First of all, i didn't ask anyone for permission to post this thread. Do not redirect me to the simple questions thread, i know that thread and this topic is not a simple question and i do not think that there will be any simple answers.
I think most of you have already seen the ladder and the incredible scores some people have there. I understand the concept of the so called "ladder inflation" (the better the ladder gets the higher scores can be obtained) but those two users on the screenshot just have scores that i can't understand.
Let's start with the definitely easier case - hogapen. As far as i understand, GXE is a percentage determined by both your rating and your win-loss ratio. The GXE rating shows a percentage chance of winning against the average player. User Hogapen has a negative win-loss ratio. If the score was only determined by rating, i wouldn't doubt the 100% chance, but comparing it to the next best user on the ladder - americann beauty, who has an Acre of 3258 and a win-loss ratio of 53-6, and his GXE is "only" 99.9. Can someone explain me how this is happening? The number one is obviously not automatically set to an 100 GXE rating, because that's not the way it is on other ladders. I just can't see someone who has lost to more people than he won to have a 100% chance of beating the "average" player.
The second user is alot harder, even though the error is much more obvious, simply because 5277 - 188 does not equal 11807. Even if it would just be a cut off number and some digits would be missing, you do not find any number where you add one digit either in the beginning or the end of 5277, subtract it with 188 and it equals 11807. It's just not happening. So here the system has obviously been glitched. I do not know if that is an error in the PS! mechanics or in the actual Glicko-2 system, but something here is fishy. I downloaded the Glicko-2 excel calculator and just inserted someone with a starting rating of 1007 and a deviation of 350 (this is what i got from wikipedia, i just assumed it is the starting deviation of PS too) and let him battle the current number one with a rating of 3705, since i assumed this would be the maximum possible gain of rating. Obviously this is something that does not really happen on PS, or at least it shouldn't.
It resulted in an outcome of 1721 ACRE for the previously 1007 rated person. I repeated the process against the exact same 3705 person and it resulted in 2437.
I repeated the whole process again, until i got 11 wins against the number 1 of the ladder. It resulted in a factual rating of 8768. Eleven wins with what i consider the maximum possible outcome of rating and it doesn't even come close to a rating of 11807. Here is a screenshot of the outcome, it is possible that i made mistakes right here, so i advise you to download the calculator yourself and check my results, this is the download link, i got it from the glicko.net website, the actual website of glicko's creator: http://www.bjcox.com/?page_id=20&did=11
I am not an expert in mathematics, nor do i even fully understand the Glicko-2 rating system, or even find a loophole or anything in either the PS mechanics or the actual Glicko-2 system, but it's not hard to see that something is wrong with at least those 2 accounts. If anyone could check what is happening there it would help me and many others understand the situation. This is not an attempt on invalidating anyone's rank, i just do not understand how numbers like that are possible in this system and think that they happened through a glitch or something similar. I also do not want to offend anyone and anyone's rating by calling it invalid.
Edit: the user alt 587300 played some more games and is currently standing on a rating of 18773, tendency rising
I think most of you have already seen the ladder and the incredible scores some people have there. I understand the concept of the so called "ladder inflation" (the better the ladder gets the higher scores can be obtained) but those two users on the screenshot just have scores that i can't understand.
Let's start with the definitely easier case - hogapen. As far as i understand, GXE is a percentage determined by both your rating and your win-loss ratio. The GXE rating shows a percentage chance of winning against the average player. User Hogapen has a negative win-loss ratio. If the score was only determined by rating, i wouldn't doubt the 100% chance, but comparing it to the next best user on the ladder - americann beauty, who has an Acre of 3258 and a win-loss ratio of 53-6, and his GXE is "only" 99.9. Can someone explain me how this is happening? The number one is obviously not automatically set to an 100 GXE rating, because that's not the way it is on other ladders. I just can't see someone who has lost to more people than he won to have a 100% chance of beating the "average" player.
The second user is alot harder, even though the error is much more obvious, simply because 5277 - 188 does not equal 11807. Even if it would just be a cut off number and some digits would be missing, you do not find any number where you add one digit either in the beginning or the end of 5277, subtract it with 188 and it equals 11807. It's just not happening. So here the system has obviously been glitched. I do not know if that is an error in the PS! mechanics or in the actual Glicko-2 system, but something here is fishy. I downloaded the Glicko-2 excel calculator and just inserted someone with a starting rating of 1007 and a deviation of 350 (this is what i got from wikipedia, i just assumed it is the starting deviation of PS too) and let him battle the current number one with a rating of 3705, since i assumed this would be the maximum possible gain of rating. Obviously this is something that does not really happen on PS, or at least it shouldn't.
It resulted in an outcome of 1721 ACRE for the previously 1007 rated person. I repeated the process against the exact same 3705 person and it resulted in 2437.
I repeated the whole process again, until i got 11 wins against the number 1 of the ladder. It resulted in a factual rating of 8768. Eleven wins with what i consider the maximum possible outcome of rating and it doesn't even come close to a rating of 11807. Here is a screenshot of the outcome, it is possible that i made mistakes right here, so i advise you to download the calculator yourself and check my results, this is the download link, i got it from the glicko.net website, the actual website of glicko's creator: http://www.bjcox.com/?page_id=20&did=11
I am not an expert in mathematics, nor do i even fully understand the Glicko-2 rating system, or even find a loophole or anything in either the PS mechanics or the actual Glicko-2 system, but it's not hard to see that something is wrong with at least those 2 accounts. If anyone could check what is happening there it would help me and many others understand the situation. This is not an attempt on invalidating anyone's rank, i just do not understand how numbers like that are possible in this system and think that they happened through a glitch or something similar. I also do not want to offend anyone and anyone's rating by calling it invalid.
Edit: the user alt 587300 played some more games and is currently standing on a rating of 18773, tendency rising
Last edited:
