Jump to content

Matchmaking Phase 3

Feedback v1.2.190

483 replies to this topic

#341 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 24 February 2013 - 10:40 AM

View PostWraithMandaka, on 24 February 2013 - 10:14 AM, said:

Current matchmaking system is for single players who do not want to join or play in a group. And that is fine, for those players. It keeps things things balanced (or not in its current form) and people will fight against others with similar skill levels.


Current matchmaking system is for stupid moneybag players who can't play for crap and getting stomped all the time because they go into the open and eat 200 LRMs or smth like this. But because PGI needs them to keep playing and feed them with money they make a matchmaker that puts them against same type of players. On the other hand a good player is dropped vs good players, but problem is - to be competitive you need to use cheesy builds. Game forces you to use mechs you don't want to or lose all the time. ELO is stupid and needs to be gone. Matchmaker we had in closed beta was broken in many ways but it was still better then what we have now with phase 3. All you needed to do is make separate ques for players in groups and solo players and everybody would have been happy. Right now you either suck in order to have fun or play well without any fun at all.

However PGI has already said that they only care about 'lone wolves' because blah blah blah they are the majority. Good players among lone wolves never cryed about how they were getting stomped in a team-concept game when they were not part of the team, and they were happy with the matchmaker as it was in closed beta. A year ago PGI said that this would be a team game, now all they do is done in favor of lone wolves (and bad ones at that), the whole idea of whom is against the idea of a team game. I guess a lot of money can really change your ideals...

#342 Rumrunner2

    Member

  • PipPipPipPipPipPip
  • 408 posts

Posted 24 February 2013 - 12:47 PM

View PostRadimentriX, on 24 February 2013 - 05:49 AM, said:

seriously: **** you, PGI!
Posted Image
5,5 lights with 4x ecm vs no lights and ecm at all


Same here, since f********* MM3 and ELO i have most uneven matches ever.

3 Assault (2 DDC with ECM, one Stalker) and 2 Raven 3L with ECM vs. one Assault (Stalker) and no Lights, no ECM.
PGI, it su*** !
But like ECM I think PGI is so beloved in there own "great" ideas that we will see not much changes
PGI started MM3 with completely wrong way:

instead:
FIRST same number of mechs in weight classes, secondary ELO, if necessary slightly uneven ELO

we have:
FIRST identical ELO at all costs, secondary mechclassen, mostly uneven mechclasses and a lot of very uneven matches.

I feel remebered to BlackProphecy and ReakktorMedia and there stupid "ideas", game crashed after 1,5years, servers shutdown, Developer insolvent...

#343 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 24 February 2013 - 12:53 PM

Now there is smth else I just thought of. Can somebody in PGI explain to me why am I getting less c-bills and exp then prior to bloody phase-3?? Surely I fight with and against people of my skill level thus getting less kills / assists / damage per game, but why for all the love of god smb with ELO = 0 is getting same c-bills and exp I do in his ELO-balanced games?? Right now there is no reason for me to play well, be an *****, drop your ELO and enjoy easy-breezy games with low ELO people...

Edit: To think about it ... why is my KDR is gonna be about same as people in other ELO-zones?

Edited by PhoenixFire55, 24 February 2013 - 01:16 PM.


#344 Tolkien

    Member

  • PipPipPipPipPipPipPipPip
  • Giant Helper
  • Giant Helper
  • 1,118 posts

Posted 24 February 2013 - 02:52 PM

Feedback so far on efficacy of matchmaking phase 3:

Prior to ECM I ran a catapult C1(F) with LRMs and had a KDR of 0.92.

Since ECM and prior to this patch I ran a craven 3L and had a KDR of 2.62.

Since the most recent patch, I am still running a Craven 3L or atlas DDC in 95% of matches and my KDR is up to 2.83.

*golfclap* to ECM balance
*golfclap* to PPC affecting ECM
*golfclap* to MM phase 3. I have played 75-100 games since MM phase 3 and I don't think it is helping since my KDR is still climbing. If 2 Raven 3L's with ECM working as a pair (my most common drop configuration with a real life friend) is able to so upend team balance, there is something seriously wrong with the item. (Some drops today were with 2 friends, but 12 on friday were solo too).

Edited by Tolkien, 24 February 2013 - 02:53 PM.


#345 Kashaar

    Member

  • PipPipPip
  • Overlord
  • Overlord
  • 75 posts

Posted 24 February 2013 - 03:32 PM

Posted Image

What is wrong with this picture?

#346 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 03:46 PM

You didn't block out the names of players in accordance with forum policy? ^_^

#347 Kashaar

    Member

  • PipPipPip
  • Overlord
  • Overlord
  • 75 posts

Posted 24 February 2013 - 03:57 PM

View PostVoid Angel, on 24 February 2013 - 03:46 PM, said:

You didn't block out the names of players in accordance with forum policy? ^_^


That is policy? I'm sorry, I didn't know. Where is this written? I've seen endgame screens posted many times, so didn't think it would be a problem. Besides, it's just supposed to illustrate the massively out of whack distribution of mech types in that game. It was ridiculous.

#348 Rumrunner2

    Member

  • PipPipPipPipPipPip
  • 408 posts

Posted 24 February 2013 - 03:57 PM

View PostKashaar, on 24 February 2013 - 03:32 PM, said:

Posted Image

What is wrong with this picture?


Its the masterpiece of of MM3 and ELO. PGI did not learn anything from ECM desaster, they still believe ECM is fine.....
Instead fixing ECM they just gave us a much bigger desaster.
It proves too its better to deinstall that f******g game instead hoping PGI is able to make a good game with fair balancing.
Btw, last week I learned PGI was involved in the legendary "Duke Nukem forever"...explains a lot to me.

#349 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 04:22 PM

View PostPhoenixFire55, on 24 February 2013 - 07:46 AM, said:

Not matching mech types is complete bull*hit. Alpine map is the size all maps should be, now imagine one team has 2 lights and other team has none and tell me who is gonna win conquest game on Alpine? Mech types should be no.1 priority for matchmaker, when it did match them everybody was happy the way it worked. Nobody needs this ELO crap which makes no sense anyway. Put 8 people of same skill in 3L Ravens and 8 people of the same skill as group one in mediums and guess who is gonna win?

the team that has no lights should win on Alpine - unless they're stupid. Because they can see they have no scouts, and they know that they might be seeing lights on the new map. So they need to set up shop near their base and just smoke the other team when they show up to the party. That's what my last no-light Alpine team did.

As for Elo being "crap which makes no sense," please. Elo is an excellent system for ensuring fun matches by matching opponents of equal skill against each other to avoid mismatches. It is unfortunate that the current system is resulting in vastly uneven weight classes in so many matches - but the logical, grown-up response to this is to say "hey, we really need to adjust the matchmaker to enforce equal weight classes a lot more." Instead, so many posters are screeching like angry monkeys pounding on a radio with a rock - because they don't like what it's playing and don't know how to change the channel.

View PostPhoenixFire55, on 24 February 2013 - 10:40 AM, said:

Current matchmaking system is for stupid moneybag players who can't play for crap and getting stomped all the time because they go into the open and eat 200 LRMs or smth like this. But because PGI needs them to keep playing and feed them with money they make a matchmaker that puts them against same type of players. On the other hand a good player is dropped vs good players, but problem is - to be competitive you need to use cheesy builds. Game forces you to use mechs you don't want to or lose all the time. ELO is stupid and needs to be gone. Matchmaker we had in closed beta was broken in many ways but it was still better then what we have now with phase 3. All you needed to do is make separate ques for players in groups and solo players and everybody would have been happy. Right now you either suck in order to have fun or play well without any fun at all.

However PGI has already said that they only care about 'lone wolves' because blah blah blah they are the majority. Good players among lone wolves never cryed about how they were getting stomped in a team-concept game when they were not part of the team, and they were happy with the matchmaker as it was in closed beta. A year ago PGI said that this would be a team game, now all they do is done in favor of lone wolves (and bad ones at that), the whole idea of whom is against the idea of a team game. I guess a lot of money can really change your ideals...


Just... wow. Elo only affects "stupid moneybag players?" You "have to use 'cheese builds?'" I've been smoking people (within the confines of my limited play time) with an Atlas D-DC using 45LRMs, a Large Laser, and an ERPPC. That's hardly a "cheese" build, but it works for me because I'm a good player and I know how to use it. Certainly some builds are more powerful than others. That's why we have Beta testing to tune things. But while I've never used a Splatcat or any "cheese" build - my KDR is positive and I win more matches than I lose by a good margin.

Then you go into misrepresenting PGI - this is called lying. You really shouldn't do it in such an obvious way. I'll bet you your last brain cell that PGI has never said they "only care about" any part of their player demographic. But I guess this conspiracy-nut diatribe explains your "moneybags" comment earlier. Because PGI is stupid, you see; they don't understand business practices and game design - as opposed to experts like you. No, PGI just pays attention to hypothetical money that you assume is being dangled in front of them by fictional "lone wolves" who "cryed [sic] about how they were getting stomped in a team-concept game when they were not part of the team." Your baseless, bitter invective is devoid of merit, insulting toward PGI, and offensive to me.

Those of us (which is to say, nearly everyone) who "cryed [sic]" about synch dropping aren't bad players who don't want to play on a team. We WERE on a team - we joined to play against a supposedly random team where no more than four players were supposed to be in their own group. What we got was a wall of 8 Liao tags shouting "SQUAAAAAAAAAAAAWK" in all chat to begin the match. This isn't about all the bad, mean "lone wolves" ruining the game for "real" players like you. This is about closing an exploit that was causing frequent hardship and frustration for the vast majority of the playerbase - in fact, the bad players are the ones you want to help. Truly good players don't have to cheat in order to feel powerful.

In closing, You can't claim that an Elo system is bad because you don't like 'mech balance. If (and I do not cede you this or any other point) game balance is really so bad that you "have" to play "cheese" builds to be competitive, then the problem is with game balance, not the Elo system. It is wholly irrational for you to pour out your hateful diatribe against the imaginary bad players (and PGI) who are "ruining the game," then claim that you'd rather play with them than the "good players" whom you assume are on your own level of skill.

Edited by Void Angel, 24 February 2013 - 04:25 PM.


#350 CG Oglethorpe Kerensky

    Member

  • PipPipPipPipPipPip
  • 420 posts

Posted 24 February 2013 - 04:25 PM

View PostVoid Angel, on 24 February 2013 - 03:46 PM, said:

You didn't block out the names of players in accordance with forum policy? ;)


These screenshots are not referencing any player committing any inappropriate behavior, rather this is directed at the Matchmaking system. The "Name and Shame" policy is not applicable as no one is accusing the players of any wrong doing.

#351 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 04:27 PM

Depends on who you ask - I'm just teasing you anyways. ;)

#352 Sable Dove

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,005 posts

Posted 24 February 2013 - 04:40 PM

Getting real sick of 8 vs 6 games.

#353 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 04:46 PM

Yeah, or 7v8 games where their 7th was the only assault 'mech in the game. That's a legitimate problem with the matchmaker, and one I'm sure they're going to fix ASAP.

#354 CG Oglethorpe Kerensky

    Member

  • PipPipPipPipPipPip
  • 420 posts

Posted 24 February 2013 - 04:57 PM

I am officially throwing in the towel.

As much as I like the Treb, I will be playing exclusively assaults and heavies until this is fixed. There is no reason not to.

#355 Kaeb Odellas

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 2,934 posts
  • LocationKill the meat, save the metal

Posted 24 February 2013 - 09:09 PM

I am so sick of playing Alpine against swarms of Ravens. If I didn't already have a burning incandescent hatred for those things, I would have after 5 straight matches of that nonsense.

#356 Vellinious

    Member

  • PipPipPipPipPipPip
  • Elite Founder
  • 254 posts
  • LocationCorn field

Posted 24 February 2013 - 10:31 PM

View PostVoid Angel, on 24 February 2013 - 04:22 PM, said:

As for Elo being "crap which makes no sense," please. Elo is an excellent system for ensuring fun matches by matching opponents of equal skill against each other to avoid mismatches. It is unfortunate that the current system is resulting in vastly uneven weight classes in so many matches - but the logical, grown-up response to this is to say "hey, we really need to adjust the matchmaker to enforce equal weight classes a lot more." Instead, so many posters are screeching like angry monkeys pounding on a radio with a rock - because they don't like what it's playing and don't know how to change the channel.



Actually, it really is a bad system. The ELO system is based on win / loss % to calculate how a player should be ranked. To me, that's a sign of luck, more than it is skill. This isn't a solo game....it's a TEAM game. Therefore, an ELO system based on wins / losses, isn't really sufficient, is it? The worst player in the world, could get lucky enough to win games based on the skill and merits of the teams that he gets dropped with. In this scenario, his ELO would go up, even though he's not really contributing anything to the match. Likewise, a really skilled player gets unlucky and gets stuck, more often than not, with a team that couldn't fight their way out of a wet paper bag. Consistently does 300+ damage and gets a few kills, contributes to the TEAM, but still loses, because the TEAM he got dropped with was just turruble. HIS ELO would go down...thus getting dropped with even MORE useless players.

An ELO system based on efficiency of the player, would be much more accurate, and give a better match making system when accompanied by one of weight class....THIS would actually ensure there are fair matches going on....

This isn't chess...where the WINS and LOSSES of a single player, measure his skill. There aren't 7 others on his team, making moves.... In a SINGLE player game, an ELO system would work fine. In a TEAM game, a win/loss ratio doesn't do anything....

Edited by Vellinious, 24 February 2013 - 10:35 PM.


#357 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 10:41 PM

Well, Raven L pilots should all be tarred and feathered. But You can help yourself a LOT on Alpine by simply setting up a defensive formation near your base if you don't have more than 2 scouts. That way you can still have your scouts do recon, but you don't get screwed if the enemy team has 6 lights to charge your base with. When dealing with a light swarm, it's important to:
  • Not FRACKING try to hug them. If you try to hump a Raven, all you're doing is obstructing your buddies' field of fire.
  • Not ignoring the other squirrels if they get in your rear arc and start walking slowly behind you unloading their weapons.
  • Kill any big 'mechs FIRST if they all charge together.
  • Do not chase the squirrels away from the main fight.
Basically in order to kill them, you need to focus on one of them as much as you can, and stay spread out a little so that you don't obstruct each other's fires. If you're in close (e.g. inside their ECM) try to find one that's pretty damaged and pound on him. If you're far away, pound on anyone that's slowing down to take careful aim because they're not the one getting focused. If they run, let them go (unless they're going to your base,) and try to focus down one of their buddies. It's really like a herd of buffalo fending off a wolf pack. If you refuse to run or be baited, you can simply use your superior firepower (some of you have weapons loadouts that weigh more than their entire Battlemech) to beat them down. Just make sure to cover your base.

#358 Vellinious

    Member

  • PipPipPipPipPipPip
  • Elite Founder
  • 254 posts
  • LocationCorn field

Posted 24 February 2013 - 10:57 PM

I pilot a Raven 3L quite a bit, because....unfortunately, taking any other light mech, other than the ECM commando, is really a waste of time. 9 times out of 10, the raven 3L will take down a non-ECM light with little or no problem..pilot skill being equal and all...


View PostVoid Angel, on 24 February 2013 - 10:41 PM, said:

When dealing with a light swarm, it's important to:
  • Not FRACKING try to hug them. If you try to hump a Raven, all you're doing is obstructing your buddies' field of fire.
  • Not ignoring the other squirrels if they get in your rear arc and start walking slowly behind you unloading their weapons.
  • Kill any big 'mechs FIRST if they all charge together.
  • Do not chase the squirrels away from the main fight.
Basically in order to kill them, you need to focus on one of them as much as you can, and stay spread out a little so that you don't obstruct each other's fires. If you're in close (e.g. inside their ECM) try to find one that's pretty damaged and pound on him. If you're far away, pound on anyone that's slowing down to take careful aim because they're not the one getting focused. If they run, let them go (unless they're going to your base,) and try to focus down one of their buddies. It's really like a herd of buffalo fending off a wolf pack. If you refuse to run or be baited, you can simply use your superior firepower (some of you have weapons loadouts that weigh more than their entire Battlemech) to beat them down. Just make sure to cover your base.



The last part of your post, I completely agree with. With the right strategy, you can eliminate the advantage that the smaller, faster Raven hoards have over you. We WORK to get you singled out....that is ALWAYS our plan. Because we know, one on one, we have a MUCH better chance of taking down the heavier, better equipped mechs. I'd MUCH rather try to take on ONE Atlas, than I would 2 of anything else. Remember that, when fighting against the light squads, and you'll be much better off. Wander off alone, and we'll pick you off faster than a lioness on a 3 legged gazelle.

#359 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,114 posts
  • LocationParanoiaville

Posted 24 February 2013 - 11:10 PM

View PostVellinious, on 24 February 2013 - 10:31 PM, said:

Actually, it really is a bad system. The ELO system is based on win / loss % to calculate how a player should be ranked. To me, that's a sign of luck, more than it is skill. This isn't a solo game....it's a TEAM game. Therefore, an ELO system based on wins / losses, isn't really sufficient, is it? The worst player in the world, could get lucky enough to win games based on the skill and merits of the teams that he gets dropped with. In this scenario, his ELO would go up, even though he's not really contributing anything to the match. Likewise, a really skilled player gets unlucky and gets stuck, more often than not, with a team that couldn't fight their way out of a wet paper bag. Consistently does 300+ damage and gets a few kills, contributes to the TEAM, but still loses, because the TEAM he got dropped with was just turruble. HIS ELO would go down...thus getting dropped with even MORE useless players.

An ELO system based on efficiency of the player, would be much more accurate, and give a better match making system when accompanied by one of weight class....THIS would actually ensure there are fair matches going on....

This isn't chess...where the WINS and LOSSES of a single player, measure his skill. There aren't 7 others on his team, making moves.... In a SINGLE player game, an ELO system would work fine. In a TEAM game, a win/loss ratio doesn't do anything....

