Jump to content

Elo Question About Light Mechs


12 replies to this topic

#1 jeffsw6

    Member

  • PipPipPipPipPipPipPipPip
  • 1,258 posts
  • LocationLouisville, KY (suburbs)

Posted 03 May 2013 - 11:26 AM

It seems like there are far, far more light mechs in the games when I play a light than when I play an assault. My kill/death radio in lights must be about 0.2 but my Win/Loss is a little better than 1.

Does the current ELO system effectively dump lights who capture, or otherwise don't pursue kills, into a trash bracket with lots of other light mechs (also getting limited kills) plus newbies? Because that is what it seems like.

Playing lights in PUGs is making me want to sell all my lights and never play them again.

#2 NinetyProof

    Member

  • PipPipPipPipPipPipPip
  • 547 posts
  • LocationSan Diego, CA

Posted 03 May 2013 - 11:58 AM

I do not believe the current ELO system does anything other then "Win / Loss". I don't think it tracks / looks at K/D Ratio, Objectives, etc, etc, etc.

Indeed, the purpose of ELO is to predict whether you will win or lose the game, and set the value you either gain or lose.

That's it ... it's not about personal performance, it's about team performance ... at least that is my understanding of it.

#3 One Medic Army

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 5,985 posts
  • LocationBay Area, California

Posted 03 May 2013 - 12:11 PM

I'm pretty sure that the current Elo score is also based at least partially on "match score"

I've noticed that my personal Elo scores seem to be Assault>Light>Medium judging by the ability and loadout of my companions and enemies.

My personal thought is that since it's harder to get a high "match score" in light mechs due to it being generally difficult to get damage/kills/assists against good/coordinated players that lights tend to have a lower Elo score given the same piloting skill.

PGI, of course, knows the truth of this, which weight classes tend to have which Elo scores.

Edited by One Medic Army, 03 May 2013 - 12:12 PM.


#4 Lupin

    Member

  • PipPipPipPipPipPipPip
  • Elite Founder
  • 955 posts
  • LocationKent, UK.

Posted 03 May 2013 - 01:14 PM

It would also depend on how many games you have played in your light vs. assault mech.
As I seem to remember ELO scores different for each class of mech.

#5 One Medic Army

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 5,985 posts
  • LocationBay Area, California

Posted 03 May 2013 - 01:20 PM

View PostLupin, on 03 May 2013 - 01:14 PM, said:

It would also depend on how many games you have played in your light vs. assault mech.
As I seem to remember ELO scores different for each class of mech.

If you're referring to me, I have far more games played in lights than assaults, but I've got enough matches in both that I can be pretty sure my Elo has settled, especially when you consider that their Elo metric includes data all the way back to the first tournament.

#6 marabou

    Member

  • PipPipPipPipPipPip
  • Warrior - Point 1
  • Warrior - Point 1
  • 370 posts

Posted 03 May 2013 - 01:32 PM

ELO is a lie.

#7 Traigus

    Member

  • PipPipPipPipPipPip
  • 303 posts

Posted 03 May 2013 - 03:05 PM

The last t thing I saw said we had a different ELO per weight class.

I know when I play my lights or heavies, I usually get easier matches then when I play my mediums.

They also turned up weight matching supposedly the week after last patch (and adjusted it up and down) So if you drive a light there should be a light on the other team. If you don't there may be none. I don't think your ELO matches the other light pilot, only some other pilot on the other team.

Since lights tend to get less of every reward then win/loss, it would be expected they would ELO lower than other types if anything but win/loss was a factor.

I wish PGI would explain it better.

#8 Lupin

    Member

  • PipPipPipPipPipPipPip
  • Elite Founder
  • 955 posts
  • LocationKent, UK.

Posted 03 May 2013 - 10:36 PM

View PostOne Medic Army, on 03 May 2013 - 01:20 PM, said:

If you're referring to me, I have far more games played in lights than assaults, but I've got enough matches in both that I can be pretty sure my Elo has settled, especially when you consider that their Elo metric includes data all the way back to the first tournament.


But did they not do an adjustment recently to level some of ELO out?

#9 hammerreborn

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,063 posts
  • LocationAlexandria, VA

Posted 03 May 2013 - 10:40 PM

View PostLupin, on 03 May 2013 - 10:36 PM, said:


But did they not do an adjustment recently to level some of ELO out?


Yes, it was heavily weighted towards games played. They fixed the numbers and reran them and now they are less silly.

The reason you see more lights when you're in a light is a) the tourney that just happened involved all classes not just the heavies and assaults as they have been lately, and b ) better ELO matching (due to the adjustment) means better tonnage matching, so you being in a light increases the chances of seeing an enemy light.


Also, there's the whole separate ELO for each class and if your Assault is high enough, everything at the higher levels of ELO is a poptard fest, so there are very few lights running around willing to get one shot for no rewards whatsoever, while in the lower ELO brackets you have new players that don't realize how crappy it's going to be as they get "better" and the game becomes progressively worse because of it.

Edited by hammerreborn, 03 May 2013 - 10:42 PM.


#10 jeffsw6

    Member

  • PipPipPipPipPipPipPipPip
  • 1,258 posts
  • LocationLouisville, KY (suburbs)

Posted 04 May 2013 - 09:58 AM

View Posthammerreborn, on 03 May 2013 - 10:40 PM, said:

better ELO matching (due to the adjustment) means better tonnage matching, so you being in a light increases the chances of seeing an enemy light.

This makes sense, but in my light mechs, I am often dropping into a team of 4 - 6 light/medium mechs. I understand why the chance of me having an opponent light-mech, because I am playing one, should be higher; but not why both teams seem to have many more lights than I see when piloting Assault mechs.

I really think this is a fault with the ELO/MM system that is causing lights to score worse, and be in low-ELO matches.

#11 hammerreborn

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,063 posts
  • LocationAlexandria, VA

Posted 04 May 2013 - 10:05 AM

View Postjeffsw6, on 04 May 2013 - 09:58 AM, said:

This makes sense, but in my light mechs, I am often dropping into a team of 4 - 6 light/medium mechs. I understand why the chance of me having an opponent light-mech, because I am playing one, should be higher; but not why both teams seem to have many more lights than I see when piloting Assault mechs.

I really think this is a fault with the ELO/MM system that is causing lights to score worse, and be in low-ELO matches.


Like I said, tourney brought out a bunch of lights. Also elo is based on win/loss only. So unless most light pilots are losers then it doesn't have anything to do with Elo.

#12 Thorqemada

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 6,385 posts

Posted 04 May 2013 - 10:44 AM

May it be that some premades use a Lance of Light Mechs and as the matchmaker tries to get a close weightmatching if possible you are thrown into these matches?

#13 One Medic Army

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 5,985 posts
  • LocationBay Area, California

Posted 04 May 2013 - 02:36 PM

View Posthammerreborn, on 04 May 2013 - 10:05 AM, said:

Like I said, tourney brought out a bunch of lights. Also elo is based on win/loss only. So unless most light pilots are losers then it doesn't have anything to do with Elo.

Standard Elo is, team-based modified Elo as used here and in several other games, is not.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users