+ Reply to Thread
Page 1 of 4 1 2 3 4 LastLast
Results 1 to 15 of 57
Like Tree18Likes

  Click here to go to the first Rift Team post in this thread.   Thread: Displaying Elo to the player-base only shows how broken the elo matchmaking system is

  1. #1
    Shadowlander
    Join Date
    Apr 2011
    Posts
    45

    Default Displaying Elo to the player-base only shows how broken the elo matchmaking system is

    1. If Trion is going to show Elo then they really need to get serious about PVP. add some new gear. add some arena perhaps someday. I don't care if it's broken. I just want it to exist.

    2. If Trion is not going to show any signs of improving pvp in this game. Then take it off displaying it to the player-base. I don't care to see my "Warfront elo"

  2. #2
    Champion R3db4ss's Avatar
    Join Date
    Jun 2012
    Posts
    584

    Default

    only thing broken about it is that you can exactly tell who was carried and who wasnt for his ELO., because it will lead to ppl(at least a few over 3k Elo) dropping ELO on purpose the next weeks...

    I think its the right start, but Trion needs to act on the Feedback, and react on Matchmaking based on the REAL Feedback aroun ELO ratings.
    But all this " Matchmaking sucks, change it, but i wont/cant give solid feedback,...." threads are eating time from the devs and wont get us players nowhere imo, so give solid feedback or let it be and move on ...

    Quote Originally Posted by Daglar View Post
    This game is not and never will be balanced for 1 vs 1.
    ~Daglar

  3. #3
    Ascendant Violacea's Avatar
    Join Date
    May 2011
    Posts
    12,128

    Default

    Broken is really a bad word for it, misused imo. I'd much rather choose the word inefficient. I would love to almost assign it as a college type project, show me the methods, procedures and data you would collect and employ to create the perfect match making system.

    Not an easy task to get an A+ on when you get down to it. I've seen games with much more tangible aspects to base their ratings off of for balance purposes, and even those have daily flaws that occur that you say "Hmm, that doesn't make sense going purely off numbers." Without role locking, I have a lot of trouble thinking of anything truly concrete that you could form even teams from.

    Wins/Losses aren't imperative.
    Damage/Healing done isn't imperative.
    Objectives interacted with isn't imperative. (maybe someone just grabs lonely objectives and dies soon after).
    Gear, many know, isn't imperative to skill either.

    If I learned anything from the 5 v 5's and 3 v 3's, the most important factor in a win/loss truly seems to be the composition and precise roles/souls used. Which WF's have thrown in the air as a free-for-all as you can change to any role and soul you wish once dropped into the WF.

    So, all the stats they are willing to look at, I feel are pretty lacking in truth telling of ones ability. And the one that might hold some value to overall team ability, isn't addressed because they want to keep your guys freedom to change souls within a WF in tact.

    Tricky eh?
    New round up of some high rank matches
    Chun-Li*E.Honda*Evil Ryu
    Abel*Ibuki*Dhalsim
    Gouken*Zangief*Ryu

  4. #4
    Ascendant Kronos v's Avatar
    Join Date
    Feb 2011
    Posts
    7,232

    Default

    Quote Originally Posted by R3db4ss View Post
    But all this " Matchmaking sucks, change it, but i wont/cant give solid feedback,...." threads are eating time from the devs and wont get us players nowhere imo, so give solid feedback or let it be and move on ...
    I've given plenty of solid feedback, suggestion, and ideas concerning matchmaking and many other aspects of pvp.
    all I've gotten for it is to have my VK giggles nerfed. twice!

  5. #5
    Shield of Telara
    Join Date
    Sep 2011
    Posts
    767

    Default

    Quote Originally Posted by R3db4ss View Post
    only thing broken about it is that you can exactly tell who was carried and who wasnt for his ELO., because it will lead to ppl(at least a few over 3k Elo) dropping ELO on purpose the next weeks...

    I think its the right start, but Trion needs to act on the Feedback, and react on Matchmaking based on the REAL Feedback aroun ELO ratings.
    But all this " Matchmaking sucks, change it, but i wont/cant give solid feedback,...." threads are eating time from the devs and wont get us players nowhere imo, so give solid feedback or let it be and move on ...

    I'm a main carrier. People i don't even know mark me dmg/skull and.... 124 elo. Elo is completely borked.
    Ezuen@Faeblight<Godlike Awesomeness> lvl 65 p100 warrior
    aka "The Healing Warrior" (before it was cool)

  6. #6
    Prophet of Telara
    Join Date
    Apr 2014
    Posts
    1,058

    Default

    Quote Originally Posted by oddiescell View Post
    I'm a main carrier. People i don't even know mark me dmg/skull and.... 124 elo. Elo is completely borked.
    You can get a lot of Elo all of a sudden. Will get back to the thousands range soon enough. I've gained a 100 to 200 points in a few warfronts. Yet on my other toon around 3K elo iti is like 1 point for winning and 4-6 points for losing. It will even out eventually. Elo means nothing until they find a way to use it properly, it has been that way forever. Matchmaking is still a joke.

  7. #7
    Ascendant Bliter's Avatar
    Join Date
    Mar 2011
    Posts
    2,564

    Default

    Isn't it just being misapplied? How do you score a solo queue versus a premade player when a team effort isn't the same as an individual effort? A premade team could have five people playing for the win, random solo queue you could have 19 potatoes looking for free gear.
    Quote Originally Posted by Hailolyergory View Post
    All, Please nerf paper it is too strong, but leave scissors alone they are fine.
    Yours Truly,
    Rock.

  8. #8
    Rift Team Ocho's Avatar
    Join Date
    Mar 2014
    Posts
    1,933

    Default

    Quote Originally Posted by Violacea View Post
    Wins/Losses aren't imperative.
    Our current ELO system measures how good you are at winning long term. It isn't a skill at killing each other or wining fights, it's skill at winning matches. That's a pretty significant distinction.

    What the matchmaker does is try to put together teams that are as equally likely to win as possible. Given that choice, how good you are at winning (compared to how good the opposing team was at winning) is the only relevant stat.

  9. #9
    Plane Walker
    Join Date
    Feb 2014
    Posts
    450

    Default

    Ocho, Is the current system fair to healers who try to queue up when few healers are queuing up? I had a horrid losing streak last week 2 days in a row when i was queing during off peak hours.

    Seemed like I kept ending up against a PM, and there were seldom more than 2 healers on my team. I always ST heal and am always near the top of the ST healers.

    It just seems like with what has been described. Healers can end up in some pretty bad situations at times.

    At one point I just put together a PM and then won 4 in a row after the first horrid streak. Then over the weekend when more healers were queuing the matches seemed closer and things seemed not nearly so lopsided again.

    And my ELO is nowhere near decent, right around 1500.

  10. #10
    Shield of Telara
    Join Date
    Oct 2011
    Posts
    758

    Default

    Quote Originally Posted by oddiescell View Post
    I'm a main carrier. People i don't even know mark me dmg/skull and.... 124 elo. Elo is completely borked.
    Why would you complain about your Elo being too low? Just enjoy it while it lasts.
    EvanX / Nihilist / Insurgent

  11. #11
    Shield of Telara
    Join Date
    Oct 2011
    Posts
    758

    Default

    Quote Originally Posted by Bliter View Post
    Isn't it just being misapplied? How do you score a solo queue versus a premade player when a team effort isn't the same as an individual effort? A premade team could have five people playing for the win, random solo queue you could have 19 potatoes looking for free gear.
    Being a premade seems to have it's disadvantages as well. I mostly solo queue, so ya I get the occasional ROFLSTOMP from a premade. But I've also been in plenty of matches where the premade is paired with low Elo players against a stacked team of solo'ers or another premade.

