|
On November 25 2013 21:56 tili wrote:Show nested quote +On November 24 2013 11:24 Thieving Magpie wrote: It took months for the "bronze league/D level" players on iccup to win their first games let alone to start stabilizing their rankings. It sounds like a matter of perception, players whining wants to believe that silver through diamond should be relatively easy when in truth it should be incredibly difficult to get into silver that way being "plat" or "masters" becomes almost legendary in status. Everyone will eventually be pushed into bronze/silver and only top level players should be able to get to gold and only exceptional players can make it to diamond. Why? Why make the learning curve so steep that no one feels like they are progressing? Plat doesn't need to be legendary; that's what masters and GM are for. That's not a sustainable game model. You're redefining what it means to be silver. Blizzard's algorithm is supposed to roughly create the following partitions (see below). If it's not doing that, then that is a problem. 8% - Bronze 20% - Silver 32% - Gold 20% - Platinum 18% - Diamond 2% - Masters 200 - GM Also, the matchmaking system should give you a 40%-60% chance to win. Not a 20%-80%. If MMR decay is significantly hampering the quality of match ups (as measured by the chance of each to win the game), then that sucks.
Then where we differ is that, to me, the numbers are arbitrary. If gold/plat is nothing but "masters players" then that will eventually be the new definition of being gold/plat. If "masters is so easy anyone can do it" then that is the definition of masters.
It's more important to accept a ladder system than to worry about what color it gives you.the best way to learn is to play versus players who are better than you.
|
The MMR decay is incredibly aggressive. I was Plat by the end of WoL before the HotS release and then spent most of my time at the start of HotS screwing about against Silver/Gold players and off-racing a bit.
I started playing again the other week finding myself matched up against Bronzes; so I played a couple games pulling off random, incredibly sloppy all-ins and winning regardless. Just complete stomps.
I feel bad about it, and the poor guys have all been incredibly polite (surprisingly), but I can't do much about it. I'm matched where I'm matched until the system corrects itself. I like the idea of preventing people from holding onto high leagues by just playing 1 game per season for years on end; but by the same token its brutal for those matched up against people far stronger.
|
That is pretty interesting. I suspected these things you mentioned in the OP. I decayed from Rank 1 Diamond to Silver lol. It is kind of a blessing because now I can do random things and win, like reactor reapers.
|
On November 25 2013 23:55 Thieving Magpie wrote:Show nested quote +On November 25 2013 21:56 tili wrote:On November 24 2013 11:24 Thieving Magpie wrote: It took months for the "bronze league/D level" players on iccup to win their first games let alone to start stabilizing their rankings. It sounds like a matter of perception, players whining wants to believe that silver through diamond should be relatively easy when in truth it should be incredibly difficult to get into silver that way being "plat" or "masters" becomes almost legendary in status. Everyone will eventually be pushed into bronze/silver and only top level players should be able to get to gold and only exceptional players can make it to diamond. Why? Why make the learning curve so steep that no one feels like they are progressing? Plat doesn't need to be legendary; that's what masters and GM are for. That's not a sustainable game model. You're redefining what it means to be silver. Blizzard's algorithm is supposed to roughly create the following partitions (see below). If it's not doing that, then that is a problem. 8% - Bronze 20% - Silver 32% - Gold 20% - Platinum 18% - Diamond 2% - Masters 200 - GM Also, the matchmaking system should give you a 40%-60% chance to win. Not a 20%-80%. If MMR decay is significantly hampering the quality of match ups (as measured by the chance of each to win the game), then that sucks. Then where we differ is that, to me, the numbers are arbitrary. If gold/plat is nothing but "masters players" then that will eventually be the new definition of being gold/plat. If "masters is so easy anyone can do it" then that is the definition of masters. It's more important to accept a ladder system than to worry about what color it gives you.the best way to learn is to play versus players who are better than you. It would not necessarily be a bad thing if this would only be a distribution change and matchmaker would give consistently similar level of opponents. The problem is that these distribution changes are caused by MMR decay and not by changed league offsets. This means that masses of players have had their MMR value tampered (lowered). Thus in different MMR ranges there are now wildly different skilled people sharing the similar MMR. As result you cannot deduct the skill level of your opponent in the beginning of match and thus cannot adjust your strategies based on that. You can read an example I wrote to the 'Is the ladder getting harder?' thread earlier.
