Jump to content

Elo Worthless


298 replies to this topic

#41 Roadbeer

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • 8,160 posts
  • LocationWazan, Zion Cluster

Posted 11 November 2013 - 07:55 PM

View PostRandalf Yorgen, on 11 November 2013 - 07:48 PM, said:

As someone who has dropped with you a fair bit Roadbeer I know how good and bad the matches can be but I have to disagree with you on how well the Elo works, at least as far as the Solo Drops are concerned. The players who have experience and who have developed some skills often find themselves grouped with players who are inexperienced and/or who are unwilling to listen, usually at a ratio of 1 old timer to 2 newer/non team players. Although this is supposed to help level the playing field and should help to make everything even it can over the long haul make it so that the old timers become bitter and not interested in doing solo drops and that feeds the belief of the soloists that vets/4man players are snobs, the circle is complete.

For example I just finished a solo drop on river city where three assaults and a medium on my team hid under the collapsed bridge between upper and lower city for 5 min of the match. When the suggestion was made that they get out from under the bridge and fight the response was "You don't know me so go yadda yadda insert suggestion here) If this was a one off it wouldn't be an issue but when it happens 40-60% of the time you cannot tell me that it's fair to the old timer to have to play with those 40 year old children over and over again.

As no one knows how MM/Elo works I can only say that suggesting Elo is good across the board is a personal opinion of yours and one I strongly disagree with from my personal experience. With CW hopefully coming online in the next few months having this system in place will do far more damage than I think people realise. A better system must be investigated by PGI and hopefully adopted for the game.


I'm going to have to plead ignorance about the shape of soloing right now as I haven't done a lot of it since early summer, I seem to spend most of my time grouped with you or others in the FWLM. So, I really don't have a horse in this race for the solo Elo argument, though, as I did state above, my stats when solo are about 25% less across the board than when i'm grouped.

#42 GRiPSViGiL

    Member

  • PipPipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 1,904 posts
  • Twitch: Link
  • LocationHillsboro, OR

Posted 11 November 2013 - 08:00 PM

View PostRoadbeer, on 11 November 2013 - 04:30 PM, said:


Posted Image

Skill will be the deciding factor.

#43 Duncan Aravain

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 416 posts
  • LocationBehind you with a sharp tool...er,mech

Posted 11 November 2013 - 08:08 PM

Roadbeer, don't suppose you can release that program you worked up......?

#44 GRiPSViGiL

    Member

  • PipPipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 1,904 posts
  • Twitch: Link
  • LocationHillsboro, OR

Posted 11 November 2013 - 08:10 PM

I don't think ELO is all it is cracked up to be. I mean I have 1.3 W/L ratio which as a PUGGer primarily is decent I would say and shows I make a little difference I hope. Sometime I have the good fortune to run in 4 man's. Our 12s mostly win so I don't know if that matter's or not. It just doesn't make sense in a team game unless it averages everyone's ELO on one team against the average ELO of the other.

#45 Roadbeer

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • 8,160 posts
  • LocationWazan, Zion Cluster

Posted 11 November 2013 - 08:13 PM

View PostDuncan Aravain, on 11 November 2013 - 08:08 PM, said:

Roadbeer, don't suppose you can release that program you worked up......?


I'm actually working on V.4 of it right now, cleaning it up, working out bugs. Got my hands on the Classic BT font to sex it up. I'm waiting on some of the changes of UI 2.0 to see if there is stuff that's still needed in it.
I'm hoping to have a release candidate of it out prior to CW, I'm even toying with the idea of putting it on the web rather than as an executable.

I'll keep you posted.

#46 Abivard

    Member

  • PipPipPipPipPipPipPipPip
  • Shredder
  • 1,935 posts
  • LocationFree Rasalhague Republic

Posted 11 November 2013 - 08:40 PM

ELO assigned to individual players based off random team w/l ratio can not work in a team game where the team can and does change..

No escaping that fact.

Indeed, ELO is worthless in MWO.

Either a more complex system taking many factors besides w/l is needed, or scrap ELO and go for something else.

#47 Randalf Yorgen

    Member

  • PipPipPipPipPipPipPipPip
  • 1,026 posts
  • Locationwith in 3m of the exposed Arcons rear ct

Posted 11 November 2013 - 09:27 PM

View PostMischiefSC, on 11 November 2013 - 07:54 PM, said:


So track the metrics. Track your score and the score of the people in the match, if you win or lose and how often your position in the top of the scoring on your side relates to a win/loss and how often your teammates carry then win vs how often you do.

If you can show that statistically you carry most of your wins and your losses but still only manage a 50/50 win/loss rate over 100 matches or so I'd absolutely believe the matchmaker is somehow utterly and completely flawed.

You're not going to find that though. You're going to track it and see that your teammates carry you far more often than you carry them. There's 11 of them and 1 of you, that's to be expected. You're also going to see that there are twits on the other team exactly as often as there are on your team.

You will see a correlation of ~8.33% of matches swinging either to win or loss based largely on your behavior. It may be a little higher or lower but over 100 matches pretty close to that.

You'll also see that the more you win the more skilled players end up on both sides of every game and fewer window-licking mouth-breathers. You'll see this number change if you're on a long losing streak. You'll also see that 10 or 20 matches either way isn't actually that significant to your overall performance, it's over 100, 300, 500 matches you'll see the impact of your playing measure out.


Track all the stats you want it's not going to change the fact, and yes I say fact because time and time and time and time again the numbers will say one thing but reality is something very different. You can hold up all the spreadsheets in the world and praise how something is but at the other end of it it's a very different story.

#48 KinLuu

    Member

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

Posted 12 November 2013 - 02:07 AM

View PostAbivard, on 11 November 2013 - 08:40 PM, said:

ELO assigned to individual players based off random team w/l ratio can not work in a team game where the team can and does change..

No escaping that fact.

Indeed, ELO is worthless in MWO.

Either a more complex system taking many factors besides w/l is needed, or scrap ELO and go for something else.


The only constant factor in your team are you.

If you are good, you will win more than you lose.
If you are bad, you will lose more than you win.

Until, after a couple of thousand games, you are at your true elo, and lose as much, as you win.

(DISCLAIMER: This only works out like this, if your performance is constant. Should you improve, while playing, you should win more than you lose, constantly.)

Elo works perfectly fine. But only after enough games. And because PGI decided to split your elo into the 4 different weight classes, you need a LOT of games before you reach your true elo.

#49 Fooooo

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,459 posts
  • LocationSydney, Aus.

Posted 12 November 2013 - 03:54 AM

View PostRoadbeer, on 11 November 2013 - 04:41 PM, said:

The thing is, isn't there supposed to be a symbiotic relationship between the Loyalists and the Mercs where when the Loyalists shift a battle line, that frees up an area where the Mercs can attack planets?

I know that CW is just barely beyond the cocktail napkin stage, but that was kind of what I took away from the Launch video.



I got something totally different.

Faction players = Fighting on moving fronts. (think heavy gears online play with North VS South)
Faction Players (PGI wishlist) = Merc style planet takeovers.

Lone Wolf = Fighting on moving fronts for whatever faction you align to for those battles, or just a random faction each battle. Also can be filler for merc corps needing an extra player. (although Im not sure how well that would work....most merc corps will have full teams and not really be a player down......)

Merc Player = Fighting on the periperary for planets. Full planet conquest is available.



Not sure how you got faction players making planets able to be attacked for mercs....thats never even mentioned afaik......


As for the OP.....

afaik Elo is not done by group avg to form the teams. Teams are formed through buckets basically. (searches for xxxx elo range players, if none found, extends range) Weither or not you gain or lose points after a win or loss however is I think.

Which means, if you are a massive high elo player and happen to get stuck with 11 , 500elo players....and end up vsing a team that is all massive high elo players (which is practically impossible)...... if you lose....you lose NO points on your elo.

However if you won that battle , you would gain a lot.

Edited by Fooooo, 12 November 2013 - 04:00 AM.


#50 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 12 November 2013 - 04:12 AM

View PostTekadept, on 11 November 2013 - 07:04 PM, said:

[/size]
Tha'ts not what I was getting at and I'm sure you know it, and charting yeh that would imply I gave a dezgra :)

I never said my team-mates keep me from dominating, I dominate just fine win or lose thanks ;)
I was just trying to point out it doesn't seem fair to those poor people who get forced up in ELO when they aren't ready for it.
For those of use from a 'more robust' generation, It builds character to be forced into a situation one is not ready for. You find out what you are made of an either sink or swim.

#51 Tekadept

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,290 posts
  • LocationPerth, Australia

Posted 12 November 2013 - 04:25 AM

View PostJoseph Mallan, on 12 November 2013 - 04:12 AM, said:

For those of use from a 'more robust' generation, It builds character to be forced into a situation one is not ready for. You find out what you are made of an either sink or swim.

I am from this "robust generation" to,i can still pity these poor little Gen Y or Z or whatever they are called these days :) who else are we going to shoot?

#52 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 12 November 2013 - 04:33 AM

View PostTekadept, on 12 November 2013 - 04:25 AM, said:

I am from this "robust generation" to,i can still pity these poor little Gen Y or Z or whatever they are called these days :) who else are we going to shoot?

...
...
... Each other? :rolleyes:
;)

#53 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 12 November 2013 - 06:32 AM

View PostAbivard, on 11 November 2013 - 08:40 PM, said:

ELO assigned to individual players based off random team w/l ratio can not work in a team game where the team can and does change..

No escaping that fact.


And yet, here we are, surrounded by evidence that it can and does work. In this thread alone we can find multiple examples of players who doubt the effectiveness of Elo matchmaking who, despite this, admit that they're being matched with high-tier players on a regular basis (ie "I see you in game all the time, eliteplayernumber1"). So we have "escaped" the "fact".

The "fact" is that most of the anti-Elo arguments are tired regurgitations: "Elo is for individuals, and this is a team game! CASE CLOSED!" What folks don't realize is that Elo matchmaking has been in use in team games for years, both online and offline (from Guild Wars to college football).

View PostAbivard, on 11 November 2013 - 08:40 PM, said:

Indeed, ELO is worthless in MWO.

Either a more complex system taking many factors besides w/l is needed, or scrap ELO and go for something else.


It is working, so it has worth. It's not working perfectly, but "in need of tweaking" or even "in need of an overhaul" is very different than "worthless." Say what you mean: "I can't conceive of how Elo works or can work, so it doesn't work." You're making an argument from ignorance... own it!

View PostTekadept, on 11 November 2013 - 07:04 PM, said:

[/size]
I was just trying to point out it doesn't seem fair to those poor people who get forced up in ELO when they aren't ready for it.


If folks are climbing in Elo, they are by definition "ready for it." If they're not, their Elo will be lowered in short order and they'll meet competition they're ready for once again. Unless by not ready for it you mean "not ready to give up stomping newbies".

Edited by FerretGR, 12 November 2013 - 06:39 AM.


#54 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 12 November 2013 - 08:00 AM

View PostRandalf Yorgen, on 11 November 2013 - 07:48 PM, said:

For example I just finished a solo drop on river city where three assaults and a medium on my team hid under the collapsed bridge between upper and lower city for 5 min of the match. When the suggestion was made that they get out from under the bridge and fight the response was "You don't know me so go yadda yadda insert suggestion here) If this was a one off it wouldn't be an issue but when it happens 40-60% of the time you cannot tell me that it's fair to the old timer to have to play with those 40 year old children over and over again.

I have similar experiences, in most matches.

The derp is strong in a good portion of my team-mates in matches.

#55 Amsro

    Member

  • PipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 3,438 posts
  • LocationCharging my Gauss Rifle

Posted 12 November 2013 - 09:31 AM

View PostFerretGR, on 12 November 2013 - 06:32 AM, said:


And yet, here we are, surrounded by evidence that it can and does work. In this thread alone we can find multiple examples of players who doubt the effectiveness of Elo matchmaking who, despite this, admit that they're being matched with high-tier players on a regular basis (ie "I see you in game all the time, eliteplayernumber1"). So we have "escaped" the "fact".

The "fact" is that most of the anti-Elo arguments are tired regurgitations: "Elo is for individuals, and this is a team game! CASE CLOSED!" What folks don't realize is that Elo matchmaking has been in use in team games for years, both online and offline (from Guild Wars to college football).

It is working, so it has worth. It's not working perfectly, but "in need of tweaking" or even "in need of an overhaul" is very different than "worthless." Say what you mean: "I can't conceive of how Elo works or can work, so it doesn't work." You're making an argument from ignorance... own it!

If folks are climbing in Elo, they are by definition "ready for it." If they're not, their Elo will be lowered in short order and they'll meet competition they're ready for once again. Unless by not ready for it you mean "not ready to give up stomping newbies".


I do agree Elo has been used and works in other implementations and versions.

Unfortunately, the way Team Elo is calculated you can have people from ANY Elo playing against or with people from ANY Elo, simply balance-ish the team average Elo. New people and very skilled people end up in my matches all the time.

In doing so this completly breaks any linear effect the rating system may have. Elo could work but it has been implemented in a completely flawed manner.

You should be able to win %80-%100 of your matches until you reach your "Elo Bracket". Right now you have a forced win/loss scenario that stems from radically different skill levels being on the same battlefield.

Only 12 man teams can benefit from this system as FAR less factors are involved.

I don't see any "proof" that Elo works,on the other hand I see a lot of its short comings.

So technically MWO Elo is completely useless.

#56 Sarsaparilla Kid

    Member

  • PipPipPipPipPipPipPip
  • The Benefactor
  • The Benefactor
  • 664 posts
  • LocationGold Country

Posted 12 November 2013 - 10:01 AM

View PostKunae, on 12 November 2013 - 08:00 AM, said:

I have similar experiences, in most matches.

The derp is strong in a good portion of my team-mates in matches.


I think for a lot of us, we just want to climb out of the derpage faster and play with others that have non-derp tendencies. The 200 point artificial Elo boost that new players get after finishing their first 25 matches seems to have a knack for launching mucho derpage into battles with seasoned vets. As Roadbeer pointed out with his stats on grouped vs. solo, the way to do that is to get on comms and git'er done.

#57 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 12 November 2013 - 10:52 AM

View PostSarsaparilla Kid, on 12 November 2013 - 10:01 AM, said:


I think for a lot of us, we just want to climb out of the derpage faster and play with others that have non-derp tendencies. The 200 point artificial Elo boost that new players get after finishing their first 25 matches seems to have a knack for launching mucho derpage into battles with seasoned vets. As Roadbeer pointed out with his stats on grouped vs. solo, the way to do that is to get on comms and git'er done.

The real problem comes in when the MM goes and puts the derpage on your team, every match by design, to "balance out" your higher Elo vs the opposing team.

#58 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 12 November 2013 - 11:29 AM

View PostAmsro, on 12 November 2013 - 09:31 AM, said:

Unfortunately, the way Team Elo is calculated you can have people from ANY Elo playing against or with people from ANY Elo, simply balance-ish the team average Elo. New people and very skilled people end up in my matches all the time.

Yes, but that's not because the matchmaker is pulling in ANY Elo vs ANY Elo, that's because the pool is very shallow and the beginner Elo is unfortunately set too high (1300).

