+ Reply to Thread
Page 1 of 3 1 2 3 LastLast
Results 1 to 15 of 31
Like Tree2Likes

Thread: This ELO system in RIFT makes no sense at all.

  1. #1
    Plane Walker Nickalispicalis's Avatar
    Join Date
    Jul 2013
    Posts
    472

    Default This ELO system in RIFT makes no sense at all.

    Logged on with my main and noted her ELO rating. 688. Terrible. Ok, played three matches, lost all three matches in big lopsided defeats. ELO rating dropped accordingly, but according to this rating, I should be grouped with some of the worst of the worst. It didn't appear that way to me. They seemed pretty skillful to me. I may discount my own ability as awful, although the numbers don't indicate that; ie; K/D ratio, dmg, objectives etc, but I would not cast that net on the people around me and certainly not the opposition that steamrolled us.

    So what is this ELO suppose to represent ? I understand some players have ELO ratings that exceed 3,000 ! Are they playing with players with a ELO of less then a 1000 ?

    The system makes no sense at all. To draw a chess analogy, a 688 ELO rated player would be a rank beginner, barely knowing the rules and how to move the pieces. The current W/C Magnus Carlsen rating is roughly 2,860. Should these two players play it would be so uneven, that Carlsen could probably give Queen odds and still win quite handily. The two would never meet apart from some type of simultaneous exhibition. I also play chess and maintain a rating of roughly 1920. I would not have much of chance of winning either, but the game would at least be competitive. The point is though, the system is remarkably accurate. If you are a 1800 rated player, you may go through a losing streak and see your rating drop or vice-versa and see it increase, but if that is your level of ability it would eventually settle to it.

    In RIFT, I have no idea what it is doing. If someone logs on and see a terrible rating the way my character has, it is not right to assume that automatically equates to you being terrible at PVP, just as someone would might have one of these crazy 3000 ELO ratings, while all their time was spent being carried on a premade, while they hardly contributed.

    I think part of the flaw is the RIFT system is based on Wins, Loses and Draws only and what the individual does or doesn't do counts for nothing.

  2. #2
    Ascendant SprawlnBrawl's Avatar
    Join Date
    May 2013
    Posts
    4,349

    Default

    from what we understand matchmaking takes everyone in the queue and splits them to try to make equal "team average ELO" for the 2 teams. So the lower your ELO the more people with higher ELO that should be on your team.

    i.e. say you had a 5v5 map

    10 people in queue with the following ELOs

    300
    500
    800
    1000
    1500
    1500
    1500
    1800
    2000
    3600

    The teams would look like this

    Team A: 3600, 300, 500, 1500, 1500 - Avg 1480
    Team B: 1000, 1500, 1500, 1800, 2000 - Avg 1560

    so Team A is one great player 2 scrubs and 2 avg

    Team B is 1 ok, 2 avg, 2 good

    now imagine this on a larger scale with 10 v 10 and 15 v 15 maps and LOTS of people with terrible ELOs and a handful of people with really high ELOs. The people with really high ELO will often get stuck having to carry all the bads against a team of average players. Some people like me can still handle this carry hence high ELO from solo queue, others can't and their ELO drops or they cry about soloing and stick to premade.

    What people would rather see is if you have high ELO you play with/against other people with high ELO but they don't do that likely due to queue times. But i would much rather the system TRY to do that first then put time limits before it adds other players.
    Last edited by SprawlnBrawl; 04-15-2015 at 12:52 PM.

  3. #3
    RIFT Fan Site Operator Riane's Avatar
    Join Date
    Apr 2011
    Posts
    5,612

    Default

    Won 4 in a row earlier and went up 6 Elo. I don't know how some people claim to have like 300 Elo and have been in WFs for months upon months.

    Wat r u doin

  4. #4
    Ascendant Violacea's Avatar
    Join Date
    May 2011
    Posts
    12,128

    Default

    These numbers are giving people false ideas I think.

    Bored, updated my sig to show low ratings can beat higher ratings.

    A warfront has too many dynamics for a simple win/loss rating record to accurately describe the competency of the player. It may ballpark it in some ways, but far from an end all explanation.
    New round up of some high rank matches
    Chun-Li*E.Honda*Evil Ryu
    Abel*Ibuki*Dhalsim
    Gouken*Zangief*Ryu

  5. #5
    Ascendant SprawlnBrawl's Avatar
    Join Date
    May 2013
    Posts
    4,349

    Default

    Quote Originally Posted by Riane View Post
    Won 4 in a row earlier and went up 6 Elo. I don't know how some people claim to have like 300 Elo and have been in WFs for months upon months.

    Wat r u doin
    if you won 4 and your ELO only went up 6 that means the ELO averages were out of wack and your teams had much higher average ELOs. I believe in most systems if you win vs an equal team you will go up/down ~15. Wins much higher rated teams are usually ~24+ and wins vs very low teams are <10 At least that is how it typically played out in WoW arenas.

    for example in wow if you were 2400+ rated (top 10 when i played) and for whatever reason there was nobody to match you with and you hit like a 1800 team you would gain like 1 point or 2 points for a win. but if you lost it would be like -30.

    The alternative is if they are making team AVERAGE and then comparing that to your actual personal ELO for points. Not sure which char you queued on but if you have a high personal ELO near the top you are always going to gain very few points unless you get a team with total average ELO close to your personal ELO. This is why they really need to show individual and team averages on the scoreboard so we can see how it actually works. I'm curious if you only gained 6 on those 4 if you had lost one theoretically you would have lost like 20+ rating.
    Last edited by SprawlnBrawl; 04-15-2015 at 01:00 PM.

  6. #6
    Shadowlander
    Join Date
    May 2011
    Posts
    28

    Default ELO misunderstandings?

    I don't think people are interpreting the ELO correctly.

    If the matchmaker system tries to balance the ELO amongst two groups from the whole queue'ing population by matching their average then ELO isn't going to tell you how good you are. In other words ELO isn't a rank of your skill. What ELO is going to tell you is how well the matchmaker system is matching folks without adjustment for the number of people bypassing the matchmaking system by simply disabling it (group queue). If the system worked perfectly and nobody group queue'd then everyone would have the same ELO, right?

    Groups would have to be matched in brackets to make this less true. Even then it won't be completely untrue, just less true. ELO would be closer to a rank at that point, but still not a rank.

    If wins/losses belonged to the individual (like chess), this wouldn't be the case, but they don't, so it is. You don't win or lose your matches. You just participate in a winning or losing match, contributing a small % either way.

    Admittedly I haven't looked into the math of this yet, but just at first glance, it seems to me that ELO does not equal E-peen rank?
    Last edited by Accountant10; 04-15-2015 at 02:19 PM.

  7. #7
    Rift Disciple Colibri's Avatar
    Join Date
    Mar 2015
    Posts
    129

    Default

    My ELO is 1003, and I have no idea what this means other than perhaps I am either average, or the token scrub.

  8. #8
    Soulwalker
    Join Date
    Feb 2011
    Posts
    3

    Default

    My ELO is 2212. Did I pass?

  9. #9
    Plane Touched Fusscle's Avatar
    Join Date
    Sep 2014
    Posts
    273

    Default

    This ELO system in RIFT makes no sense at all.-wl.jpg

    Win Loss percentage looks a little strange. The game seems to think I win more than I actually do. It's not lifetime Win/loss either, because it moves quite a bit after each warfront.
    Fusscle of Cheat Engine

  10. #10
    RIFT Fan Site Operator Riane's Avatar
    Join Date
    Apr 2011
    Posts
    5,612

    Default

    Quote Originally Posted by Fusscle View Post
    Attachment 23150

    Win Loss percentage looks a little strange. The game seems to think I win more than I actually do. It's not lifetime Win/loss either, because it moves quite a bit after each warfront.
    Of course it's not lifetime. It just started tracking win/loss via that UI today.

    4 7 = 57%

  11. #11
    Plane Touched Xgame's Avatar
    Join Date
    Apr 2014
    Posts
    258

    Default

    My rift elo is 189 i have no idea if it's a good thing or not ,but i finished all my dailies pretty quick and have a good amounts of win. WHOEVER HAS THE 1 ELO OUT THERE I CHALLENGE YOU :<

    Xgame@deepwood

    Title known as: Xgame The Legendary Rogue

  12. #12
    Plane Touched Fusscle's Avatar
    Join Date
    Sep 2014
    Posts
    273

    Default

    Quote Originally Posted by Riane View Post
    Of course it's not lifetime. It just started tracking win/loss via that UI today.

    4 7 = 57%
    4 wins + 7 losses = 11 games played

    4 / 11 = 36%
    Fusscle of Cheat Engine

  13. #13
    RIFT Fan Site Operator Riane's Avatar
    Join Date
    Apr 2011
    Posts
    5,612

    Default

    Quote Originally Posted by Fusscle View Post
    4 wins + 7 losses = 11 games played

    4 / 11 = 36%
    Oh huh yeah I forgot that Win % is Wins/Overall. The math I gave correlates to yours though so I guess that's how it's calculating it and no one noticed because no one did WFs on the PTS to see that. Bug report ahoy

  14. #14
    Plane Touched Fusscle's Avatar
    Join Date
    Sep 2014
    Posts
    273

    Default

    Quote Originally Posted by Riane View Post
    Oh huh yeah I forgot that Win % is Wins/Overall. The math I gave correlates to yours though so I guess that's how it's calculating it and no one noticed because no one did WFs on the PTS to see that. Bug report ahoy
    I guess it's actually a ratio shown as a percent.
    Fusscle of Cheat Engine

  15. #15
    Rift Disciple
    Join Date
    May 2011
    Posts
    139

    Default

    It has to do with the number of games, too. I'm casual, so my 660 is pretty darn good.

+ Reply to Thread
Page 1 of 3 1 2 3 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