Jump to content

Pgi - Please Enlighten Us About Mm!

Gameplay

140 replies to this topic

#21 Drogra

    Member

  • PipPipPip
  • Legendary Founder
  • Legendary Founder
  • 58 posts
  • LocationSLC, Utah

Posted 02 September 2014 - 07:51 AM

I would love to see Drop Weight in solo queue, with clan mechs having an "effective tonnage" increase (its been talked about all over). That would do more to balance a lot of drops then most other changes, and would be fairly easy to implement.

#22 KodiakGW

    Member

  • PipPipPipPipPipPipPipPip
  • The Jaws
  • The Jaws
  • 1,775 posts
  • LocationNE USA

Posted 02 September 2014 - 07:55 AM

Regarding this. Has anybody else noticed that if you leave a match before it ends (usually because your team melted and you don't want to watch a bunch of heavy and assaults chase an ECM Spider on Terra Therma), that your attempt to find a match in another mech (even one that has extremely low percentage in queue) takes MUCH longer?

That and after waiting 30-45 seconds, you hit Cancel and suddenly it finds you a match?

I'd be interested to see how many people also experience this.


#23 Shlkt

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 319 posts

Posted 02 September 2014 - 08:06 AM

Quote

He was solo pugging on the stream to get the last 25 wins... I dont know how many it ended up taking, but it was well over 50.


Assuming a 50% win rate (and in the solo queue it should be, even for Jager), there's a 10% chance that you'll need more than 60 games to get 25 wins. Unlikely but not bizarrely so. In that time you have a 10% chance of seeing a losing streak of 6 or more games.

TBH the best way to break a losing streak (in the solo queue) is to just keep doing what you're doing. If you're losing then your Elo is dropping, and eventually you'll get matched against opponents that you can beat. If instead you get fed up and switch mech classes then all those losses become irrelevant to the matchmaker - suddenly you're working with a totally different Elo rating; a new chance to be ground into the dust :)

What some players may not appreciate is that your Elo can fluctuate based on meta behavior that has nothing to do with your skill. Let's say I like to run a NARC-equipped Jenner. Week 1 sees a new patch which buffs LRMs; suddenly I go on a huge winning streak because my NARC is a huge asset to all the guys trying out the new LRMs. My Elo skyrockets. Week 2, people get bored of LRMs and switch back to laser barf. I keep dropping in my NARC mech, though, which worked so well last week... and I experience a huge losing streak. The matchmaker is expecting me to work miracles because of my success last week. But in week 2 that same mech is worth much less (regardless of how well I NARC), and my higher Elo is unsustainable in that mech.

Edited by Shlkt, 02 September 2014 - 08:11 AM.


#24 MrMadguy

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,199 posts

Posted 02 September 2014 - 08:38 AM

You want to know, how this system works? It's so called 50/50 winrate + ELO ranking system. It's being used in most of modern PVP games due to having some decent pros from developer point of view. How it works? It simply tries to keep your winrate close to 50/50. If you're winning too much and your winrate tends to rise - your ELO rank being boosted, so you'll start fighting against better and better players. If you're losing too much, your rank, and therefore level of skill of your enemies, starting to drop. Why it's so good from developer's point of view? Because developers don't even need to actually balance their game, because game is self-balancing in this case. Every player in such a game has a fair equal chance to win. P2W is impossible in this game, because those, who donated into this game, won't be able to roflstomp other players - they will play against other donators. For example MWO don't have rank ladder, so there is no point in getting higher rank. Dire Wolfs and TImber Wolfs are OP? They literally one-shotting you (true story)? Ok. I don't see any of them in my matches! How could it be? They all simply have their rank boosted and are playing against each other now. But this system has obvious cons from players' point of view. What cons? It's like in Matrix movie: "most people agree with rules if they have a choice, even if this choice exists in their minds only". So. The problem is in fact, that your ability to affect result of the match somehow - is only an illusion, that exist only in your mind. I.e. if in order to win one match, you have to lose one match, no matter how hard you'll try to prevent this from happening - then why should you even bother about improving your skill somehow? If you have to lose one match in order to win other one - then why should you try to win the match, you have no chances to win? Why can't you just slack, suicide or just quit one match in order to get maximum fun from other one? Yeah... That's why it against rules to do so. You have to carry. Cuz if everybody will realize this truth about 50/50 + ELO ranking system, then, like in Zeon's citizens case, it will cause catastrophic consequences to the game (Matrix). Ok. Let's imagine, that I agree with 50/50 + ELO system rules. But. There is another problem. In ideal case, this system should tend to offer you balanced matches. And it doesn't. If fails to balance matches and therefore have to use simple tactic - in order to be sure that you'll win/lose next match, it offers you a match, where you just can't lose/win. I.e. instead of 20 balanced matches it offers you 10 win rolfstomping streak + 10 lose being roflstomped streak. That's why I don't care about rules for a long time already. If I see, that I have to 100% lose in this match without any chances to win - then I just trying to lose as fast as I can and move on to the next match.

