Jump to content

Impossible Odds, Time After Time


117 replies to this topic

#61 Ubikjam

    Rookie

  • 8 posts

Posted 10 August 2013 - 10:18 AM

Personally I haven't found stomps to be as regular after 12v12, (which makes sense because the probability of having uneven teams gets mathmatically smaller the bigger the sides).

I would still say the major problem in creating walkovers is a lack of cohesion/tactics - that really is a probem that sucks the fun away. It's also vhard to fix...

#62 Majorfatboy

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 623 posts
  • LocationBound and gagged on The Island

Posted 10 August 2013 - 10:20 AM

OP, I feel your pain. I was just complaining about this very thing in a different thread, only to get the "Working as intended" line from a white knight.

I pug mostly, and get plenty of stomps, but gets worse if I'm with My buddies. Then the matchmaker actively balances matches against us.
Our lance: Me in a ac5/med laser jager, new guy in a stalker, Buddie #1 in an lrm catapult, Buddie #2 in a spotter spider.
Bravo lance: ALL BRAND SPANKING NEW PLAYERS IN TRIAL MECHS

Opposing force: All custom-tailored mechs driven by relatively experienced players.

We get stomped, they win. Happens. All. The. Time. I've been getting My butt kicked for the past two days now, and what few buddies I have in this game regularly give up and just run lights and fast-cap the other guy's base, just so they can actually win a match for once.

#63 kuangmk11

    Member

  • PipPipPipPipPipPipPip
  • Veteran Founder
  • Veteran Founder
  • 627 posts
  • LocationW-SEA, Cascadia

Posted 10 August 2013 - 10:49 AM

My take is that now with 12v12 the matchmaker is trying to find 24 players the instant you click launch instead of 16, so the acceptable skill gap widened. The matchmaker needs to take some more time to build better matches.

I seem to get put on a team with "first-timers" in trial mechs every match (A1's asking where their lasers are) . Is MWO experiencing a massive influx of new players? Instead of planning a strat I end up explaining basic game mechanics. I thought new players suppose to only be put against other new players for 100 matches of something?

Match after match I get teams that make a lot of noise but get no results. I check the score screen to see how we are doing and all the kills we have are my own. I have had a couple really good, fun matches but they are WAY too far between.

I can't trust my team to pull any weight so my new tactic is to use them as a meat shield. I hang back watching them fire lasers at the sky, TK each other, and walk in circles and while the enemy team rips them to shreds I can get a few surprise kills in.

Edited by kuangmk11, 10 August 2013 - 10:49 AM.


#64 Appogee

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 10,967 posts
  • LocationOn planet Tukayyid, celebrating victory

Posted 10 August 2013 - 10:53 AM

View PostDeaconW, on 10 August 2013 - 06:23 AM, said:

Um...do you even understand the contradiction in your statement? You want "Similarly skilled players", "reasonably even odds of winning or losing", but the 50/50 thing has got to go? Dude, you just ASKED FOR the 50/50 thing.


No I didn't. The difference is between making 50:50 a goal - which the matchmaker relentlessly tries to achieve by creating impossible matchups across players of quite differing skill levels - vs just creating matches between teams comprised of similarly skilled players... which incidentally would likely end up for many as a 50:50 W/L.

We are not currently having games with similarly skilled players. That's the problem.

#65 Sug

    Member

  • PipPipPipPipPipPipPipPipPip
  • The People's Hero
  • The People
  • 4,630 posts
  • LocationChicago

Posted 10 August 2013 - 11:02 AM

View Postkuangmk11, on 10 August 2013 - 10:49 AM, said:

I seem to get put on a team with "first-timers" in trial mechs every match (A1's asking where their lasers are) . Is MWO experiencing a massive influx of new players? Instead of planning a strat I end up explaining basic game mechanics. I thought new players suppose to only be put against other new players for 100 matches of something?


New players get a New Player Elo score for their first 25 matches along with a Standard Elo score. For the first 25 matches they are matched based on their NP Elo but any change in score also effects the Standard Elo.

After 25 matches the NP Elo disappears and they are placed in matches based on their Standard Elo.

Unfortunately the original Standard Elo score they get is roughly the AVERAGE Elo score among ALL players. And since Elo changes very slowly.....and most players are within a standard deviation of the average score......yeah. You end up with people in trial mechs with 26 matches played put on teams of veterans.


