38 replies to this topic

[occusoj]

First off, Im a noob (just a bit more than 500 drops) and play at the very bottom of ELO.
edit: my post got cut off here, more follows

Ive read a lot about the new MM and some claim it works wonders. While it did shorten the time it takes to get into a match many of them are still stomps. One team usually has no coordination whatsever and at least a whole lance that has not the slightest clue of what its doing. Thats not the bad thing, I belong to that scrub level.
In most cases though the other team isnt burdned with a load of total derps and just rolls over them.

So, dear PGI, please fix it. Its no fun, neither when you roll the other team nor when they wipe you. Only a small minority of matches ended as stomps because of equipment imbalance, for example one team with ecm and 3 hardcore missile boats and the other one with no ECM at all. On caustic. Most wipes resulted of bad gameplay on one side. Isnt that what MM should attempt to sort out? Make both teams so they have roughly the same skill?

The last 29 matches Ive had ended:
12-2
3-12
12-5
11-4
12-4
12-1
11-4
12-3
12-9
12-3
4-12
12-3
11-4
12-5
12-6
11-4
12-5
9-12
12-2
12-5
4-11
12-5
11-1
5-12
12-2
12-4
12-5
4-12
3-12

Even though I endet up on the winning team quite often, its not gameplay that I enjoy. Just walk over them and dont derp hard enough to get sloughtered. The numbers dont do some of these stomps no justice.

Previous MM wasnt any better, just to mention it.

TL;DR: Not all is fine and perfect, new MM still puts me into a lot stomps, even though its faster in doing so.

Edited by occusoj, 06 July 2014 - 12:34 PM.

[Karamarka]

It's very hard to make a comeback in MWO i've found for most matches.

Because your average player has no idea about mech lab, combat and tactics - most of the people are cannon fodder and can only succeed if they can sit still and turret, if you get a few of those on your team you are screwed.

Even at a 55% win/rate and 2.0KDR i just had this game not 5 minutes ago.

Edited by Karamarka, 06 July 2014 - 01:06 PM.

[Willard Phule]

Wow. That's rough. Sorry, but, it's working as intended.

The developers, in their infinite wisdom, decided that new players should start out in the middle of the Elo range...as opposed to starting out lower and working their way up.

I'm not sure why they decided that, considering the matchmaker does it's level best to keep you at a 1:1 win/loss ratio. I can only guess the point is to get you interested in the game, hit you up for $30-60 bucks on a brand new cash grab, then make you lose interest because it's simply not fun for you. It's really not a big deal for them if you quit and leave because there will be a fresh batch of new guys next month. And the cycle will repeat itself.

[Tripzter]

They "fixed" the MM like they fixed the UI. To be honest i'm afraid that they will start messing around with the earnings of xp and credits formulas next...

Edited by Tripzter, 06 July 2014 - 12:58 PM.

[WhiteRabbit]

Karamarka, on 06 July 2014 - 12:50 PM, said:

It's very hard to make a comeback in MWO i've found for most matches.

SNIP

That's right. A Shooter without respawn tends to develop those kind of matches. 1 down is bad..2 down is worse and as soon as 3 of your team are dead and the enemy still has 12 (that are not walking sticks without weapons)....well you're boned. It's just the core-mechanic of that game--> lose a mech and your team loses firepower while the enemy still has the same firepower and from that point on it snowballs

[EgoSlayer]

"Stomps" or overwhelming wins/losses *ARE* normal because there are no respawns. Once a team has a 2-3 kill advantage, especially if they are still at their full 12 members, they have about a 30% firepower advantage over the enemy.
The only way a team can overcome that deficit is through errors done by the team with the advantage, or through superior teamwork on the team that is down even if everything thing else is equal (Skill, loadouts, etc). In PUGs the former sometimes happens if they are too spread out instead of the rolling deathball group, the latter is exceptionally rare in PUGs because that level or coordination is difficult to perform though chat.

[Karamarka]

WhiteRabbit, on 06 July 2014 - 01:06 PM, said:

That's right. A Shooter without respawn tends to develop those kind of matches. 1 down is bad..2 down is worse and as soon as 3 of your team are dead and the enemy still has 12 (that are not walking sticks without weapons)....well you're boned. It's just the core-mechanic of that game--> lose a mech and your team loses firepower while the enemy still has the same firepower and from that point on it snowballs

Yep, and with the idea of armor and a high time-to-kill. It's not like Counterstrike where you can pull off 5 headshots and secure the victory, if you are 1v5 you have almost 0% chance, even if you get the jump on someone they will get shots off and hurt you at least.

[ShinobiHunter]

The MM can't read a players actual abilities because if you get placed with a good team your stats will go up even if you are terrible. Likewise, even a stellar player will be brought down if placed with a lousy team.

[Bront]

12 - 4 is not always the landslide it looks like it is. Quite often, the difference between winning and losing a game is if mechs are dropping from damage or not. I've seen my team come back from being down 8-2 several times simply because 7 of the enemy mechs were on their last legs, while the 4 we had remaining were still doing pretty good.

