#1
Posted 05 April 2014 - 03:03 AM
All of a sudden the MM decides I've won too much and puts our 4-man with 8 complete newbs. No amount of carrying is enough, we get smashed.
Rinse and repeat.
I'm lucky to get 1-2 games out of ten that resemble anything like a good fight.
I DON'T WANT TO PLAY WITH OR AGAINST NEWBS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
I've been playing over a year and consider myself a good pilot. I should not have to put up with casuals, newbs, etc (no offense) in my games. It's BS.
There should be tiers: Soldier, Veteran, Elite, to separate players.
Don't give me "there aren't enough people playing", if the community is that small is time to close shop.
I just want good fights with/against people of a similar skill level as me.
#2
Posted 05 April 2014 - 03:07 AM
If you are in a certain ELO level you get all the newbs, because they start with pretty much the same ELO level.
#3
Posted 05 April 2014 - 03:14 AM
Klappspaten, on 05 April 2014 - 03:07 AM, said:
If you are in a certain ELO level you get all the newbs, because they start with pretty much the same ELO level.
ELO never worked in Team Games, in the end, you rate people based on the perfomance of 23 OTHER people....
#4
Posted 05 April 2014 - 03:28 AM
Androas, on 05 April 2014 - 03:14 AM, said:
Yes it does - it just takes a lot longer to get to your proper Elo.
With law of large numbers, the other 23 people will eventually average out, leaving only one common denominator.
And - by the way Elo works - new players by definition start in the middle. (though I believe during the cadet bonus they get a 200 point Elo nerf that falls away after game 25)
#5
Posted 05 April 2014 - 03:35 AM
#6
Posted 05 April 2014 - 03:39 AM
#7
Posted 05 April 2014 - 03:43 AM
#8
Posted 05 April 2014 - 04:18 AM
Papaspud, on 05 April 2014 - 03:39 AM, said:
Really - on a match to match basis - a better indicator is how you do in comparison to the rest of your team.
If I pull out 300 damage and only one other person on the team broke 200, I feel pretty good about the match even if we did get rolled. (though I don't think stomps happen nearly as often as one would think by reading the boards )
(conversely, if I overextended early and end up with 80 damage I feel lame, even if my team ends up winning)
Oh - and a pet peeve of mine is people who hide in a corner for the last 2-5 min when they know they're gonna lose, just to protect their precious K/D. Gah!
Edited by Charons Little Helper, 05 April 2014 - 04:19 AM.
#9
Posted 05 April 2014 - 04:20 AM
#11
Posted 05 April 2014 - 04:22 AM
#12
Posted 05 April 2014 - 04:43 AM
Charons Little Helper, on 05 April 2014 - 03:28 AM, said:
With law of large numbers, the other 23 people will eventually average out, leaving only one common denominator.
...
Assuming that the skill levels of individual players don't fluctuate too wildly fo course. Which they do.
Your own performance gets drowned in all that static. The resulting ELO is merely an indicator of average team performance, not singular performance.
Once you've played a couple thousand games, you just realize that your individual performance depends much more on the other players than you may like to admit. If you play against vets, chances are that you don't have a chance to shine. If you play against newbies however ...
Likewise, whether you are teamed with vets or newbies also makes a huge difference regarding your performance. You may not even be able to get more than one shot on target before it explodes, simply because the vets are practicing focus fire. With newbies around, there is more opportunity to deal damage and score kills because they lack cohesion and experience. By the same token you are also much more likely to lose when facing vets, regardless of your skill and performance.
Long story short: ELO works in 1vs1 games. Outside of that it fails, rather predictably.
#13
Posted 05 April 2014 - 04:51 AM
CCC Dober, on 05 April 2014 - 04:43 AM, said:
Wrong. And more importantly, mathematically proven to be wrong.
Elo works just fine in multiplayer games, and has been repeatedly proven to do so. It just takes longer for your rating to stabilize.
The problem with MWO is that the matchmaker uses too wide of a rating spread when creating matches, which is necessary to create matches in a reasonable amount of time because the player base is so small.
#14
Posted 05 April 2014 - 04:58 AM
Roadkill, on 05 April 2014 - 04:51 AM, said:
Elo works just fine in multiplayer games, and has been repeatedly proven to do so. It just takes longer for your rating to stabilize.
The problem with MWO is that the matchmaker uses too wide of a rating spread when creating matches, which is necessary to create matches in a reasonable amount of time because the player base is so small.
Of course it can work in multiplayer games, if you consider 1vs1 a multiplayer game and leave it at that.
I'm curious what kind of assumptions were made to "prove" that ELO works past that point. Don't be shy, please do share with us.
#15
Posted 05 April 2014 - 04:58 AM
CCC Dober, on 05 April 2014 - 04:43 AM, said:
Your own performance gets drowned in all that static. The resulting ELO is merely an indicator of average team performance, not singular performance.
Wrong.
They skill levels of individual players do fluctuate.
However, eventually they will average out. With enough games, your average opponent will be equivelent to your average teammate. You are the only common denominator.
It's like how casinos work. In the short term table games like roulette and craps seem totally random. That you can get lucky and come out ahead. And in the short term, you can. But eventually the house always wins.
#16
Posted 05 April 2014 - 05:05 AM
The combination of team average ELO, custom mech builds, and a low player population make it difficult to have really well balanced matches.
The launch module and 3/3/3/3 will help some but it seems to me like a temporary fix for a more complicated problem.
#17
Posted 05 April 2014 - 05:09 AM
Charons Little Helper, on 05 April 2014 - 04:58 AM, said:
Wrong.
They skill levels of individual players do fluctuate.
However, eventually they will average out. With enough games, your average opponent will be equivelent to your average teammate. You are the only common denominator.
It's like how casinos work. In the short term table games like roulette and craps seem totally random. That you can get lucky and come out ahead. And in the short term, you can. But eventually the house always wins.
Which is what I wrote. Read more closely next time.
The problem with you suggesting that MWO players average out like dice in a casino or roulette is a pretty bold claim.
They do not and with a bit of thinking you can come to that conclusion rather quickly. Each player develops differently. You don't meet the same players and you don't have the same team compositions and maps every time you play. With dice or a roulette table that's not an issue, because the playing pieces and their environment stay the same.
In MWO there is no basis for any of that to average out. Not even the players that participate handle the same input devices or play by the exact same rules. None of that is visible, but it has an impact regardless. It's utter chaos to put it lightly. Try to average that out. I guess if nothing else, it makes for good entertainment, heh.
#18
Posted 05 April 2014 - 10:06 AM
Edited by MonkeyDCecil, 05 April 2014 - 10:06 AM.
#20
Posted 05 April 2014 - 10:56 AM
MonkeyDCecil, on 05 April 2014 - 10:06 AM, said:
It's the nature of Elo unfortunately. They do get a slight nerf during the Cadet period. One could argue to increase and/or extend it, but their base Elo has to start there.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users