View Postkuangmk11, on 10 August 2013 - 10:49 AM, said:

I can't trust my team to pull any weight so my new tactic is to use them as a meat shield. I hang back watching them fire lasers at the sky, TK each other, and walk in circles and while the enemy team rips them to shreds I can get a few surprise kills in.


Pretty much it. My new tactic is to not do anything until the 10min mark. After all the nubs rushed and the snipers are all dead or soften up I'll do a bit of flanking and blindside the other team. So far it's been a very effective strat for a 100% solo pugger.

#66 DeaconW

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 976 posts

Posted 10 August 2013 - 11:07 AM

View PostAppogee, on 10 August 2013 - 10:53 AM, said:


No I didn't. The difference is between making 50:50 a goal - which the matchmaker relentlessly tries to achieve by creating impossible matchups across players of quite differing skill levels - vs just creating matches between teams comprised of similarly skilled players... which incidentally would likely end up for many as a 50:50 W/L.

We are not currently having games with similarly skilled players. That's the problem.


I agree with you about the problem, but you are still not getting why your statement was non-sensical. By saying "I want reasonably even odds" you are, by definition, asking for 50/50 results or close to it. They are logically and mathematically equivalent statements. I would like "reasonably even odds" as well...I just disagree with PGI's way of doing it...

#67 ztac

    Member

  • PipPipPipPipPipPipPip
  • 624 posts

Posted 10 August 2013 - 11:16 AM

Well don't count on anything serious being done, the proposed fixes are ludicrous at best. Despite PGI saying they are aware of the problem , except they will not do what has to be done, got to keep certain people happy you know!

#68 DeaconW

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 976 posts

Posted 10 August 2013 - 11:20 AM

View Postztac, on 10 August 2013 - 11:16 AM, said:

Well don't count on anything serious being done, the proposed fixes are ludicrous at best. Despite PGI saying they are aware of the problem , except they will not do what has to be done, got to keep certain people happy you know!


What is this "certain people list" you speak of. And how do I get on it. :)

#69 Zolaz

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 3,510 posts
  • LocationHouston, Tx

Posted 10 August 2013 - 11:59 AM

Just had a match that had 6 Assaults vs 3 Assaults. But at least we had more medium mechs than they had Heavies. Match Maker sucks in MWO.

#70 Royalewithcheese

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,342 posts

Posted 10 August 2013 - 12:05 PM

Posted Image

#71 Angel of Annihilation

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,881 posts

Posted 10 August 2013 - 12:06 PM

I got dropped in a 6 vs 12 match the other night. Naturally I was on the side with 6 hehe.

My explaination was that us 6 must have been ELO Gods and the only way the game could balance us out was to give us 2 to 1 odds....or at least that is the story I am sticking with <wink>.

On a positive note, I did manage to kill one of the enemy before dying, though I was the only one on my team who got a kill hehe.

#72 Appogee

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 10,967 posts
  • LocationOn planet Tukayyid, celebrating victory

Posted 10 August 2013 - 12:43 PM

View PostViktor Drake, on 10 August 2013 - 12:06 PM, said:

I got dropped in a 6 vs 12 match the other night. Naturally I was on the side with 6 hehe...
And I thought I was getting a raw deal... :)

#73 Hauser

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 976 posts

Posted 10 August 2013 - 02:35 PM

View PostDeaconW, on 10 August 2013 - 08:49 AM, said:

If you really want to think about what is wrong with using ELO for MWO, think about this: ELO was designed for rating individual players in a one-on-one game with a fixed board and fixed piece capabilities. It is being used by PGI on a multiplayer game with a constantly changing "board"(different maps) and constantly changing "pieces"(different mechs, different weapons on mechs). Just doesn't sound right when you look at it that way, does it?


Think some more. :)

You're constantly playing with and against different players so in the long run the influences of each individual cancel each other out. You however are the only constant in all the teams you are in (duh). So the result of a match can be used, even in a team game.

To understand why in the long run your team mates cancel out try to imagine all possible matchups, each and every one of them. There will be some where your team is just plain better, there are some where the other team is plain better. There will be some where your team is slightly better, there will be some where your team is slightly worse. If you are a good player there are more combinations of matchups where your team has the advantage then when you're a bad player.

Now the match maker already preselects those combinations of matches where you'll have teams that are only slightly different. So if your true rating is much higher your team will actually be much better and you'll win those games. This will send your Elo up until the match maker has estimated your correctly. It also works the other way around when you're are say **** drunk and not living up to your rating.