View PostAmsro, on 12 November 2013 - 09:31 AM, said:

In doing so this completly breaks any linear effect the rating system may have.

Why? This isn't ideal, but it doesn't break anything. Even if the Elo ratings vary widely on a team, one team will have a higher overall or average Elo rating. That team should win. This is true of all Elos are between 1495 and 1505 or if all Elos are between 1000 and 2000. One team still has the higher Elo and that team still has the better chance of winning. Wide ranges of Elo are problematic, but they don't break the system or make winning and losing completely random. Are tighter ranges better? Of course. Will the system give a better indicator of which team "should" win when the ranges are tighter? Of course. Does a "loose range" make the system completely worthless? Of course not... it might lead to wonky results from time to time, it might mean it takes a lot longer for us to get to our "true Elo bracket", but it doesn't mean that the system doesn't work at all.

View PostAmsro, on 12 November 2013 - 09:31 AM, said:

You should be able to win %80-%100 of your matches until you reach your "Elo Bracket".

Says who? That'd be the case if there only if there were wildly different levels of ability between beginners and veterans, ie. if you threw NBA players into a pick-up game. I don't agree that's the case here. If you're expecting to see this and you're not, that's not Elo's fault.

View PostAmsro, on 12 November 2013 - 09:31 AM, said:

Only 12 man teams can benefit from this system as FAR less factors are involved.

How so? I could play with a different 12-man every day. Every match if I felt like it. My Elo rating is just as valid in 100% solo play as it is in 100% 12-man play. The only difference is that you'll likely be in a higher Elo "bracket" if you're training with the same 12 guys all the time: not because Elo has more meaning for 12-mans, but because you're simply better and your Elo rating will reflect that.

View PostAmsro, on 12 November 2013 - 09:31 AM, said:

I don't see any "proof" that Elo works,on the other hand I see a lot of its short comings.
So technically MWO Elo is completely useless.

I call ********. You see the proof every time you play. I play with you, so I know you see the same bunch of names in every second game. We're getting matched with good opposition... GOTE and the rest. How is that not proof? How is Elo "completely useless" when we, our team, sees the benefit from it every time we group up and play?

As for the shortcomings, what are you seeing that you can unequivocably blame on Elo matchmaking failing to work, but not on factors like time of day, small player base, too-high initial Elo, and so on?

Edited by FerretGR, 12 November 2013 - 12:00 PM.


#59 Malino

    Rookie

  • 7 posts

Posted 12 November 2013 - 12:33 PM

Hi, ok to clarify a few points,

The point I'm trying to make is ELO as a matchmaking tool in 12-v-12 is pretty much useless without also matching weights.

Especially, If as is said above the process attempts to match ELO across the team to balance. So you end up with a mix of experienced and inexperienced players.

Then when you take into acount there is no weight matching taking place so you see an imbalance of 2-300 tones between the teams you're ending up with a majority of matches ended up just exercises in being or doing the steamrolling.

Afterall, it dosn't matter if you're in the top 10% of the game if half your team's gone off and suicided and you're 3 Hunchie's facing 6 Atlas's you're pretty much screwed.

I'm not saying this in complaint at losing allot, I'm making this observation form being on both sides of the fence.

Just my opinion form playing since this change that matches would be allot more challenging if weight was used to balance rather than ELO.

And in regards to statistics and experience to make these observations I'm currently at:

Wins / Losses 1,808 / 1,203

And before someone trolls thats an attempt to be taken seriously.

Mal

#60 Malino

    Rookie

  • 7 posts

Posted 12 November 2013 - 12:40 PM

One last comment,

Especially if third of your team is locusts (third=4) and the other team's lightest mech is a Kintaro.

4x Locusts, fun yes, especially if you're on the other team and you have no lights. If it was weight balanced maybe it would be fun for both teams....

Mal





7 user(s) are reading this topic

0 members, 7 guests, 0 anonymous users