Also this link takes you to a table that shows how league distributions have changed from start of last season to this season. From it you can easily deduct that major percentage of 1v1 ladder population goes inactive for 2 weeks or more during each season and are thus affected by the MMR decay. As a result player population shifts towards the lower leagues. And yes, Blizzard measures the league distribution by 'active players' (those sc2ranks figures include all players who have played their 1v1 placements). But that measurement does not tell how messed up the ladder actually is (wildly different skill level people sharing similar MMR in different MMR ranges).
|
Not fun at all. I used to be top Master in WOL, I stopped playing before HOTS until now. I was expecting to play my 5 games and be Plat or something like that and then stabilize at Diamond. But I ended up being Gold, and after 61 games I am 32/29.
I know my skill is not the same and people got better at the game, but seriously I feel no progress at all and I'm clueless who am I really playing (some times other masters in the same situation, other times truly gold players who will be demoted the next season because of players like me).
There is no point to have leagues if everybody is in the same league.
|
Seems like getting promoted to plat, diamond and master is pretty hard now! You all have my respect! 140k+ players in EU ladder is nice, no chance of the game dying any time soon.
|
Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make
|
On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make
Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem.
A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze.
A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week.
No MMR decay means that players stay on masters playing 1-3 games a season.
Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken.
|
On November 26 2013 02:28 Thieving Magpie wrote:Show nested quote +On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem. A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze. A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week. No MMR decay means that players stay on masters playing 1-3 games a season. Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken.
It's pretty clearly broken? The percentages are all off and blizzard has acknowledged there's an issue, so how could it not be broken? Their methods are not leading to their intended result, is that not the very definition of it being broken?
You can explain and justify the system and decay as it is now all you want, but the fact of the matter is the current method is leading to wrong results, so by the very definition of the word, the system is broken and needs fixing. In this case, the fixing is likely the method, which will have to be re-looked at.
|
United States12181 Posts
On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make
I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one.
As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons.
The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists.
|
On November 26 2013 03:38 Excalibur_Z wrote:Show nested quote +On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one. As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons. The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists.
Do you really think that many players have quit in the past 2-3 seasons though? Honest questions. Would seem hard to believe.
|
On November 26 2013 03:25 ffadicted wrote:Show nested quote +On November 26 2013 02:28 Thieving Magpie wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem. A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze. A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week. No MMR decay means that players stay on masters playing 1-3 games a season. Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken. It's pretty clearly broken? The percentages are all off and blizzard has acknowledged there's an issue, so how could it not be broken? Their methods are not leading to their intended result, is that not the very definition of it being broken? You can explain and justify the system and decay as it is now all you want, but the fact of the matter is the current method is leading to wrong results, so by the very definition of the word, the system is broken and needs fixing. In this case, the fixing is likely the method, which will have to be re-looked at.
Getting unwanted results =/= broken system.
Getting unwanted results is simply getting unwanted results which means desiring a different system. The disdain for the specific set of results is a philosophical issue. Blizzard first gave us a system allowing players who don't play everyday to maintain their ranks. People complained, so Blizz added decay. People now complain since decay means they actually have to keep up with other ladder players. Blizzard will change the system again and people will come back to complain about it.
There is no "correct" system since the problem is not with the numbers but with the philosophy behind the numbers. Add a decay and people who don't play often whine. Remove decay as people who play often whine. Get a middle ground and both sides will whine that the game is too easy. And so on and so forth.
This is a player base problem, not a blizzard one.
|
On November 26 2013 03:43 ffadicted wrote:Show nested quote +On November 26 2013 03:38 Excalibur_Z wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one. As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons. The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists. Do you really think that many players have quit in the past 2-3 seasons though? Honest questions. Would seem hard to believe.
There was another thread about the ladder getting harder and someone compiled a graph from nios, over 2-3 seasons it "appeared" (again, not really sure of the source) that about half as many people play ladder now.
