test content
What is the Arc Client?
Install Arc

It doesn't make sense to drop on leader board after winning

biibiisaibiibiisai Member, NW M9 Playtest Posts: 133 Arc User
edited October 2014 in PvE Discussion
So winners drop places on leaderboard, get matched up with lower rank people, win even more, drop even more, win more no-brainer matches, drop even more.... This is just a positive feedback vicious cycle that leads to unbalanced matches. Can we perhaps revisit this logic?

EDIT:
And... this is about unbalanced match making, not about leaderboard ranking.
Post edited by Unknown User on

Comments

  • eldartheldarth Member Posts: 4,494 Arc User
    edited October 2014
    biibiisai wrote: »
    So winners drop places on leaderboard, get matched up with lower rank people, win even more, drop even more, win more no-brainer matches, drop even more....

    By "dropping" on the leader board, while you are winning against lower ranked people, and thereby gaining very few ELO rating points perhaps others are playing and winning against higher ranked opponents and gaining more ELO rating points and PASSING you by? You mean that?

    If only Cryptic would actually show everyone's ELO rating score people would understand what's actually occurring.
  • frishterfrishter Member Posts: 3,522 Arc User
    edited October 2014
    eldarth wrote: »
    By "dropping" on the leader board, while you are winning against lower ranked people, and thereby gaining very few ELO rating points perhaps others are playing and winning against higher ranked opponents and gaining more ELO rating points and PASSING you by? You mean that?

    If only Cryptic would actually show everyone's ELO rating score people would understand what's actually occurring.

    There was a way to check that supposedly and people found out that your elo was actually decreasing on an easy win. They fixed the ability to check though.
  • biibiisaibiibiisai Member, NW M9 Playtest Posts: 133 Arc User
    edited October 2014
    eldarth wrote: »
    By "dropping" on the leader board, while you are winning against lower ranked people, and thereby gaining very few ELO rating points perhaps others are playing and winning against higher ranked opponents and gaining more ELO rating points and PASSING you by? You mean that?

    If only Cryptic would actually show everyone's ELO rating score people would understand what's actually occurring.

    I'm not sure what exactly goes on in the background, but I monitor my leaderboard status right before match starts, right before match ends, and right after match ends, for both winning matches and losing ones. My page number does not change between matches, even after a night's sleep. It only changes right after matches. Since I reached page 25, I have been constantly dropping for about eight matches, about 6 of which were winning matches. I closed the game so I'm not 100% sure about the numbers atm. Most of my later games were against people with at least 2k less gs than ours and white mounts... This is not fun or fair. I only pug so there's no premade issues involved.
  • adamy2004adamy2004 Member Posts: 226 Arc User
    edited October 2014
    I wouldn't even worry about leaderboard positions, dropping positions by winning is only part of the problem, the fact that many players still cheat the leaderboard by leaving matches is making the whole system a complete joke.
    Don't waste my time.
  • qq88ppqq88pp Member Posts: 143 Arc User
    edited October 2014
    frishter wrote: »
    There was a way to check that supposedly and people found out that your elo was actually decreasing on an easy win. They fixed the ability to check though.
    Seems to have been easier than fixing the leader board mechanic.

    "These aren't the droids you're looking for".
  • rashylewizzrashylewizz Member Posts: 4,265 Bounty Hunter
    edited October 2014
    I like the intent. Cryptic didn't want the leaderboard to be who played the most matches.

    The problem is, there is no way for players to get matched with other premades/other good players 100% of the time, so they get matched with PUGs they don't want to fight.
  • koralzombiekoralzombie Member Posts: 0 Arc User
    edited October 2014
    adamy2004 wrote: »
    I wouldn't even worry about leaderboard positions, dropping positions by winning is only part of the problem, the fact that many players still cheat the leaderboard by leaving matches is making the whole system a complete joke.

    This.

    Look at the leaderboard. Look at page 1. Divide the total number of players on page 1 by 7. Subtract 7 from that number. You now are looking at the number of players on that board who do not cheat the system by playing against premade bot teams or leaving matches before they end etc.
  • magiquepursemagiquepurse Member Posts: 0 Arc User
    edited October 2014
    frishter wrote: »
    There was a way to check that supposedly and people found out that your elo was actually decreasing on an easy win. They fixed the ability to check though.
    Exactly this.

    There is a hidden Rating Score, that when you win against low ELO pugs, goes down. It is as simple as that.

    People PROVED it, with NUMBERS, here in the forums. You win, rating goes down, you drop on the leaderboard regardless of what happens to other people playing games.

    Then, instead of fixing it, or at least giving an official explanation, they HID the ratings so we now cannot sustain our claims with math anymore.

    Nice job.
  • lewel555lewel555 Member Posts: 616 Arc User
    edited October 2014
    biibiisai wrote: »
    So winners drop places on leaderboard, get matched up with lower rank people, win even more, drop even more, win more no-brainer matches, drop even more.... This is just a positive feedback vicious cycle that leads to unbalanced matches. Can we perhaps revisit this logic?

    Elo is fine. If Elo is correctly implemented you don't lose points when winning. Carlsen won't lose Elo by defeating me over the chessboard despite our 600 points difference. Elo in itself is also self-balancing. The higher you are in Elo, the more difficult it is to gain more ponts.
    The thing is, Carlsen won't play tournaments where people like me are playing. He will only enter tournaments with people his size (nearly) and where he can win rating if he performs well.
  • djarkaandjarkaan Member Posts: 883 Arc User
    edited October 2014
    eldarth wrote: »
    By "dropping" on the leader board, while you are winning against lower ranked people, and thereby gaining very few ELO rating points perhaps others are playing and winning against higher ranked opponents and gaining more ELO rating points and PASSING you by? You mean that?

    This, if you would be match against players close to your ELO ranking it would not happen that often but could still happen, the issue is matchmaking not that you go down after winning.

    1)Winning against higher rank then yourself = Gain lots of points
    2)Winning against lower rank then yourself = Gain few points or non at all.

    If the player behind you gets 1 and you get 2 he might/will pass you. Not hard to understand really.
  • hefisdohefisdo Member Posts: 709 Bounty Hunter
    edited October 2014
    I was in page 13 when I stopped PvPing. I didn't PvP for like 2 weeks and I was still in page 13; then I went to a match, won easily and dropped to page 22.

    If everyone in the first pages were actually PvPing every day, the flow of people going and leaving the tops would be much better; they're definitely stuck there by doing nothing, since it's the best thing to do when you reach there.
    Look at the leaderboard. Look at page 1. Divide the total number of players on page 1 by 7. Subtract 7 from that number. You now are looking at the number of players on that board who do not cheat the system by playing against premade bot teams or leaving matches before they end etc.

    Looks like -4,14 players of the first page don't cheat.
    (´・ ω ・`)
  • blacksladdiblacksladdi Member Posts: 359 Arc User
    edited October 2014
    you may not climb leaderboard pages but you do get a win so you can flaunt your win:loss ratio - even if you steamroll pug teams or people who just hit 60 and still don't know how to play their class in pvp (that's me, except on my rogue i know a bit there)
  • magiquepursemagiquepurse Member Posts: 0 Arc User
    edited October 2014
    Yet again:

    - There is a hidden Rating Score. It looks like a number ranging from 15 or so to 53 or so, such as 52.899 Rank 1 or 15.0001 for low ranks
    - the score was once VISIBLE by console command. It was posted publicly in the forums for everybody to see
    - if you win against very low ELO pugs, YOUR SCORE WILL DECREASE.

    This is NOT about people playing against a better team, winning and getting a higher Score overtaking you. You will drop on the leaderboard purely based on your own score going down.

    All was posted here on the forums with images and proof.
  • lewstelamon01lewstelamon01 Member Posts: 7,415 Arc User
    edited October 2014
    To make an analogy here, and apologies to those who don't understand college football in the States:

    Teams shouldn't drop in ranks in college football because they keep winning, yet some do. Why? Strength of schedule. Example? Florida State has a win streak going back to last season, yet recently slipped out of the #1 spot, because another school won against a much stronger slate of teams. It may not be formally used anymore (e.g. BCS poll), but when coaches are voting for teams, they're going to take a look at who those teams played against. If you have a cupcake schedule, your chance of getting a good ranking isn't as great as a team that played a grueling schedule and won against all comers.

    Same applies here, except it's codified in an algorithm. If you're winning, but those wins are coming against figurative cupcakes, then you're actually going to drop in ranking because...well...it's not real proof of your standing.

    Case in point: If one player won a match against a target dummy, and another won a match against an uber PVP geared player, who do you think should be ranked higher, all other things being equal?
    ROLL TIDE ROLL

    Great Weapon Fighter: Because when is today not a good day to die?

    PC and PS4 player. Proud Guildmaster for PS4 Team Fencebane. Rank 5 Officer for PC Team Fencebane. Visit us at http://fencebane.shivtr.com
  • magiquepursemagiquepurse Member Posts: 0 Arc User
    edited October 2014
    To make an analogy here, and apologies to those who don't understand college football in the States:

    Teams shouldn't drop in ranks in college football because they keep winning, yet some do. Why? Strength of schedule. Example? Florida State has a win streak going back to last season, yet recently slipped out of the #1 spot, because another school won against a much stronger slate of teams. It may not be formally used anymore (e.g. BCS poll), but when coaches are voting for teams, they're going to take a look at who those teams played against. If you have a cupcake schedule, your chance of getting a good ranking isn't as great as a team that played a grueling schedule and won against all comers.

    Same applies here, except it's codified in an algorithm. If you're winning, but those wins are coming against figurative cupcakes, then you're actually going to drop in ranking because...well...it's not real proof of your standing.

    Case in point: If one player won a match against a target dummy, and another won a match against an uber PVP geared player, who do you think should be ranked higher, all other things being equal?

    You're 100% wrong.

    Please read my post above yours. What you're talking about is a whole different thing.

    It was already proven your Rating decreases when you win against low rating pugs. Was posted numerous times before. This happens INDEPENDENTLY from what other people do, win/lose etc.

    Let's take an example:
    - let's consider a fully dead game, NOBODY ONLINE, besides 10 people in queue
    - first team is composed of Page 1 people, the other team page 2000
    - the first team wins
    - the people in the first team will drop in the leaderboard

    Why are you people ignoring it? I honestly cannot understand this attitude.
  • djarkaandjarkaan Member Posts: 883 Arc User
    edited October 2014
    You're 100% wrong.

    Please read my post above yours. What you're talking about is a whole different thing.

    It was already proven your Rating decreases when you win against low rating pugs. Was posted numerous times before. This happens INDEPENDENTLY from what other people do, win/lose etc.

    Let's take an example:
    - let's consider a fully dead game, NOBODY ONLINE, besides 10 people in queue
    - first team is composed of Page 1 people, the other team page 2000
    - the first team wins
    - the people in the first team will drop in the leaderboard

    Why are you people ignoring it? I honestly cannot understand this attitude.

    So winning against much weaker/lower rank teams decreases your Score? That is great news actually, Top players might think twice before stomping weak new players who just want to learn. Again the issue is with matchmaking as these match ups should not happen.
  • magiquepursemagiquepurse Member Posts: 0 Arc User
    edited October 2014
    djarkaan wrote: »
    So winning against much weaker/lower rank teams decreases your Score? That is great news actually, Top players might think twice before stomping weak new players who just want to learn. Again the issue is with matchmaking as these match ups should not happen.
    Indeed it does. And I hate stomping pugs as much as you do, "stomping" for me is actually going mid, throwing a daily, killing a few people than AFKing if I am not attacked. I ain't gonna hunt 10K Gs newbies with with my 20K. I once had bad gear; it was really ****ty. I have not forgot how it is to be stomped and taunted by nabs in full premades.

    These being said, because of the leaver penalty, I CANNOT AVOID the pugstomps. If we leave against pugs, we take a loss and go down, and have 30 minutes penalty. If we stay and pugstomp, we win and go down and have no penalty.

    These are the "options"...
  • hefisdohefisdo Member Posts: 709 Bounty Hunter
    edited October 2014
    To make an analogy here, and apologies to those who don't understand college football in the States:

    Teams shouldn't drop in ranks in college football because they keep winning, yet some do. Why? Strength of schedule. Example? Florida State has a win streak going back to last season, yet recently slipped out of the #1 spot, because another school won against a much stronger slate of teams. It may not be formally used anymore (e.g. BCS poll), but when coaches are voting for teams, they're going to take a look at who those teams played against. If you have a cupcake schedule, your chance of getting a good ranking isn't as great as a team that played a grueling schedule and won against all comers.

    Same applies here, except it's codified in an algorithm. If you're winning, but those wins are coming against figurative cupcakes, then you're actually going to drop in ranking because...well...it's not real proof of your standing.

    Case in point: If one player won a match against a target dummy, and another won a match against an uber PVP geared player, who do you think should be ranked higher, all other things being equal?

    It's new information for me that the US football teams have a blindfold when they're forming and accepting a match.

    It must be very disappointing when you're in the match and then you notice that you're playing against kids of the elementary school.
    (´・ ω ・`)
  • djarkaandjarkaan Member Posts: 883 Arc User
    edited October 2014
    hefisdo wrote: »
    It's new information for me that the US football teams have a blindfold when they're forming and accepting a match.

    It must be very disappointing when you're in the match and then you notice that you're playing against kids of the elementary school.

    His Analogy was bad, take the ATP and WTA. #1 beats #300 and #3 Beats #2 good depending on the 50 previous week #3 may past #1.
  • biibiisaibiibiisai Member, NW M9 Playtest Posts: 133 Arc User
    edited October 2014
    To make an analogy here, and apologies to those who don't understand college football in the States:

    Teams shouldn't drop in ranks in college football because they keep winning, yet some do. Why? Strength of schedule. Example? Florida State has a win streak going back to last season, yet recently slipped out of the #1 spot, because another school won against a much stronger slate of teams. It may not be formally used anymore (e.g. BCS poll), but when coaches are voting for teams, they're going to take a look at who those teams played against. If you have a cupcake schedule, your chance of getting a good ranking isn't as great as a team that played a grueling schedule and won against all comers.

    Same applies here, except it's codified in an algorithm. If you're winning, but those wins are coming against figurative cupcakes, then you're actually going to drop in ranking because...well...it's not real proof of your standing.

    Case in point: If one player won a match against a target dummy, and another won a match against an uber PVP geared player, who do you think should be ranked higher, all other things being equal?

    I don't care the least about my leaderboard ranking. I started tracking it after I heard that the ranking determines who I'm matched up against. If I win, drop in ranking, and keep getting matched up with worse and worse players with lower and lower rankings, giving me easier and easier wins, it is against what I'm looking for in pvp. So whether it makes sense to drop in ranking for the ranking itself is not relevant for me. What is relevant is how this influences who I am matched up against.
  • biibiisaibiibiisai Member, NW M9 Playtest Posts: 133 Arc User
    edited October 2014
    djarkaan wrote: »
    So winning against much weaker/lower rank teams decreases your Score? That is great news actually, Top players might think twice before stomping weak new players who just want to learn. Again the issue is with matchmaking as these match ups should not happen.

    Intentional stomping happens if only you can choose who you are matched up against... With the current matching system, it's extremely hard to pick who to match up against unless you make 2 premades and queue together. How many guilds now have the power to do that?
  • frishterfrishter Member Posts: 3,522 Arc User
    edited October 2014
    djarkaan wrote: »
    So winning against much weaker/lower rank teams decreases your Score? That is great news actually, Top players might think twice before stomping weak new players who just want to learn. Again the issue is with matchmaking as these match ups should not happen.

    So a system that punishes you for being better is great news? I'm sorry, but what kind of logic is that. If you're competitive then you don't want to hold back, you want to give it your all. If the matchmaking system puts you against a worse team, it's not your fault or theirs. Though I don't know why more people don't abuse the system by letting the enemy cap some so that they don't fall. It boggles my mind that people defend blatant inferior systems. Compare the ladder to other games ladders and this clearly doesn't compare. I really have no clue why people keep doing it.
  • yawulfyawulf Member Posts: 139 Arc User
    edited October 2014
    The system makes sense u get more points if u win against a formidable team and get less points if u win against weaker team. They explained it when the board was announced but honestly if it was a league where u get set points for winning and no point for losing it wodnt be fair in a 1000-987 match but maybe most of these problems with ppl falling in rank wodnt arise. But the thing that needs fixing is the matchmaking system either introduce different tiers of pvp with say tier 1 highest enchantments r lessers and rank 6 and blue Artifact lock them at those level irrespective of enchantment level
  • izatarizatar Member, NW M9 Playtest Posts: 1,161 Arc User
    edited October 2014
    I think Neverwinter is using a statistical approximation that does as good as it can with limited information.

    It is not an exact calculation of competative rank.

    Cryptic actually made the right choice when it hid the rank numbers because the algorithm is designed to improve matchmaking, it is not for organizing the e-peens by size.

    So, the reason your rank can go down when you win against a much lower ranked team is because information is added that makes the calculation work out that way. So lewstelamon01's explanation is a relatively correct vast oversimplification of the problem.

    So, the solution is: stop taking the rank number so seriously. It is not a measure of your worth as a player or a person.
Sign In or Register to comment.