play level all clicking LAUNCH within a couple seconds of each other. My theory, then, is that it's tossing in random terribads to try to "balance" the match so it's not just a pugstomp.

Early Elo was, of course, very buggy and needed more seeding - lots of people don't play a lot (myself included) so even with the month headstart, it needed some time to level out.

To see, try running some random solo drops, or just grouped with one other fellow, during a busy time. If my theory is correct, you should get much better matches.

The solo drops are exactly what I am talking about, the groups thing was just another example. I have played plenty of solo drops recently. What I said applies to both situations, sadly. Also, just to clarify, I am NOT saying the other team is way better than I. They generally seem to be average, with maybe one or two players that I would call above average. The issue is the number of players on my team far below average.

Wintersdark, on 28 April 2013 - 02:47 PM, said:

But still... like I said, I don't ever see what you talked about. I play roughly 6-8 matches a day, and since Elo was fixed in the recent patch, virtually all my matches have been closely fought, awesome games, where pretty much everyone is at least decent. Most have been 8-4 or closer.

See now that is very interesting. About 80% of the games I've played since elo are still ridiculously one sided. 8-0 or 8-1 generally (win or lose). Of course this type of game (so heavily team based) is quite likely to have many situations like that because any small advantage quickly snowballs if a team takes advantage of it. But it shouldn't happen that often.

Edited by and zero, 28 April 2013 - 09:00 PM.

[Ryvucz]

and zero, on 28 April 2013 - 02:19 PM, said:

..... The other 4 on our team would do like 50 damage while we would all do 500+ with a few kills. It seemed as though the matchmaker was taking 8 average players and putting them against 4 very good players (my team) and 4 terrible players to average out the elo ratings. Of course we didn't know our elo scores, but we all had high k/d ratios and very good win loss ratios.

Like this?

[and zero]

Ryvucz, on 28 April 2013 - 08:57 PM, said:

Like this?

I am quite depressed to say; yes. Precisely like that.

[Ryvucz]

I'm such a badass, I take up an entire 4 man lance solo.



[Please note, not a name and shame, unless you count me as being such a badass]

[LeShadow]

Wintersdark, on 28 April 2013 - 10:36 AM, said:

[...] but I simply don't see matches where only one person did any damage at all and got lots of kills.

The tournament leaderboard tells me these matches do happen quite frequently.