I think the ladder is working the way that Blizzard had originally intended, but if you have that many less players, I would assume that a lot of players would get shafted to higher percentage leagues, just to make the system happy. Of course this doesn't affect the 2%'s of our world, but it's definitely discouraging to the lower player, as the numbers would clearly indicate, if correct of course.
I would personally love to see some new features put into the ladder system. Just so many hidden variables, etc, it's seemingly impossible for a new player to grasp. For simplicity, I would love to see a "Promotion" game, where in, you would play a set of 5 matches vs opponents in a higher league, if you win over 50% of them (3) you get promoted.
Of course, this probably isn't all that realistic, I'm just throwing an example of how they can be more up-front about the promotion system, and at least give players something to look forward to.
|
On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make
You have any link for that?
|
On November 26 2013 03:57 Thieving Magpie wrote:Show nested quote +On November 26 2013 03:25 ffadicted wrote:On November 26 2013 02:28 Thieving Magpie wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem. A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze. A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week. No MMR decay means that players stay on masters playing 1-3 games a season. Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken. It's pretty clearly broken? The percentages are all off and blizzard has acknowledged there's an issue, so how could it not be broken? Their methods are not leading to their intended result, is that not the very definition of it being broken? You can explain and justify the system and decay as it is now all you want, but the fact of the matter is the current method is leading to wrong results, so by the very definition of the word, the system is broken and needs fixing. In this case, the fixing is likely the method, which will have to be re-looked at. Getting unwanted results =/= broken system. Getting unwanted results is simply getting unwanted results which means desiring a different system. The disdain for the specific set of results is a philosophical issue. Blizzard first gave us a system allowing players who don't play everyday to maintain their ranks. People complained, so Blizz added decay. People now complain since decay means they actually have to keep up with other ladder players. Blizzard will change the system again and people will come back to complain about it. There is no "correct" system since the problem is not with the numbers but with the philosophy behind the numbers. Add a decay and people who don't play often whine. Remove decay as people who play often whine. Get a middle ground and both sides will whine that the game is too easy. And so on and so forth. This is a player base problem, not a blizzard one. No it is Blizzard's problem. With the changes they made (added too steep maximum decay to a game, where a large portion of the player base goes inactive from time to time), they caused a cascading effect. MMR often does not represent skill anymore. Regarding matchmaking SC2 was arguably one of the best in the world. It is not anymore. Even if the matchmaker is working technically fine, the matchmaking results are often not fine (the matchmaker thinks that players A and B who have similar MMR are equally skilled, even if player A might belong typically to that level (e.g. gold) and the more skilled player B has just decayed there (e.g. from master).
Even if the player base at the moment is healthy, due to the current situation (not to mention that the situation gets worse as time passes), players might start quitting in accelerating pace.
Edit: Changed "Even if the matchmaker is working fine ... " to "Even if the matchmaker is working technically fine ... " Edit 2: added missing 'due to the' to 'current situation'
|
United States12181 Posts
On November 26 2013 03:43 ffadicted wrote:Show nested quote +On November 26 2013 03:38 Excalibur_Z wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one. As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons. The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists. Do you really think that many players have quit in the past 2-3 seasons though? Honest questions. Would seem hard to believe.
Well let's see. I looked up the numbers based on snapshots from archive.org so they're not perfect.
Season 14 (Aug 19 2013 [ends Aug 26 2013], SC2Ranks 2.0 overhaul): 471,590 Season 15 (Nov 7 2013 [ends Nov 11 2013]): 354,694 Season 16 (Nov 2013, today): 182,591
Now, under the old SC2Ranks 1.0, it kept a running tally of all players that it had within its system, including new team compositions. Under SC2Ranks 2.0 it separates by game mode, so that 471,590 and 182,591 are the number of global 1v1 players. Archive.org doesn't really let you freely browse around because it doesn't capture every page, so what I can get is limited
The other thing is that the SC2Ranks 2.0 numbers are seasonal. When the new season starts, the numbers get reset to 0, and they'll increase over time until the next season roll. So, a week before 2013 Season 4/Season 14 ended, there were 471,590 1v1 players. Half a week before 2013 Season 5/Season 15 ended, there were 354,694 1v1 players. Today, about 2 weeks into the new season, there are 182,591 1v1 players so that doesn't really say too much yet.
|
On November 26 2013 04:38 anessie wrote:Show nested quote +On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make You have any link for that?
http://us.battle.net/sc2/en/forum/topic/10636803688
On November 26 2013 04:48 Excalibur_Z wrote:Show nested quote +On November 26 2013 03:43 ffadicted wrote:On November 26 2013 03:38 Excalibur_Z wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one. As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons. The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists. Do you really think that many players have quit in the past 2-3 seasons though? Honest questions. Would seem hard to believe. Well let's see. I looked up the numbers based on snapshots from archive.org so they're not perfect. Season 14 (Aug 19 2013 [ends Aug 26 2013], SC2Ranks 2.0 overhaul): 471,590 Season 15 (Nov 7 2013 [ends Nov 11 2013]): 354,694 Season 16 (Nov 2013, today): 182,591 Now, under the old SC2Ranks 1.0, it kept a running tally of all players that it had within its system, including new team compositions. Under SC2Ranks 2.0 it separates by game mode, so that 471,590 and 182,591 are the number of global 1v1 players. Archive.org doesn't really let you freely browse around because it doesn't capture every page, so what I can get is limited The other thing is that the SC2Ranks 2.0 numbers are seasonal. When the new season starts, the numbers get reset to 0, and they'll increase over time until the next season roll. So, a week before 2013 Season 4/Season 14 ended, there were 471,590 1v1 players. Half a week before 2013 Season 5/Season 15 ended, there were 354,694 1v1 players. Today, about 2 weeks into the new season, there are 182,591 1v1 players so that doesn't really say too much yet.
I see... I guess there definitely is a noticible player decrease from s14 to s15. On korona's point above, this is what I'm getting at... I know the system is "working as written" or however you want to call it, but there's no way these were the intended results when they did write them. The match-making does seem pretty poor right now, from reports and from personal experience. And as with any other system, when the intended result isn't achieve, generally tweaks need to be made (whether they be fixes or changes) to achieve that result. That's all I hope blizzard has recognized at this point.
|
On November 26 2013 04:46 korona wrote:Show nested quote +On November 26 2013 03:57 Thieving Magpie wrote:On November 26 2013 03:25 ffadicted wrote:On November 26 2013 02:28 Thieving Magpie wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem. A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze. A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week. No MMR decay means that players stay on masters playing 1-3 games a season. Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken. It's pretty clearly broken? The percentages are all off and blizzard has acknowledged there's an issue, so how could it not be broken? Their methods are not leading to their intended result, is that not the very definition of it being broken? You can explain and justify the system and decay as it is now all you want, but the fact of the matter is the current method is leading to wrong results, so by the very definition of the word, the system is broken and needs fixing. In this case, the fixing is likely the method, which will have to be re-looked at. Getting unwanted results =/= broken system. Getting unwanted results is simply getting unwanted results which means desiring a different system. The disdain for the specific set of results is a philosophical issue. Blizzard first gave us a system allowing players who don't play everyday to maintain their ranks. People complained, so Blizz added decay. People now complain since decay means they actually have to keep up with other ladder players. Blizzard will change the system again and people will come back to complain about it. There is no "correct" system since the problem is not with the numbers but with the philosophy behind the numbers. Add a decay and people who don't play often whine. Remove decay as people who play often whine. Get a middle ground and both sides will whine that the game is too easy. And so on and so forth. This is a player base problem, not a blizzard one. No it is Blizzard's problem. With the changes they made (added too steep maximum decay to a game, where a large portion of the player base goes inactive from time to time), they caused a cascading effect. MMR often does not represent skill anymore. Regarding matchmaking SC2 was arguably one of the best in the world. It is not anymore. Even if the matchmaker is working fine, the matchmaking results are often not fine (the matchmaker thinks that players A and B who have similar MMR are equally skilled, even if player A might belong typically to that level (e.g. gold) and the more skilled player B has just decayed there (e.g. from master). Even if the player base at the moment is healthy, the current situation (not to mention that the situation gets worse as time passes), players might start quitting in accelerating pace.
Higher level players unable to keep up with a new logarithmic paradigm is not a flaw of the system. Lower level players upset that higher level players are no longer able to keep up with said paradigm is also not a flaw in the system since the system rewards regular play and punishes sparse play.