View PostDeaconW, on 10 August 2013 - 09:03 AM, said:

But the truth is that wins that are accomplished by actively cheating or simply being given to you by the MM aren't "real" wins.


I'm sorry. But that is not what the match maker is trying to do. No wins are handed out.


View PostSug, on 10 August 2013 - 11:02 AM, said:

New players get a New Player Elo score for their first 25 matches along with a Standard Elo score. For the first 25 matches they are matched based on their NP Elo but any change in score also effects the Standard Elo.

After 25 matches the NP Elo disappears and they are placed in matches based on their Standard Elo.

Unfortunately the original Standard Elo score they get is roughly the AVERAGE Elo score among ALL players. And since Elo changes very slowly.....and most players are within a standard deviation of the average score......yeah. You end up with people in trial mechs with 26 matches played put on teams of veterans.


Actually when Elo changes it changes with at least 25 points at the time. So actual players S-Elo will be identical to his N-Elo in 8 losses. All those losses are unexpected because the match maker puts him in a match with N-Elo players but calculates with his S-Elo. He is expected to win but fails to do so.

#74 DeaconW

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 976 posts

Posted 10 August 2013 - 03:30 PM

View PostHauser, on 10 August 2013 - 02:35 PM, said:


Think some more. :)

You're constantly playing with and against different players so in the long run the influences of each individual cancel each other out. You however are the only constant in all the teams you are in (duh). So the result of a match can be used, even in a team game.

To understand why in the long run your team mates cancel out try to imagine all possible matchups, each and every one of them. There will be some where your team is just plain better, there are some where the other team is plain better. There will be some where your team is slightly better, there will be some where your team is slightly worse. If you are a good player there are more combinations of matchups where your team has the advantage then when you're a bad player.

Now the match maker already preselects those combinations of matches where you'll have teams that are only slightly different. So if your true rating is much higher your team will actually be much better and you'll win those games. This will send your Elo up until the match maker has estimated your correctly. It also works the other way around when you're are say **** drunk and not living up to your rating.


Even if I spot you that "multiple players doesn't matter because it evens out"...you still have the changing mapboard and pieces.

Besides, the matchmaking isn't random...I can pick 2,3,4 or 12 constants on my side if I wish. Play with 3 other people I know and trust "inflates" my ELO but going back to solo pugging my ELO is "deflated". Like I said, IMO there are too many variables to make this type of ELO system work.

Quote

I'm sorry. But that is not what the match maker is trying to do. No wins are handed out.


I didn't say that is what it was *trying to do*...I was stating an unintended consequence of the MM's failure. I don't feel it is a true win either, when I get the occasional 12-0 on the win side.


Quote

Actually when Elo changes it changes with at least 25 points at the time. So actual players S-Elo will be identical to his N-Elo in 8 losses. All those losses are unexpected because the match maker puts him in a match with N-Elo players but calculates with his S-Elo. He is expected to win but fails to do so.


You know the way you phrased this reinforces my "ELO gives people wins" statement, right? :D

#75 Steel Claws

    Member

  • PipPipPipPipPipPipPip
  • The Clan Cat
  • The Clan Cat
  • 665 posts
  • LocationKansas

Posted 10 August 2013 - 03:32 PM

If the weight limits were more balanced I'm fine with ELO.

In all seriousness it isn't the match maker that is most at fault and it isn't pugging. It's the people. Almost every drop I loose it's because a number of the team either won't pay attension to what is going on, go off chasing light mechs, don't go to where the enemy is, or just rush in foolishly.

Take for instance yesterday, I typed in that there were 4 mechs in a location on Terra Therma, and three mechs came over and helped me kill a Dragon - then they just left - leaving me to face three more mechs alone - despite my typing in that there were more there repeatedly. End result was they got surrounded and taken down.

I continue to see teams on River City going the opposite way that the other team is going (even when it's been called out where they are going and that they are pushing) and then stopping and milling around in confusion for several minutes while the rest of their team is killed trying to defend the base. Then outnumbered they get rolled up and wonder what went wrong (must be those nasty premades!).

Then there are those wonderful times we have found the enemy and have a line built to stop them and face them down when suddenly a light pops up behind us and half to three quarters of the team turns and tries to chase it down. Then the enemy pushes are weakened line and we get rolled.

Then there are those that think they can just rush over a ridge into the middle of the enemy team and fight a one on one with a mech and that his team mates will just set there and let them do this. Then they wonder why they died so fast (It's those PPCs and gauss I tell you!)