I don't see a lot of 1 or 2 team kills on your list, and those are the real landslides. At least a 12-4 losing team took out 4 mechs, and likely damaged a few more.

[Fleeb the Mad]

There's a concept called the N Squared Law which applies to attrition-based single spawn games like MWO and is the reason behind a lot of blowouts, even on evenly matched teams.

It's based on the fact that a badly damaged mech can still have most or all of its offensive ability intact. Even if mechs on your team are inflicting damage, if they don't remove enemy mechs from the battlefield as they go down themselves, the enemy is increasingly more able to bring unequal firepower down on your outnumbered force. As one side loses mechs, they lose more faster because of the inequality in firepower, unless they're able to kill off all the wounded enemies quickly.

Have a look at the patterns of kills in your games. If you see these blowouts begin with a couple people being killed in the first few minutes without taking out any enemies, then it could be something so simple as a couple players making bad decisions that push your team past the tipping point early.

[ScarecrowES]

Also... please understand that your personal Elo rating does not ensure you will be playing with or against only players of similar rating. It only means that at least one player on the enemy team will have a similar Elo rating as you, to make the teams "balanced." Inevitably, some players on a team will have high Elo rating, some will have low, and others will be in the middle in varying degrees. The idea is that both teams will have similar composition, not that all players in the match are similarly ranked.

[Willard Phule]

ScarecrowES, on 06 July 2014 - 01:22 PM, said:

Also... please understand that your personal Elo rating does not ensure you will be playing with or against only players of similar rating. It only means that at least one player on the enemy team will have a similar Elo rating as you, to make the teams "balanced." Inevitably, some players on a team will have high Elo rating, some will have low, and others will be in the middle in varying degrees. The idea is that both teams will have similar composition, not that all players in the match are similarly ranked.

Which is crap.

New players should have a lower Elo than people that have been around for a while. They shouldn't be thrown into the deep end with the big boys...they should be in their own little bracket and as they get better, they move upward into higher levels. It's relatively simple.

As it is right now, there are BASIC skills that they simply don't have the time to learn. Shooting without stopping and zooming in, not shooting their own team in the back in spawn (although, arguably, you'd think that even a complete mouthbreather would know that one), sticking together, locking targets.......odds are fairly good that they're dead in the first few minutes of each match, so they never get a chance to practice it.

Granted, I'm SURE they could learn all those things with the tutorials...er....nevermind. Scratch that. They CAN read a tutorial in the forums because, well, the written word is far better than actual experience. At least that's the theory PGI works under with their Elo system. Elo...Extremely Laughable Outcome.

How come you never see Russ, Bryan or any of the other PGI goons dropping solo into the public queue? Hmmm....makes you wonder, doesn't it?

And, alternatively, once the 6-8 new players on the team die, everyone else goes with them.....that is, if they haven't been teamkilled by Derp McDerpalot and his alpha strike to the back in spawn.

Edited by Willard Phule, 06 July 2014 - 01:28 PM.

[occusoj]

Comebacks are extremely rare and difficult, I second that.

But in many of my matches the other team is dominated from 0:00 on. One team usually starts with 2-4 people that are vastly below the rest in terms of "skill" - or whatever one likes to call it. So one team is 4 down from the beginning - very tough job to pull a win from that if the other team gets its act together.

Now thats where I would have expected MM to balance it out and not add, for example, 4 total beginners to one team but 2 to each. If ELO isnt suitebale for that task then toss it. Or supplement it with other information like hours played, accuracy, asssists,.... so much information available.

Bront, on 06 July 2014 - 01:12 PM, said:

12 - 4 is not always the landslide it looks like it is. Quite often, the difference between winning and losing a game is if mechs are dropping from damage or not.

Some of them have been, the numbers dont do it justice. On many 12-3 or 12-4 matches the ones getting killed on the winning team where either rambos that charged in and died or some poor lights/mediums/damaged heavies that were a bit too ambitious at the end. Entierly avoidable deaths. Not in every match, some 12-4s werent THAT bad, but most really should have ended 12-2 or even less.
The domination was noticeable from the point of first confrontations right to the end. Speaking of fun or enjoying the games - nothing to be found here. Doesnt matter on what side I am - its not a fair or balanced fight and thats what matters to me. Even if the enemy team wins it can still be a great game with demanding fights going on.

Though there are some really good matches around and thats what keeps me at MWO. PGI is about money but they wont get mine until some things are fixed - MM beeing one of em. New players beeing assigned an average ELO is pure madness IMHO. Whoever came up with that must posses some very deep hatred towards noobs.

The lack of better introduction/tutorials for batlle isnt nice either, it took me some games to learn the basics of battle - not beeing a great help to my team. Not the very basic stuff like how to group weapons or how to move but when to be where and do what.

Quote

Have a look at the patterns of kills in your games. If you see these blowouts begin with a couple people being killed in the first few minutes without taking out any enemies, then it could be something so simple as a couple players making bad decisions that push your team past the tipping point early.