People getting butt hurt by this new system is not a flaw or mistake of the system, it is a disdain of the market reacting to an unwanted revelation presented by the product. Ladder decay will always pull down masters level players to lower leagues and it will always "force" supposedly "lower level" players to face people who are more difficult that previously faced. This will eventually push down lower level players even lower as the upper echelons of the system become harder to reach--this is the result that will always be created by MMR decay.
People disliking this result is personal problem, not a Blizzard mistake because the system as is works just as an MMR decay system should. Right now, playing once a week is considered often enough to not suffer decay. If you only play 2-3 times a month is when the ladder decay hits you. Currently only those types of players are dropping down.
People realizing that MMR decay actually sucks for you even more than if MMR decay was not present is a negative reaction of the public to a successful implementation of a product. Whether that product is something Blizzard wishes to stand by or not is a different issue entirely.
|
On November 26 2013 04:30 Ctone23 wrote:Show nested quote +On November 26 2013 03:43 ffadicted wrote:On November 26 2013 03:38 Excalibur_Z wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make I don't want there to be any confusion here, but this is not a matter of the community informing Blizzard about some new development and Blizzard doing something about it. They already have all the data that korona and I have been presenting, and much more. They know the player burnout rate, they know the activity levels of players, and they know the community perceives that there is an issue with the matchmaker and the league distribution. These complaints by the community aren't new, they're only more commonplace because of threads like this one. As best I can tell, the matchmaker is still functioning properly in that it's pairing one player's number with another player's similar number. But, as korona said above, the definitions of those numbers can vary greatly due to the rate of decay. Some players think they're decaying but aren't ("I only get to play once a week" means you're not decaying). Nevertheless, all of these reports contribute to a collective perception that something is wrong, which can be just as problematic to the developers as something that's actually wrong for various reasons. The league distributions, because of the activity metric, also are probably functioning properly as I point out in that other thread. That thread was created after decay had been active for 4 seasons. Now: is the decay causing deflation? That's hard to say and will require a new snapshot. What I do know is that after 4 seasons it still looked pretty close to the target distribution for active players even though the distribution across all players was quite different. Another two seasons probably wouldn't change it that much, if at all. What's really suffering here is the size of the playerbase. If players play the game to go up in leagues, but they don't play very often, they're never going to get promoted and may even get demoted, which is discouraging, and possibly reason to quit altogether. Is that what's happening? Not sure, all I can present are the numbers and not the psychology. It's the same for the developers, really. They have to make a judgment based on what might be a psychological problem even if no mechanical problem exists. Do you really think that many players have quit in the past 2-3 seasons though? Honest questions. Would seem hard to believe. There was another thread about the ladder getting harder and someone compiled a graph from nios, over 2-3 seasons it "appeared" (again, not really sure of the source) that about half as many people play ladder now. Interpretation of that graph was invalid. The first figures were not from the start of S14 (S4/2013), but from the end of S13 (S3/2013). The second numbers were from the beginning of this season S16 (S6/2013). You cannot directly compare numbers from the end of one season to numbers from start of another season.
But if we compare the numbers from the end of S13 (S3/2013) to numbers from end of last season S15 (S5/2013) we notice that the player base has declined only little: http://www.teamliquid.net/forum/viewpost.php?post_id=20273329
Of course S13 was much shorter season. Thus if it would have been as long as S15, more players would have played their 1v1 placements. Thus the actual decline is likely larger than the numbers suggest, but not as big as the graph you mentioned suggested. Also the numbers don't tell how active the players were / are. In general each player might have played more games per certain time frame during earlier seasons than now.