Fix your formatting. Your color codes don't work.

As for Elo being a bad system - no actually it's not. Thank you for pointing out that this is a team game, though. I had not noticed! If I were inclined to return the favor, I might advise you that water is wet, fire is a form of combustion, and computer games are played on computers. Your argument, in using the word "TEAM" somewhat inapropriately, demonstrates a misunderstanding of both team mechanics and statistical probability.

First, one of the core skills in Mechwarrior Online is tactical cooperation. You cannot just bring your 'mech and go do what you want to do - because you'll probably lose. That'll suck for your more competent teammates, but because they try to cooperate (hypothetically) in all their games, they're far more likely to win their next matches than you are. So even though you rise and fall as teams in matches, individuals' Elos will still move toward their own level of skill. You cannot disregard tactical cooperation in favor of pure piloting and marksmanship - because this is indeed a team game.

Similarly, your examples are comparing apples to oranges, creating a false cause fallacy in your argument. Take your first example: You start with the "worst player in the world," meaning that he is consistently bad - this refers to his performance over time. Then you postulate a match where his teammates are good enough that the team wins despite his bad performance - this is a single instance. You then claim that this breaks Elo as a system because a single instance of playing the game might yield a counter-intuitive result - such as giving a win to the "worst player in the world." Your reasoning in this instance is invalid. Our archetypical Bad Player is not going to be matched with the same teammates and the same enemies every match. If he really is the world-wide worst, he's not likely to be so lucky in the next match - and the next, and the next, and the next. So while it sucks slightly for the people he tarnishes on the way to the lowest level of Elo Hell, his Elo will still come to approximate his skill level over time.

Thus, Elo does indeed work for MWO. There is nothing magical about the team environment that breaks Elo and makes it an unworkable system.

#360 Vellinious

    Member

  • PipPipPipPipPipPip
  • Elite Founder
  • 254 posts
  • LocationCorn field

Posted 24 February 2013 - 11:24 PM

View PostVoid Angel, on 24 February 2013 - 11:10 PM, said:

Fix your formatting. Your color codes don't work.

As for Elo being a bad system - no actually it's not. Thank you for pointing out that this is a team game, though. I had not noticed! If I were inclined to return the favor, I might advise you that water is wet, fire is a form of combustion, and computer games are played on computers. Your argument, in using the word "TEAM" somewhat inapropriately, demonstrates a misunderstanding of both team mechanics and statistical probability.

First, one of the core skills in Mechwarrior Online is tactical cooperation. You cannot just bring your 'mech and go do what you want to do - because you'll probably lose. That'll suck for your more competent teammates, but because they try to cooperate (hypothetically) in all their games, they're far more likely to win their next matches than you are. So even though you rise and fall as teams in matches, individuals' Elos will still move toward their own level of skill. You cannot disregard tactical cooperation in favor of pure piloting and marksmanship - because this is indeed a team game.

Similarly, your examples are comparing apples to oranges, creating a false cause fallacy in your argument. Take your first example: You start with the "worst player in the world," meaning that he is consistently bad - this refers to his performance over time. Then you postulate a match where his teammates are good enough that the team wins despite his bad performance - this is a single instance. You then claim that this breaks Elo as a system because a single instance of playing the game might yield a counter-intuitive result - such as giving a win to the "worst player in the world." Your reasoning in this instance is invalid. Our archetypical Bad Player is not going to be matched with the same teammates and the same enemies every match. If he really is the world-wide worst, he's not likely to be so lucky in the next match - and the next, and the next, and the next. So while it sucks slightly for the people he tarnishes on the way to the lowest level of Elo Hell, his Elo will still come to approximate his skill level over time.

Thus, Elo does indeed work for MWO. There is nothing magical about the team environment that breaks Elo and makes it an unworkable system.


And you're still ignoring the fact, that they're ranking individual players, based on a win / loss % that takes a team to build. If a player plays with a group of friends, it's quite possible, that he gets carried for 100 or so matches, EASILY. But then decides to PUG a few games, and just gets owned, because his ELO is inflated. Likewise, a PUG player, that's REALLY good, could lose more matches, than he wins, because he gets in drops with players that don't work together. Can he make up for that EVERY time? Likely not. Pilot skill does play a part in the team dynamic...sure, you can take 8 average players that are working together, and defeat 8 really good individuals that AREN'T. That's a given....but player skill DOES affect the outcome. To say that it doesn't, is just silly. And to weigh the entire ELO system on wins and losses is too.

Didn't use any color codes, btw....not sure what you're seeing, but it looks pretty normal to me.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users