Something is rotten in the state of GAC

Prev1345
Acrofales
1363 posts Member
First off, huge congrats to the top 10 players, and particularly to THΞ WOLFP∆CK, who have 3 players in the top 10!

However, upon closer analysis, we see this cannot be "the 10 best SWGOH players" by any stretch of the imagination. I don't know what happened and am definitely not claiming any of these players cheated: matchmaking is just as suspect here. But in particular I wish to draw your attention to the #1 player, who has:
Zetas: 58
Arena ranks (squad/fleet): Rank 59 / Rank 54
Number of G11: 12
Number of G12: 88
Number of G13: 1
Number of 6 dot mods: 46
Number of 10+ speed mods: 193
Number of 15+ speed mods: 58
Number of 20+ speed mods: 14
Number of 25+ speed mods: 0
Number of 100+ offense mods: 31

There is simply no way this person faced competition at the level I did (I scraped into kyber). So, you say, 1 person lucked their way into a top spot? That isn't too bad.

So I draw your attention to the #5, #8 and #13 players:
#5:
Zetas: 60
Arena ranks (squad/fleet): Rank 13 / Rank 49
Number of G11: 46
Number of G12: 52
Number of G13: 5
Number of 6 dot mods: 62
Number of 10+ speed mods: 246
Number of 15+ speed mods: 84
Number of 20+ speed mods: 19
Number of 25+ speed mods: 1
Number of 100+ offense mods: 27
no Padme

#8:
Zetas: 66
Arena ranks (squad/fleet): Rank 17 / Rank 56
Number of G11: 73
Number of G12: 40
Number of G13: 4
Number of 6 dot mods: 38
Number of 10+ speed mods: 238
Number of 15+ speed mods: 63
Number of 20+ speed mods: 17
Number of 25+ speed mods: 1
Number of 100+ offense mods: 17

and #13:
Zetas: 53
Arena ranks (squad/fleet): Rank 17 / Rank 2
Number of G11: 97
Number of G12: 41
Number of G13: 2
Number of 6 dot mods: 44
Number of 10+ speed mods: 176
Number of 15+ speed mods: 38
Number of 20+ speed mods: 8
Number of 25+ speed mods: 2
Number of 100+ offense mods: 23
and Padme g8 no zetas.

No offense to these players. They probably played their hearts out. But between matchmaking and the specter of cheating, the leaderboard is meaningless. Seeing surprises at the top is great, and can give us hope that we can make it too. But in reality, match any of these with the majority of the players in the top 50 and they will get destroyed... and having 3 "upsets" in the top 10 and more continuing down doesn't mean there were awesome upsets: it means the ranking system is broken. Matchmaking, the point system and cheaters are ruining it.

Replies

  • leef
    13458 posts Member
    Chet manly is the man !
    Save water, drink champagne!
  • What tool have you used for the player analysis?
    "He's no good to me dead"
  • At this point, does anyone still think the leaderboard means anything? It's completely meaningless. At best it tells you how weak your opponents are
  • Excellent post. I think they did themselves a major disservice by creating a leaderboard and calling it a championship. If they just called it Progressive Grand Arena and left the rankings out of it, a lot of the disdain for the feature would have gone away. At least mine would have.

    Top end players like to compete but, based on your findings, that competition does not appear to be happening on an "even" playing field.

    I'd prefer they scrap the whole thing, go back to the old GA format (with whatever MM you want, couldn't care less at this point) with the old rewards structure.
  • Krjstoff
    633 posts Member
    edited August 2019
    Matchmaking works perfectly! GAC is a perfect way to establish who the top dog is! Anti-cheating measures work perfectly! I'm not crying, you're crying!
  • leef
    13458 posts Member
    leef wrote: »
    Chet manly is the man !

    different chet manly than i thought, still the man though !
    Save water, drink champagne!
  • leef
    13458 posts Member
    Top end players like to compete but, based on your findings, that competition does not appear to be happening on an "even" playing field.

    It's not that i disagree about the leaderboard being meaningless, but the the definition of "even playing field" ea/cg are using for the ranking is that everyone faces players with roughly equal rosters. So a "mediocre" roster ending up in the top of the leaderboard is arguably a argument in favour of the "even playing field".
    Save water, drink champagne!
  • Tiggus
    766 posts Member
    What are you talking about?

    I for one will cheer that no one from the top-10 guilds in term of GP has been admitted into the top-10 or this GA and will call THAT a refreshing "World's First", hahaa!

    Power to the Underdogs, yeah!

    ☮ Consular ☮ - https://swgoh.gg/u/tiggus/
  • Prosser
    278 posts Member
    Just an observation, but the guy in first has 88 g12. At his gp you have to field what, 8 teams? That means you need 80 toons at least, 40 def, and 40 offense, if you clear all in one battle. None of the rest of the ones you shared can field a full 80 g12+ toons on off and def, so they have to go weaker on def and keep a strong offense or strong def and have a weaker offense. He can do both. Plus if he is diligent to finish the feats, that will bolster his score more.
  • 3pourr2
    1927 posts Member
    Just because they bought the best characters and gear doesn’t mean they know how to play the game
  • No_Try
    4051 posts Member
    @Acrofales does your pertinent gp match with any of them? If not, naturally they didn't meet any competition such as you.

    One in the list is from my guild and I know how crazy tactical that person can play it to the utmost capability of his roster.
  • Kyno
    32087 posts Moderator
    There are several points to consider when looking at the scoring.

    Firstly we are still in an exhibition, this means that it is possible this will change as time moves on.

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    Third, as has been stated before, "top player" doesn't have any direct correlation to GP. This is what makes GP a good base for matching, it allows the player to shine through with skill.
  • leef
    13458 posts Member
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.
    Save water, drink champagne!
  • Kyno wrote: »
    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    I think you mean that in the context of GAC and the way we understand matchmaking to work it's not surprising. It definitely doesn't "make sense" if you consider this to be a competition and the leaderboard to be reflective of the best.
  • Acrofales
    1363 posts Member
    edited August 2019
    What tool have you used for the player analysis?

    DSR bot's recruit function.
    Prosser wrote: »
    Just an observation, but the guy in first has 88 g12. At his gp you have to field what, 8 teams? That means you need 80 toons at least, 40 def, and 40 offense, if you clear all in one battle. None of the rest of the ones you shared can field a full 80 g12+ toons on off and def, so they have to go weaker on def and keep a strong offense or strong def and have a weaker offense. He can do both. Plus if he is diligent to finish the feats, that will bolster his score more.

    Once again, I am not saying he didn't get there legitimately. I am saying his roster doesn't hold up against many other players who took GAC seriously. As for his roster... g12s are nice, but mods and zetas are the real dealbreaker. A 278 speed Darth Revan is not scaring anybody (just to be clear: my thrawn is faster than that), nor are a 244 speed Bossk without his unique zeta... or a Mother Talzin without her unique zeta.

    Now given the matchmaking system he may simply have gotten matched with other people who don't have the number of zetas or mods to compete with the rest of the top 50 in division 1. And that's completely fair, but that the ranking system subsequently "decides" he is the very best player in SWGOH means that ranking system is simply broken.
    Tiggus wrote: »
    What are you talking about?

    I for one will cheer that no one from the top-10 guilds in term of GP has been admitted into the top-10 or this GA and will call THAT a refreshing "World's First", hahaa!

    Power to the Underdogs, yeah!


    That is another way of looking at it. But does it really make sense to have a top 10 with players who have absolutely no chance at beating the real top rosters? And I'm not saying "pick a random whale". I am saying that players like Maraxus (top ranked TI player), Zetre (top ranked MAW player) or any of those Wolfpack players in the top 10 would wreck these players without breaking a sweat. In fact, I'd hazard a guess that MOST of the players from any of the guilds that dominate the monthly TW top 10 will all beat them without breaking a sweat.
    3pourr2 wrote: »
    Just because they bought the best characters and gear doesn’t mean they know how to play the game

    There are plenty of players who take PvP seriously and have rosters built for it, and not just random whales. In fact, there are whole guilds built around PvP.

    Now, once again, I am not claiming they climbed to their ranks illegitimately. Everybody got matched in CG's system and the points are scored the way they are. But clearly the ranking is not working.
    No_Try wrote: »
    @Acrofales does your pertinent gp match with any of them? If not, naturally they didn't meet any competition such as you.

    One in the list is from my guild and I know how crazy tactical that person can play it to the utmost capability of his roster.


    I didn't fight any of these players. I am also not accusing them of cheating. And in fact, I think it's great that there are upsets happening. But the ranking as is, unfortunately gives us no way of distinguishing someone who legitimately punches way above his weight class, someone who was matched with chumps and lucked his way to the top, and, yes, someone who cheated.


  • For it to be a fair system that would truly determine a top leaderboard, it would need to include 3 things:
    1. No cross division matchup.
    2. Elo based system.
    3. Recalculating the scoring system.

    Our current system is based largely on luck, both in who you're matched against and inside matches. And, oddly, more even opponents tend to lead to lower scores, not higher.
  • No_Try
    4051 posts Member
    I'll leave cheating aside as that's whole another subject. But what would really give us a clue someone at the top punched above their weight class by playing better? What kind of an mm system are imagining that which simply puts forth the match and not the rosters?

    Lastly it's obvious rosters that everyone would gawk at, those ones with 40-50 g13s are also getting matched with eachother. Why would anyone expect them to be at the top? Is this a -best roster to exist in swgoh- tournament?
  • Boros
    507 posts Member
    Excellent post. I think they did themselves a major disservice by creating a leaderboard and calling it a championship. If they just called it Progressive Grand Arena and left the rankings out of it, a lot of the disdain for the feature would have gone away. At least mine would have.

    Top end players like to compete but, based on your findings, that competition does not appear to be happening on an "even" playing field.

    I'd prefer they scrap the whole thing, go back to the old GA format (with whatever MM you want, couldn't care less at this point) with the old rewards structure.

    Old reward was better
  • Acrofales
    1363 posts Member
    No_Try wrote: »
    I'll leave cheating aside as that's whole another subject. But what would really give us a clue someone at the top punched above their weight class by playing better? What kind of an mm system are imagining that which simply puts forth the match and not the rosters?

    Lastly it's obvious rosters that everyone would gawk at, those ones with 40-50 g13s are also getting matched with eachother. Why would anyone expect them to be at the top? Is this a -best roster to exist in swgoh- tournament?

    Well, something ELO-like would go a long way. I know CG said there were weird knock-on effects that made it perform poorly, but the mathematician in me doesn't understand their reasoning. Yes, it's a collection game, and growing your collection is a way of improving. But so are Hearthstone and to a certain extent LoL and they use MMR with great success. Moreover, what is so categorically different between "growing your collection" and "increasing in skill" that invalidates the use of ELO? The main thing I can think of is the release of new legendary/mythic characters that shake up the meta completely (JKR, Malak). That will no doubt cause instability in the ranking. Maybe have these coincide with seasonal ladder resets, when you shake up the ladder anyway?

    As for rosters everybody gawks at vs minnows: if the minnows can't fight the krakens, they shouldn't be in the same league, should they? I mean, the little league champion is a real accomplishment, but them beating the Baltimore Orioles would still be a giant upset, let alone challenge any of the actual contenders for the World Series, right?
  • Kyno
    32087 posts Moderator
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.
  • ReveNCZ
    52 posts Member
    edited August 2019
    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins
    mid level GP and this slightly easier defenses made me laugh a lot :D not sure if serious or trolling...really. i am mid level GP (not even 4.8M before GAC started), atm 7k char GP less than the first guy in kyber and i would crush him in GA. yet in first two GA i was lowest GP in our brackets, 200-400k less char GP than others and big g13/g12/zeta/6E disadvantage and a lot of players had very similar or even worse matches...and i lost like only 1-2 single GA matches before championship, but here in GAC literally no chance to beat guys like i met. mid level GP facing slightly easier defenses...guess chance like to get full carb from bronziums...
  • Kyno
    32087 posts Moderator
    jhbuchholz wrote: »
    Kyno wrote: »
    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    I think you mean that in the context of GAC and the way we understand matchmaking to work it's not surprising. It definitely doesn't "make sense" if you consider this to be a competition and the leaderboard to be reflective of the best.

    Why not, to get there that player had to win almost the exactly the same amount as the players around them, and then also found ways to score more points within the given system.

    You dont have to have the beat vocabulary to win at scrabble, a little luck and the better understanding of the point system will prevail.

    The major driving force is absolutely the win record, after that there is some "skill" to gaining points.
  • Acrofales
    1363 posts Member
    Kyno wrote: »
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.

    Lets put it this way: at the very least it places serious questions what the developers understand as "being good" at their game.

    I mean.... anybody can come up with a scoring system. The developers *could* have given scores for the number of times you used an ugnaught on defense, or the number of teams you beat using Tuskens.

    That would definitely be a scoring system. But if the developers tried to claim that that is the score that determines who is the best player in SWGOH, I don't think you'd find many players agreeing with them. And that is what is at stake here.
  • No_Try
    4051 posts Member
    Acrofales wrote: »
    No_Try wrote: »
    I'll leave cheating aside as that's whole another subject. But what would really give us a clue someone at the top punched above their weight class by playing better? What kind of an mm system are imagining that which simply puts forth the match and not the rosters?

    Lastly it's obvious rosters that everyone would gawk at, those ones with 40-50 g13s are also getting matched with eachother. Why would anyone expect them to be at the top? Is this a -best roster to exist in swgoh- tournament?

    Well, something ELO-like would go a long way. I know CG said there were weird knock-on effects that made it perform poorly, but the mathematician in me doesn't understand their reasoning. Yes, it's a collection game, and growing your collection is a way of improving. But so are Hearthstone and to a certain extent LoL and they use MMR with great success. Moreover, what is so categorically different between "growing your collection" and "increasing in skill" that invalidates the use of ELO? The main thing I can think of is the release of new legendary/mythic characters that shake up the meta completely (JKR, Malak). That will no doubt cause instability in the ranking. Maybe have these coincide with seasonal ladder resets, when you shake up the ladder anyway?

    As for rosters everybody gawks at vs minnows: if the minnows can't fight the krakens, they shouldn't be in the same league, should they? I mean, the little league champion is a real accomplishment, but them beating the Baltimore Orioles would still be a giant upset, let alone challenge any of the actual contenders for the World Series, right?

    I don't do irrelevant sports corellations, sorry. It's just too context specific to fall on that trap.

    Not against an elo-like system whatsoever. But that doesn't get rid of the necessity for further matchmaking either. Or do you just randomly bunch up people and expect winning ratios to sort it out? This is not chess and everyone is not playing by the same pieces obviously. No matter whatever is further implemented, matchmaking itself needs improvement.

    Currently we have a league-based system which is very abruptly elo-like. You can only get matched inside your league...which then only kicks into real effect in later weeks as there are less accomplishers in higher leagues. And even then it's also mingled with cross-division still.

    Going elo-like success based system persistent through seasons, I'm neutral about that. It has to be pondered on really well. i.e. take a case where a new starter spender has been very successful with a beginner roster in a season, that sets his precedent which he has to carry out in following seasons...and he keeps spending which gives the roster mobility in divisions reaching far tougher opponents, and now he's getting matched with people above his punching power due to previous success. I dunno if a persistent elo is the right choice for an actively developing game, namely this particular one and not X football team, or go championship or whatever.
  • Kyno
    32087 posts Moderator
    evoluza wrote: »
    Kyno wrote: »
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.

    You can't call the leaderboard a competition if some in front can't knock back others on there.

    Really?

    In many sports there is a time, usually as the finals are approaching (but not always)that some teams no matter what they do cant effect others whether they win or lose and those teams could have beaten the top team earlier in the season.

    This leader board just represents the pointed gained. It doesn't necessarily represent who would win when face to face. I'm not saying this represents a player to player "skill", this represents a player to player skill of this game mode as a whole.
  • Kyno
    32087 posts Moderator
    Acrofales wrote: »
    Kyno wrote: »
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.

    Lets put it this way: at the very least it places serious questions what the developers understand as "being good" at their game.

    I mean.... anybody can come up with a scoring system. The developers *could* have given scores for the number of times you used an ugnaught on defense, or the number of teams you beat using Tuskens.

    That would definitely be a scoring system. But if the developers tried to claim that that is the score that determines who is the best player in SWGOH, I don't think you'd find many players agreeing with them. And that is what is at stake here.

    Over time the lifetime board will represent that. Will it represent it at the very minute level of who is first or 5th, no I dont think it ever will, but that is a product of the tournament style, since we will never be able to pit each player against each other the way its currently set up.
  • Acrofales
    1363 posts Member
    edited August 2019
    Kyno wrote: »
    evoluza wrote: »
    Kyno wrote: »
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.

    You can't call the leaderboard a competition if some in front can't knock back others on there.

    Really?

    In many sports there is a time, usually as the finals are approaching (but not always)that some teams no matter what they do cant effect others whether they win or lose and those teams could have beaten the top team earlier in the season.

    This leader board just represents the pointed gained. It doesn't necessarily represent who would win when face to face. I'm not saying this represents a player to player "skill", this represents a player to player skill of this game mode as a whole.

    It doesn't even do that accurately, because I am quite confident the #2 would kick the #1s derriere (sorry for the earlier word) in the GA "game mode". But because the matchmaking never matched them, and the #1 probably had far more fortuitous matches, we have this higgledypiggledy "leaderboard".
  • leef
    13458 posts Member
    Kyno wrote: »
    leef wrote: »
    Kyno wrote: »

    Secondly, it actually makes sense that from a division a "mid level" GP would have a chance to get more points than the "top dogs", some possible reasons are- facing slightly easier defenses, which could lead to more feats being completed, more points gained from teams/zones cleared, and general wins

    That may make sense, but it's definately an argument in favour of the leaderboard being meaningless.

    Meaningless, not entirely. It's just not going to show what players expect. That doesnt mean that the player in position 1 didnt complete more feats and score more points in the current system we have.

    The competition as a whole is a series of ways to score points, some people will always be better at that then others. Winning is a major point system, but after that it breaks down to a player by player basis.

    I don't think the difference between "meaningless" and "not entirely meaningless" matters when it comes to leaderboards.
    I'm not sure where you're going with the whole showing something other than what players expect. I thought it was pretty obvious that you were giving reasons as to why some players got a higher rank which had nothing to do with player skill. Some players being able to earn points more easily due to the circumstances than other players within the same division makes the leaderboard (not entirely) meaningless.
    Save water, drink champagne!
  • No_Try wrote: »
    Currently we have a league-based system which is very abruptly elo-like. You can only get matched inside your league...which then only kicks into real effect in later weeks as there are less accomplishers in higher leagues. And even then it's also mingled with cross-division still.

    thats not true, at the start of 3rd GA me and others were chromium and one guy was already in aurodium league
Sign In or Register to comment.