GAC Squish

Same happens to me. 

Very frustrating, instead of getting into K1 I almost dropped to K3. Please return us out logs rating points and stop squishing at all.

I’d love an update from CG on their goals with the squish because the vagueness is frustrating.  But it looks like they overshot their initial goal, especially with regards to Kyber.  At the start of June, it was nearly twice the size of Bronzium, Chromium or Aurodium (~85k players compared to ~40-50k for the other leagues).  Now that they’ve got Carbonite back to a reasonable size, it looks like they’re trying to reel back in the size of Kyber, and get closer to an even distribution between leagues.

I would like to see a response too. On the old forums I have expressed concerns about the squish from day 1 (often against a chorus of people quoting very poorly grasped mathematics at me). The underlying problem as it appears to me isn’t that the maths isn’t working but that CG have applied a mathematical solution (admittedly badly) to an arithmetic problem - and one that could quite possibly have been solved better by doing nothing

In game looking at the top 50 leaderboards it is quite confusing.  There are numerous high scores without rosters that would not seem to fit Kyber 1.  (Numbers are post-squish) Anyway, the highest is 4855.  Kyber 2 the highest is 3534 (the top end is supposed to be 3610 which will soon have players from Kyber 1).  Kyber 3 the highest is 3249.  (3309 is what the top end is supposed to be.)  Kyber 4 has 2 players at 3205 and 3203 and then a big grouping at 3078.  Kyber 5 has 2926 at the top.

Then it gets really confusing as the top end of Aurodium 1 has a group of players from 4961 to 4379 and then it drops to 3707.  Aurodium 2 has 2 players at 3005 and 2946 before dropping to 2900.  Aurodium 3 has 2 players at 2925 and 2835 before dropping to 2780.  Aurodium 4 has someone at 2711 before a large group at 2680.  Aurodium 5 has two players at 2624 and 2598 before a group at 2558.    Then Chromium 1 has 7 players at 5000 before dropping to 2876, then 2793. 

I don't understand those scores and it looks like there will be massive movements of players between levels, many of which may find themselves completely overwhelmed by their opposition soon.  

I did try to look at numbers over on swgoh.gg but it is not retrieving them well at this time.   

Anyway, some explanation as to what is supposed to be the situation and what is actually being done would be appreciated.

The players with weird roster/scores that don’t fit their league is due to how the squish has always treated players who stop playing for extended periods.

After a few weeks of non-participation players stop losing points as though they were getting losses.  But they still have their SR impacted by the squish each month.  Depending on what league they stopped playing in, this can result in them getting an increasing number of bonus SR every month until they start playing again.

Eventually people max out at 5000 SR, when they rejoin they’re in whatever league they left in.  They mostly get matched with other people just rejoining.  Since you can only be promoted one league at a time, they’ll move up the ranks one league at a time until they hit the top of Kyber, and eventually they fall low enough to get matched against real K1 accounts, at which point they’ll just stop winning till they get back to an appropriate ranking.

They absolutely need the squish.  Without it Carbonite would be the largest league by an overwhelming margin.  The primary function of the squish across all of the leagues (not just looking at Kyber) from day one has been to move new players up the ranks and redistribute them upwards.  Without it more than half the player base would be clustered in Carbonite. 

You mean redistribute downwards, don’t you?  How many people have been squished from Kyber 2 into Kyber 1?  Personally, I’ve been squished from Kyber 1 to Kyber 2 multiple times. 

Nope.  On average across all of GAC the squish has always moved people upwards, sometimes (like during the recent Oct-May squish) by massive amounts.  The top of Kyber has always been negative, but at the macro level, the squish moves the distribution of players up even if it pulls the top end back to the rest of the pack.

I mean, we can't have everyone in Kyber 1, that makes no sense.

I don't think anyone has suggested that.  We just want to know if they know they know what they are doing and if it is actually doing what is intended.  Do they really need to reduce K1 by half (or whatever figure it is) at the top of each season?  Just give us an explanation why things have been changing.