I submit that the issue isnt the match maker. The problem is the dumb things people do. I've seen teams with only 1 assault and 2 heavies take out a 5 assaults and 3 heavies 8:1 simply because they played smart (and were all pugs). If people play smart (don't go running off) and pay attension to what is going on and the big picture - there wouldn't BE any "impossible odds." Stop running off and leaving your team mates hanging when a light gets on base or there is a report of something in the tunnel. Stay on the line and fight instead of trying to chase that Jenner in your Atlas or Stalker.

Yes you can and will get steam rolled on occation but for the most part all of the lop sided victories I've seen were the result of foolish play on the loosing teams part.

Edited by Steel Claws, 10 August 2013 - 03:36 PM.


#76 DeaconW

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 976 posts

Posted 10 August 2013 - 03:35 PM

View PostSteel Claws, on 10 August 2013 - 03:32 PM, said:

Yes you can and will get steam rolled on occation but for the most part all of the lop sided victories I've seen were the result of foolish play on the loosing teams part.


But that IS the MM...it decided those people were close enough to your skill level to play with you.

#77 Hauser

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 976 posts

Posted 10 August 2013 - 03:38 PM

View PostDeaconW, on 10 August 2013 - 03:30 PM, said:

Even if I spot you that "multiple players doesn't matter because it evens out"...you still have the changing mapboard and pieces.


Those are just more permutations.

View PostDeaconW, on 10 August 2013 - 03:30 PM, said:

Besides, the matchmaking isn't random...I can pick 2,3,4 or 12 constants on my side if I wish. Play with 3 other people I know and trust "inflates" my ELO but going back to solo pugging my ELO is "deflated". Like I said, IMO there are too many variables to make this type of ELO system work.


There are many variables but Elo only looks at three. The outcome of the match, your teams Elo and your opponents Elo. The Elo ratings are under control of the algorithm so the only free variable is the match out come. Anything you do that influences the outcome will be reflected in your Elo rating. I can't think of a more effective system really.

View PostDeaconW, on 10 August 2013 - 03:30 PM, said:

You know the way you phrased this reinforces my "ELO gives people wins" statement, right? :)


Only if the new player isn't actually a new player but that won't last because his N-Elo will go up pretty fast.

Edited by Hauser, 10 August 2013 - 03:53 PM.


#78 KHETTI

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,329 posts
  • LocationIn transit to 1 of 4 possible planets

Posted 10 August 2013 - 03:41 PM

The problem right now in 12v12 is largely due to no tonnage restrictions per team, all the scrubs are running assaults and just blobbing.
70% of every team, every game is made up of assaults, there should be 2 assaults, 2 heavies, 2 lights and the rest mediums per team.

#79 Steel Claws

    Member

  • PipPipPipPipPipPipPip
  • The Clan Cat
  • The Clan Cat
  • 665 posts
  • LocationKansas

Posted 10 August 2013 - 03:45 PM

View PostDeaconW, on 10 August 2013 - 03:35 PM, said:


But that IS the MM...it decided those people were close enough to your skill level to play with you.


Actually match maker tries to average things out. To do this it will stick good players with bad. This isn't a bad thing. Then there is the whole queue thing. If the queue isn't very full you get who ever is available. The idea behind match maker isn't bad it just needs to have a bit of weight balancing thrown in for good measure but ultimtely it all comes down to who is in the queue when you drop and what they do during the drop - and this is something that simply cannot be coded.

Edited by Steel Claws, 10 August 2013 - 03:47 PM.


#80 Hauser

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 976 posts

Posted 10 August 2013 - 04:00 PM

View PostSteel Claws, on 10 August 2013 - 03:45 PM, said:

Actually match maker tries to average things out. To do this it will stick good players with bad.


I'm afraid you are mistaken.

http://mwomercs.com/...-making-update/

Quote

How does the match maker compose a teams Elo rating, is it average rating or closest to a target?

It's closest to a target value, so the match maker starts trying to make a match for an Elo of say 1300 and will pull in players to those teams closest to those values; however, as mentioned earlier within growing thresholds and those curves will be tuned. Currently it may be a bit 'sloppy' about how it's filling those buckets but over time it will be tuned to be much more precise.

We need to do this carefully over time as generally the cost of precision is time to find a match we want to slowly find a very nice balance between time to find a match and the number of matches that are correctly composed.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users