Thats exactly what can be observed in many games. It starts with 2-3 people derping up and then the roll gains traction - especially if one team looses loads of firepower and/or damage-soaking ability. Ive seen entire matches beeing decided by 1 (one) Jenner that caused half of the heavies in a firing line to chase him.

Edited by occusoj, 06 July 2014 - 01:45 PM.

[Bront]

occusoj, on 06 July 2014 - 01:44 PM, said:

Thats exactly what can be observed in many games. It starts with 2-3 people derping up and then the roll gains traction - especially if one team looses loads of firepower and/or damage-soaking ability. Ive seen entire matches beeing decided by 1 (one) Jenner that caused half of the heavies in a firing line to chase him.

This is part of why I don't mind playing aggressive at times. When your team can take advantage of it, getting the enemy to chase a squirrel can turn, or drawing and disorienting the enemies fire as a DDC charging down on them while teammates come though behind and mop up unnoticed is just plain awesome, even if the in game rewards aren't that great for it.

[Sandpit]

No need for another queue.

Simply start new players at the bottom of Elo. That way their natural skill and ability improve while they're continuously playing against more skilled opponents without ever getting in over their head.

Alternatively...

I've said for years now that new players need to have their own cadet queue, only players under the cadet protection period allowed in, vet a few experienced players to allow them to drop into the queue in order to help, answer questions, etc. Give them a forum badge and a little bonus to their winnings and they'll happily donate time to the cadet queue as a drill instructor.

That would improve the NPE, mitigate roflstomps, and increase retention of players as they don't start their first experience in MWO with a string of roflstomps

[Goose]

… It's also possible it'll take a few weeks for Elo KB to adjust/ switch over to the new de jure …

Edited by Goose, 09 July 2014 - 12:46 AM.

[OneEyed Jack]

ScarecrowES, on 06 July 2014 - 01:22 PM, said:

Also... please understand that your personal Elo rating does not ensure you will be playing with or against only players of similar rating. It only means that at least one player on the enemy team will have a similar Elo rating as you, to make the teams "balanced." Inevitably, some players on a team will have high Elo rating, some will have low, and others will be in the middle in varying degrees. The idea is that both teams will have similar composition, not that all players in the match are similarly ranked.

Willard Phule, on 06 July 2014 - 01:27 PM, said:

Which is crap.

What it is, is untrue. The MM doesn't remotely try to balance each player against a similar player. It uses total team Elo, only.

[ScarecrowES]

OneEyed Jack, on 06 July 2014 - 02:36 PM, said:

What it is, is untrue. The MM doesn't remotely try to balance each player against a similar player. It uses total team Elo, only.

I would ask you to refer to the full description of matchmaking system priorities straight from PGI, and specifically the recent update.

Regardless... even if you make the assumption that the system operates on a total team basis, it still ensures you have players of varying skill level on the same team, and a total overall balance between teams. There is no BRACKET that you're placed in, and thus only play other players in the same bracket. It doesn't work that way.

Further regardless... Elo is a useless rating system for any competition in which the competitors are fundementally random, as they are in any matchmaking system. For all intents and purposes, Elo has no practical effect on team composition in this game anyway, as an individual's Elo score in no way reflects that individual's actual ability. As Elo, then, is a randomizer for matchmaking, you can discount its effect on the outcome entirely.

Edited by ScarecrowES, 06 July 2014 - 03:19 PM.

[OneEyed Jack]

ScarecrowES, on 06 July 2014 - 03:06 PM, said:

I would ask you to refer to the full description of matchmaking system priorities straight from PGI, and specifically the recent update.

Really? Do please link this complete list which states that they've reversed the system from the team-based Elo it's been since Elo was introduced to an individual 1:1 Elo system. Nothing I've seen has suggested this.

[ScarecrowES]

OneEyed Jack, on 06 July 2014 - 03:16 PM, said:

Really? Do please link this complete list which states that they've reversed the system from the team-based Elo it's been since Elo was introduced to an individual 1:1 Elo system. Nothing I've seen has suggested this.

http://mwomercs.com/...302-02-jul-2014

• The matchmaker will attempt to match based on both Elo and weight class composition (3’s)
  
• Over time, the matchmaker will relax both Elo and weight class composition constraints. In the case of weight class composition, the matchmaker guarantees both teams end up matching.

The statements imply a player-for-player match for both points of composition.

Like I said, though... it doesn't matter either way. Elo was designed for direct player to player matches. It can be adapted to team-based competition, but only to apply to the entire team, and not to individual players. However, it requires that team composition remain static from event to event, and is thus treated as a single "player" made up of multiple participants rather than a team. Elo doesn't work in matchmaking as the team composition changes from match-to-match. It's a bit like trying to rank NFL teams when the players on each team are randomized every week - all the players going into a pool and randomly thrown out to each team just before match time - and then trying to say what effect a single player in the entire player community is having on their particular team week to week based on who won what game. Using Elo in a game like this is beyond stupid... and as I said, has NO influence whatsoever on the outcome of matchmaking games.

Edited by ScarecrowES, 06 July 2014 - 03:35 PM.