Jump to content

Elo Vs Battle Value


48 replies to this topic

#1 Diablobo

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,014 posts
  • LocationOn your six

Posted 22 February 2013 - 12:17 AM

Why does PGI think that ELO matchmaking is going to work in MWO better than the Battle Value system of Battletech? It seems to me that when a game system has such a huge variety in the equipment that a player can bring into the game, then it is even more necessary to compare equipment along with pilot skills. In what other ELO rated game engines do the configuration of a players' equipment vary so much? If the ELO system does not factor equipment configurations in a game where it makes such a big difference, then that seems to be a huge flaw in the matchmaker, and needs to at least take tonnages into consideration.

#2 Zaptruder

    Member

  • PipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 716 posts

Posted 22 February 2013 - 12:25 AM

Elo is a system that ignores the noise in favour of the signal.

A thousand factors contribute towards you winning.

You could be a team communicator, run in a good group, be a good marksman, run OP mechs, know all the strategies, etc, etc, etc.

All those things contribute towards how likely you are to win. So... why not just measure how often you win instead of trying to sort out through thousands of variables?

The only concession they've made is the use of class player based Elo ratings rather than a flat player Elo rating. If you're better in your assault mechs, then you'll fight tougher opponents... likely opponents that are better in their assault mechs too.

But if your Elo with a light is 2000... it's because you punch at the same level as another player running an assault with an Assault Elo of 2000. Maybe that means your Assault Elo is 1500; you specialize in lights. Or your Assault Elo is 2300; You're a really good player, and your Lights aren't your best class.


If you ever see a player in top tier leagues with a main of a Spider 5k... you'll know that he's utterly amazing with machine guns.

#3 p00k

    Member

  • PipPipPipPipPipPipPipPip
  • 1,661 posts

Posted 22 February 2013 - 12:28 AM

battle value doesn't take into account ability and experience

elo may not directly take into account mech builds, but it's not unreasonable to assume that someone who wins a lot has a decent mech build. the reverse is not true.

battle value is garbage. is a raven really only a fraction of the usefulness of an atlas? of course not. are mechs a sum of their parts (and thus their parts' bvs)? of course not. battle value is awful, was awful in battletech, but less egregiously so because you rolled die

#4 Dukarriope

    Member

  • PipPipPipPipPipPipPip
  • 923 posts
  • Twitter: Link
  • Locationa creative suite

Posted 22 February 2013 - 12:35 AM

You know what I think?
Take something like Battle Value. Make elo a factor of said Battle Value.

#5 Vassago Rain

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 14,396 posts
  • LocationExodus fleet, HMS Kong Circumflex accent

Posted 22 February 2013 - 12:38 AM

Why do you think battlevalue would work? That's a much better question.

#6 OneEyed Jack

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,500 posts

Posted 22 February 2013 - 12:52 AM

BV would be a useful in limiting factor for teams in premade queues, but that's the only place I see it being a good thing.

#7 Loken73

    Member

  • PipPipPip
  • Bad Company
  • Bad Company
  • 88 posts
  • LocationPoland

Posted 22 February 2013 - 01:01 AM

OP have you tried to build and use swayback in TT? Its bv will be very low, but in MWO its very useful. Same with Raven 3L and other ECMed mechs. In MWO they are very useful, sometimes decimating enemy teams. I cannot imagine same situation in TT. http://www.masterunitlist.info/
AS-7D is 1897bv and Raven 3L is 708bv. You need almost 3 Ravens to counter Atlas. And in TT it makes sense, in MWO not.
So, no, bv will not be good in MWO.

#8 Rocket Puppy

    Member

  • PipPipPip
  • Overlord
  • 97 posts

Posted 22 February 2013 - 01:04 AM

ELO is a ranking system that has been proven effective for 50+ years, albeit with some modification in situations.

Battle Value has been proven to be exploitable in its couple decades of existence.

Neither is fool proof, but an ELO system with enough variables taken into consideration is an extremely good method of ranking players.

If you solo queue you are going to hit hard times, playing in a group you might hit a wall.

If you are losing consistently you need to find a clan or a group of players to chill with, if you have a ranking high enough to hit the pre-made player gap that should not be a problem.

I'm not going to tell you what to do, but if you hit the top 1/3 of players it is probably time to reach out to your peers and join an active community. If your skill level is that high finding a group of people willing to play with you should not be a problem.

Edited by Rocket Puppy, 22 February 2013 - 01:05 AM.


#9 Ramrod

    Member

  • PipPipPipPipPipPip
  • Urban Commando
  • Urban Commando
  • 205 posts
  • LocationSwitzerland

Posted 22 February 2013 - 01:20 AM

OP, if it helps calm your thoughts, just imagine Elo-rating is Battle Value, but for players instead of mechs.

Edited by Ramrod, 22 February 2013 - 01:20 AM.


#10 Diablobo

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,014 posts
  • LocationOn your six

Posted 22 February 2013 - 01:45 AM

Obviously Battle Values are not going to translate properly to MWO, but it is equally obvious that Elo is not going to work as well either. Elo is meant to measure players' skill levels. In MWO, equipment plays a huge role, so any system that ignores that factor is flawed at best. It seems a mix of the two will be necessary.

#11 KinLuu

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,917 posts

Posted 22 February 2013 - 01:49 AM

A good player won't bring bad equipment to the fight.
If he would, he would be no good player.