Edited by MrMadguy, 02 September 2014 - 08:54 AM.


#25 Apnu

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 2,083 posts
  • LocationMidWest

Posted 02 September 2014 - 08:44 AM

MM is a mistake. It is obviously not working.

PGI should have a lobby system and tie it to locations in CW. Done all problems solved. You can find a lobby, wait for it to fill and launch. If you have a good team and good chemistry, then you should be able to stick with them for more games. If you have a bad team, you'd be free to find another lobby. For the lazy, have a "autofind" feature that would look for a lobby mostly filled with your faction mates and off you go.

As for newbies... just make a newbie lobby. Call it whatever you want. Solaris... Training Grounds... Kentaries IV... I don't care. And only allow/force players who haven't finished getting their training bonuses to go there. When they have enough XP, they can graduate and see the public lobbies.

Come on PGI, get on this!

#26 StandingInFire

    Member

  • PipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 152 posts

Posted 02 September 2014 - 08:48 AM

Seriously can people stop saying it adds bad players to balance good players, PGI showed screenshots of the ELO system the way it works is it takes AVERAGE with LOW VARIANCE (square distance from average) meaning it groups players of about the same elo on the same team.

Even if you are amazing sometimes you will have loss streaks even if you play well at any elo, e.g. look at LoL.

View PostJohn1352, on 02 September 2014 - 03:41 AM, said:

  • If you are the predicted winner, and you win, your ELO doesn't change. Same if you're the predicted loser and you lose.


I hope thats not how they implemented it because even if the prediction is correct your elo should be adjusted proportionally inversely to the elo difference i.e. big difference and prediction is accurate small change, small difference and prediction is accurate moderate change. Otherwise elo change would be very very slow and new and bad players would take a very long time to drop elo.

EDIT:
http://mwomercs.com/...79-matchmaking/
4th Post
It does do it a bit and may adjusts ELO even if prediction is accurate, this maybe could be enlarged a bit though.

Edited by StandingInFire, 02 September 2014 - 08:59 AM.


#27 ei8ht

    Member

  • PipPipPip
  • Bridesmaid
  • 82 posts
  • Locationobscured by ECM

Posted 02 September 2014 - 09:03 AM

View PostKodiakGW, on 02 September 2014 - 07:55 AM, said:

Regarding this. Has anybody else noticed that if you leave a match before it ends (usually because your team melted and you don't want to watch a bunch of heavy and assaults chase an ECM Spider on Terra Therma), that your attempt to find a match in another mech (even one that has extremely low percentage in queue) takes MUCH longer?

That and after waiting 30-45 seconds, you hit Cancel and suddenly it finds you a match?

I'd be interested to see how many people also experience this.


Raises fist. Yes, getting this too. I could wait for a coupla minutes, then finally give up hit cancel and it suddenly it says connecting and gets you a match.

#28 Michael Abt

    Member

  • PipPipPipPipPipPip
  • The People's Hero
  • The People
  • 470 posts
  • LocationEurope

Posted 02 September 2014 - 09:23 AM

Very Interesting, it matches my own experience. Prior to last event i was +70 wins ahead, roughly a 55/45 w/l ratio. Just like the OP i streaked down, to +30 wins. What really makes me worried is that i did top damage often enough. In my HBK. No matter if i died early, or fought smart, or only had first enemy contact when score was already something like 2-12, i was top damage for my team. During some brawls i noticed that the other team also had a share of new or bad players so i was even more puzzled how on earth it was possible to have lost against them.

Anyway, my only solution so far is to play 3-5 matches, and when they are horrible losses, i am taking a break.

Edited by Michael Abt, 02 September 2014 - 09:24 AM.


#29 Duke Nedo

    Member

  • PipPipPipPipPipPipPipPipPip
  • CS 2023 Top 12 Qualifier
  • CS 2023 Top 12 Qualifier
  • 2,184 posts

Posted 02 September 2014 - 09:51 AM

View PostStandingInFire, on 02 September 2014 - 08:48 AM, said:

Seriously can people stop saying it adds bad players to balance good players, PGI showed screenshots of the ELO system the way it works is it takes AVERAGE with LOW VARIANCE (square distance from average) meaning it groups players of about the same elo on the same team.


Problem here is if they match the average elo well, that would mean that any prediction is a 50:50 guess and doesnt add anything to the equation other than only making every second game count. If the elo are not matched though, then prediction would be correct most of the time, but these matches are very unlikely to count as well since the high elo will very likely beat the low elo team... so, if it truly does this it doesn't sound very productive... atleast if elos are reasonably close, all matches should count to speed up convergence by 100%.

View PostMrMadguy, on 02 September 2014 - 08:38 AM, said:

You want to know, how this system works? It's so called 50/50 winrate + ELO ranking system.


If they actually do this they are truly ********... where do you have that information from?

Edited by Duke Nedo, 02 September 2014 - 09:52 AM.


#30 Duke Nedo

    Member

  • PipPipPipPipPipPipPipPipPip
  • CS 2023 Top 12 Qualifier
  • CS 2023 Top 12 Qualifier
  • 2,184 posts

Posted 02 September 2014 - 10:17 AM

Ok, I have had it now. 8 straight losses now and counting... and I get this???? What. The. Frock!?!?!? One trial and one that wants to lock with SRMs in my lance. Yeah right. Needless to say we lost big and I did top damage with 2 cLL in my kitfox.

PGI, FIX YOUR GAME!

Posted Image

#31 Too Much Love

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 787 posts

Posted 02 September 2014 - 10:19 AM

View PostDuke Nedo, on 02 September 2014 - 10:17 AM, said:

Ok, I have had it now. 8 straight losses now and counting... and I get this???? What. The. Frock!?!?!? One trial and one that wants to lock with SRMs in my lance. Yeah right. Needless to say we lost big and I did top damage with 2 cLL in my kitfox.

PGI, FIX YOUR GAME!


My only advice - play in the good (not any, but good) group of 6 and more people. Solo is broken right now. And will be, I suppose.

#32 Almeras

    Member

  • PipPipPipPipPipPip
  • The Slayer
  • The Slayer
  • 294 posts
  • LocationLondon

Posted 02 September 2014 - 10:20 AM

The carry ceiling per player hasn't been factored in.

Sure a single player my get 8 or even 12 kills in a game but he doesn't do it solo ever. Every mech there's a theoretical limit to its power (health and DPS). I think its unfair of the match maker to assume anyone can carry more than 2.5 mechs it sounds like a really low number but I can assure you if I placed even the best pilot 1 vs 2.5 he'd struggle with damage/heat unless he very lucky.

Its not fair on the new pilots who didn't ask to be a chew toy nor the better pilots who just want a team that know what they are doing.

#33 MrMadguy

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,199 posts

Posted 02 September 2014 - 10:22 AM

View PostDuke Nedo, on 02 September 2014 - 09:51 AM, said:

If they actually do this they are truly ********... where do you have that information from?

Most games have this system now. Just google "50% winrate MMR system". DOTA2, LOL have it, Blizzard HOTS will have it too.
Forced 50% winrate in DOTA2 for example.

Edited by MrMadguy, 02 September 2014 - 10:24 AM.


#34 Duke Nedo

    Member

  • PipPipPipPipPipPipPipPipPip
  • CS 2023 Top 12 Qualifier
  • CS 2023 Top 12 Qualifier
  • 2,184 posts

Posted 02 September 2014 - 10:35 AM

If that is true there is no point in trying really. Can it be confirmed? If so I cant think of a reason to play this game.

#35 terrycloth

    Member

  • PipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 769 posts

Posted 02 September 2014 - 11:12 AM

The idea is that you have fun when you have a closely fought battle, and not when you get to stomp newbies with your l33t skillz.

So in order for people to have fun, you have to be put in matches that you might lose (and in fact, will lose half the time).

Something related to your ELO should be visible, though, so that you can see some reward for improving your skills (since W/L can't be a good indicator and KDR is only a good indicator if it's forced to mix good people and bad people on the teams, which is suboptimal but not as bad as putting all the good people on one team).

(in reality, people have the most fun when they win easily but had the illusion that their opponents were playing well when they actually weren't... this isn't very useful for a PvP game though since being on the losing side in that situation is no fun)

Edited by terrycloth, 02 September 2014 - 11:17 AM.


#36 Clit Beastwood

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,262 posts
  • LocationSouthern California

Posted 02 September 2014 - 11:20 AM

What I've observed is the following

- Pre-clans, when pugging it was more accepted by individual players that they needed their team in order to survive.
- Post-clans, players will grab a dwf or tbr and head off to be a hero, totally neglecting their team. KDR's go up, WLR's go down.

#37 El Bandito

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Big Daddy
  • Big Daddy
  • 26,736 posts
  • LocationStill doing ungodly amount of damage, but with more accuracy.

Posted 02 September 2014 - 04:13 PM

View PostFaith McCarron, on 02 September 2014 - 04:34 AM, said:

I think the best example of this was Jager's stream last week during the final day of the 50 win challenge. He was solo pugging on the stream to get the last 25 wins, and it was painful to say the least. He was carrying as hard as ever, dropping in good builds, all of that, and yet he was regularly running loss streaks of 5-10 games. In a normal night you figure it would take Jager what, 30-35 drops at most to get 25 wins? I dont know how many it ended up taking, but it was well over 50. It was just giving him garbage teams, no matter how many kills, how much damage he did, how many openings he made for the team, how many enemies he tied up, his teams were just so terribly bad that a win was simply impossible.


Damn, that just makes me feel like a scrub who got carried by his teammates. :P

How else could I get positive WLR in the 50 win challenge otherwise?

#38 KHETTI

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,328 posts
  • LocationIn transit to 1 of 4 possible planets

Posted 02 September 2014 - 05:08 PM

Solo dropping with the current Match making is pointless, total waste of time.
I suspect the MM needs a healthy player base to even come close to working, which sadly this game hasn't got, and which is most likely slowly shrinking.
MWO and "Fun" don't belong in the same sentence now, poor performance, buggy everything, horrible Ui, poor balancing and garbage match making.
Pretty much sums up MWO right now, but then again i'm just a random smuck stuck on an island.

#39 occusoj

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • 452 posts

Posted 02 September 2014 - 05:45 PM

Quote

LOW VARIANCE (square distance from average)

Please correct me If Im wrong but I seem to remember that somewhere it was stated that PGIs "variance" isnt what one would expect variance to be. Its more or less just max-min. Again, I could be totally off.

However, MM is going totally crazy these days. Had games were one team got 2-4 new players in trial mechs and the other was given guys from Lords, SJR and the like. That were some of the fastest stomps Ive ever been in.

So either this system is totally screwed or the player base got too small to provide a suiteable pool of players causing it not to open release valves but damn floodgates.

#40 John1352

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,025 posts
  • LocationConnecting....

Posted 02 September 2014 - 06:11 PM

View PostAlmeras, on 02 September 2014 - 10:20 AM, said:

The carry ceiling per player hasn't been factored in.

Sure a single player my get 8 or even 12 kills in a game but he doesn't do it solo ever. Every mech there's a theoretical limit to its power (health and DPS). I think its unfair of the match maker to assume anyone can carry more than 2.5 mechs it sounds like a really low number but I can assure you if I placed even the best pilot 1 vs 2.5 he'd struggle with damage/heat unless he very lucky.

Its not fair on the new pilots who didn't ask to be a chew toy nor the better pilots who just want a team that know what they are doing.

This could well be the problem. Short of maybe PPFLD or pulse lasers into a side torso or cockpit, fighting 2 v 1 just doesn't work, you get too hot trying to kill both enemies and take too much damage. Usually one person struggles to get multiple kills without an ally distracting the enemies, the only common exception is when fighting damaged mechs in the endgame.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users