-- Edit: End of S14 numbers might be available in Excalibur_Z:s post (if sc2ranks numbers are only from that season. sc2ranks 1.0 calculated also people from previous seasons who had not played their placements yet): http://www.teamliquid.net/forum/viewmessage.php?topic_id=429734¤tpage=16#316
S14 & S15 were equal length seasons. Of course S14 happened during holiday season.
|
On November 26 2013 04:55 Thieving Magpie wrote:Show nested quote +On November 26 2013 04:46 korona wrote:On November 26 2013 03:57 Thieving Magpie wrote:On November 26 2013 03:25 ffadicted wrote:On November 26 2013 02:28 Thieving Magpie wrote:On November 26 2013 01:20 ffadicted wrote: Blizzard has seemed to acknowledge (through bnet MVP posters) that there is indeed a problem with the ladder right now and they're looking into it. I guess that can put to rest the argument of "ladder is fine players just got better" some were trying to make Players have gotten better, but players getting better is not the cause of MMR decay. MMR decay simply shifts the focus of placement to that of regular players as opposed to rarely logs in players. It's not a balance problem, or a design problem, it's a philosophy problem. A harsh MMR decay means that you need to be good AND play regularly to be at the top. That means good players who don't play regularly eventually crowd the middle shoving former plat/gold/silver players down to Bronze. A weak MMR decay means that you will have players who maintain masters/Grandmaster positions only playing a few times a week. No MMR decay means that players stay on masters playing 1-3 games a season. Which philosophical view do you prefer in a ladder system? I don't care, number rankings are arbitrary. But the problem with the current ladder is not that it is broken. It's pretty clearly broken? The percentages are all off and blizzard has acknowledged there's an issue, so how could it not be broken? Their methods are not leading to their intended result, is that not the very definition of it being broken? You can explain and justify the system and decay as it is now all you want, but the fact of the matter is the current method is leading to wrong results, so by the very definition of the word, the system is broken and needs fixing. In this case, the fixing is likely the method, which will have to be re-looked at. Getting unwanted results =/= broken system. Getting unwanted results is simply getting unwanted results which means desiring a different system. The disdain for the specific set of results is a philosophical issue. Blizzard first gave us a system allowing players who don't play everyday to maintain their ranks. People complained, so Blizz added decay. People now complain since decay means they actually have to keep up with other ladder players. Blizzard will change the system again and people will come back to complain about it. There is no "correct" system since the problem is not with the numbers but with the philosophy behind the numbers. Add a decay and people who don't play often whine. Remove decay as people who play often whine. Get a middle ground and both sides will whine that the game is too easy. And so on and so forth. This is a player base problem, not a blizzard one. No it is Blizzard's problem. With the changes they made (added too steep maximum decay to a game, where a large portion of the player base goes inactive from time to time), they caused a cascading effect. MMR often does not represent skill anymore. Regarding matchmaking SC2 was arguably one of the best in the world. It is not anymore. Even if the matchmaker is working fine, the matchmaking results are often not fine (the matchmaker thinks that players A and B who have similar MMR are equally skilled, even if player A might belong typically to that level (e.g. gold) and the more skilled player B has just decayed there (e.g. from master). Even if the player base at the moment is healthy, the current situation (not to mention that the situation gets worse as time passes), players might start quitting in accelerating pace. Higher level players unable to keep up with a new logarithmic paradigm is not a flaw of the system. Lower level players upset that higher level players are no longer able to keep up with said paradigm is also not a flaw in the system since the system rewards regular play and punishes sparse play. People getting butt hurt by this new system is not a flaw or mistake of the system, it is a disdain of the market reacting to an unwanted revelation presented by the product. Ladder decay will always pull down masters level players to lower leagues and it will always "force" supposedly "lower level" players to face people who are more difficult that previously faced. This will eventually push down lower level players even lower as the upper echelons of the system become harder to reach--this is the result that will always be created by MMR decay. People disliking this result is personal problem, not a Blizzard mistake because the system as is works just as an MMR decay system should. Right now, playing once a week is considered often enough to not suffer decay. If you only play 2-3 times a month is when the ladder decay hits you. Currently only those types of players are dropping down. People realizing that MMR decay actually sucks for you even more than if MMR decay was not present is a negative reaction of the public to a successful implementation of a product. Whether that product is something Blizzard wishes to stand by or not is a different issue entirely.
I think you're arguing different points here entirely. Nobody is saying there was like an error in the MMR calculation code the blizzard made or a bug in the system that caused all of this. All we're saying is that the system isn't producing the correct outcomes. You want people to accept the new outcomes, which isn't really fair to ask when the old outcome produced even matches almost every time and distributed leagues accordingly, and the new one is pretty awful at producing even matches and seems to have thrown the league distributions into whack.
|
|
|
|