If you use bad builds, you are bad and you should feel bad. And your Elo will reflect that.

#12 semalferuzA

    Member

  • PipPipPipPipPip
  • 125 posts

Posted 22 February 2013 - 01:56 AM

View PostKinLuu, on 22 February 2013 - 01:49 AM, said:

A good player won't bring bad equipment to the fight.
If he would, he would be no good player.

If you use bad builds, you are bad and you should feel bad. And your Elo will reflect that.


What about when you are playing a new mech? Basic and elite skills make a huuge difference in pretty much every mech.

How about the fact that you have to play some of the terrible variants in order to get elite/master skills for the good variants?

Edited by semalferuzA, 22 February 2013 - 01:57 AM.


#13 KinLuu

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,917 posts

Posted 22 February 2013 - 02:00 AM

You will be a liability to your team. You will drop in Elo. Nothing you can do.

The problem here is the flawed XP system or the lack of good variants.

#14 semalferuzA

    Member

  • PipPipPipPipPip
  • 125 posts

Posted 22 February 2013 - 02:06 AM

View PostKinLuu, on 22 February 2013 - 02:00 AM, said:

You will be a liability to your team. You will drop in Elo. Nothing you can do.

The problem here is the flawed XP system or the lack of good variants.


Another problem with this is that it now takes *longer* to get the bad variants the xp they need. Pretty annoying. I'm not sure that they are willing to take any action to make the situation better either.

#15 Gaan Cathal

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 2,108 posts

Posted 22 February 2013 - 02:06 AM

View PostDiablobo, on 22 February 2013 - 01:45 AM, said:

Obviously Battle Values are not going to translate properly to MWO, but it is equally obvious that Elo is not going to work as well either. Elo is meant to measure players' skill levels. In MWO, equipment plays a huge role, so any system that ignores that factor is flawed at best. It seems a mix of the two will be necessary.

Actually, ELO is a metric of how much you win in a given weight class. Nothing else. Both player skill and equipment piwer factor into win/loss ratio and ergo into ELO score. In effect not only is there an activwly evolving BV involved already, it takes account of a pilot's ability with said weapon (gauss/PPC for example are very ping-dependant).

View PostsemalferuzA, on 22 February 2013 - 01:56 AM, said:


What about when you are playing a new mech? Basic and elite skills make a huuge difference in pretty much every mech.

How about the fact that you have to play some of the terrible variants in order to get elite/master skills for the good variants?


This is a global phenomenon and related ELO fluctuation will be temporary, same as having an 'off' weekend or being ill. ELO is dynamic and ergo self-correcting.

#16 semalferuzA

    Member

  • PipPipPipPipPip
  • 125 posts

Posted 22 February 2013 - 02:25 AM

View PostGaan Cathal, on 22 February 2013 - 02:06 AM, said:

This is a global phenomenon and related ELO fluctuation will be temporary, same as having an 'off' weekend or being ill. ELO is dynamic and ergo self-correcting.


I understand that it is temporary and happens to everyone but it isn't good for the game. People don't like playing knowing that they are disadvantaged and are likely to have poor performance for an extended period of time until the grinding is over. I don't have any solutions to the problem that aren't obvious or limiting in other ways.

#17 Dukarriope

    Member

  • PipPipPipPipPipPipPip
  • 923 posts
  • Twitter: Link
  • Locationa creative suite

Posted 22 February 2013 - 02:27 AM

A problem with ELO is that it is traditionally used to match people who will be using equal items. Hence why it is successful on something with two players that have identical pieces, e.g. chess.

I don't think having an Elo system be a primary matchmaker in a game where your loadout is just as meaningful is ideal.

#18 Gaan Cathal

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 2,108 posts

Posted 22 February 2013 - 02:28 AM

View PostsemalferuzA, on 22 February 2013 - 02:25 AM, said:


I understand that it is temporary and happens to everyone but it isn't good for the game. People don't like playing knowing that they are disadvantaged and are likely to have poor performance for an extended period of time until the grinding is over. I don't have any solutions to the problem that aren't obvious or limiting in other ways.


The problem isn't the effect of bad variants on ELO though, it's the existence of bad variants. Address that and the ELO sorts itself out.

#19 Gaan Cathal

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 2,108 posts

Posted 22 February 2013 - 02:33 AM

View PostDukarriope, on 22 February 2013 - 02:27 AM, said:

A problem with ELO is that it is traditionally used to match people who will be using equal items. Hence why it is successful on something with two players that have identical pieces, e.g. chess.

I don't think having an Elo system be a primary matchmaker in a game where your loadout is just as meaningful is ideal.


Apologies for the doublepost, but my phone hates the edit function on this forum.

Equipment selection contributes to your ELO score, since it affects your chancea of victory. We even have seperate ELO ratings for the weight classes. There is no need to factor in equipment twice when it comes to the matchmaker. Especially not given the abusesit would open up for minimal-to-zero benifit.

#20 Dukarriope

    Member

  • PipPipPipPipPipPipPip
  • 923 posts
  • Twitter: Link
  • Locationa creative suite

Posted 22 February 2013 - 02:35 AM

Again what I think is that make elo a component of a system that also considers equipment used. Rather than have Elo be the parent system.





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users