    In my experience, premade does not equal autowin.
    EvanX / Nihilist / Insurgent

  12. #12
    Plane Touched Sinsalot's Avatar
    Join Date
    Mar 2011
    Posts
    257

    Default

    Quote Originally Posted by EvanX View Post
    Being a premade seems to have it's disadvantages as well. I mostly solo queue, so ya I get the occasional ROFLSTOMP from a premade. But I've also been in plenty of matches where the premade is paired with low Elo players against a stacked team of solo'ers or another premade.

    In my experience, premade does not equal autowin.
    The last 5 man HK premade I was in won 11 in a row. The one before that was 9/10 wins when we lost to another HK PM.

    So you are correct premade doesn't mean autowin, just autowin 20/21 times.

    Are you sure the team you are playing is a PM? If you were to see me in a warfront you'd probably call it a PM, but I solo queue the vast majority of the time.

  13. #13
    Plane Touched
    Join Date
    May 2011
    Posts
    289

    Default

    Quote Originally Posted by Ocho View Post
    Our current ELO system measures how good you are at winning long term. It isn't a skill at killing each other or wining fights, it's skill at winning matches. That's a pretty significant distinction.

    What the matchmaker does is try to put together teams that are as equally likely to win as possible. Given that choice, how good you are at winning (compared to how good the opposing team was at winning) is the only relevant stat.
    It's actually just flat out not working.

    Was q'ing with someone 2k lower rating, we both intentionally lost and he lost 6 more points than I did. Explain.

  14. #14
    Plane Touched BigTerj's Avatar
    Join Date
    Aug 2012
    Posts
    286

    Cool

    Quote Originally Posted by JudeFromKale View Post
    It's actually just flat out not working.

    Was q'ing with someone 2k lower rating, we both intentionally lost and he lost 6 more points than I did. Explain.
    Not sure if you even bothered to read any of Ocho's threads, but he explained that the Elo is a LONG TERM calculation, so your friend with the low Elo has probably only been playing a short time or not played many WFs, so his Elo is more volatile than a player with a longer PvP history.

    Example:
    You've got a .500 match record with 1000 matches, your friend has the same record but only 100 matches under his belt. You both win, now your score is 501/1001=0.5005 and his is 51/101=0.505. His score change is an order of magnitude larger than yours.

    It's simple maths.

    Edit: The exact Elo system is slightly more complex as the +/- of your score also depends on the predicted outcome and actual outcome of each match, but the concept is same: Long term averages are affected less by a single match outcome than short term averages.
    Last edited by BigTerj; 04-20-2015 at 06:38 PM.
    Bigterj - <Grievance> - Wolfsbane

  15. #15
    Shield of Telara Spungey's Avatar
    Join Date
    Mar 2011
    Posts
    747

    Default

    Quote Originally Posted by BigTerj View Post
    Not sure if you even bothered to read any of Ocho's threads, but he explained that the Elo is a LONG TERM calculation, so your friend with the low Elo has probably only been playing a short time or not played many WFs, so his Elo is more volatile than a player with a longer PvP history.

    Example:
    You've got a .500 match record with 1000 matches, your friend has the same record but only 100 matches under his belt. You both win, now your score is 501/1001=0.5005 and his is 51/101=0.505. His score change is an order of magnitude larger than yours.

    It's simple maths.

    Edit: The exact Elo system is slightly more complex as the +/- of your score also depends on the predicted outcome and actual outcome of each match, but the concept is same: Long term averages are affected less by a single match outcome than short term averages.
    That just doesn't make sense to me. I know of people who have played this game over a longer period of time than myself. 8 months in 2011 + 3 1/2 months in 2015. Players who have played a ton of matches from vanilla til now gain/lose more than myself. Think I'm at 791. Given I pvpd all the time during vanilla except for raid hours. But was in pm with a 3k player. He gains 4 points for win.. myself gains 2. Given your theory I should gain more points. Idk. Still think its borked.
    "TO PRO PHYS HEALS FOR JUUUU"

    (Vanilla: Spongey@Keenblade "Nexilis")
    (Nightmare Tide: Excuse@Deepwood

+ Reply to Thread
Page 1 of 4 1 2 3 4 LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts