Ranks Need To Be Reset Once A Year

@midnightmusicnetwork said in #10:
> Maybe for some, but it would also increase motivation to play for many others. It would not devalue the rating of someone who is truly playing at their level. It would only devalue the ones who don't deserve it.
These are extremely vague terms you are throwing around. "those who don't deserve it"? "someone who is truly playing at their level"?

@InkyDarkBird said in #11:
> These are extremely vague terms you are throwing around. "those who don't deserve it"? "someone who is truly playing at their level"?

Read first post.

Ever heard of provisional rating?

@InkyDarkBird said in #13:
> Ever heard of provisional rating?

Yeah, but its only for 10 games or so. Need a larger number of games, and the rank reset to clear the rating pool of inactive accounts.

@midnightmusicnetwork the "problem" you are describing is solved by provisional rating, showing everyone that the player has not played enough games to get a solid ranking. If they somehow "get lucky" for 10+ games and stay in the 2000 range, then i guess they got really really lucky and the matchmaking system failed because:

let's say player A is around 1750 blitz and creates a new account. Their first game is vs a 1500, they win, get 500 points. Their next game is vs a 1950 and they get lucky and win, earning about 200 more points. Now they're at 2200, and it's almost physically impossible to get lucky in 8 more games against players 500+ rating higher than you. It's highly, highly improbable, and their rating WILL balance out to or very near to their true rating

Now consider this: wouldn't a rank reset every January 1st be even worse than the current "problem"? Suddenly, if you're a 1500, you have a higher chance of playing someone above your rating than at your rating. 2500 rating players are going to be crushing everybody in their path, giving them easy victories and the <1500s unfair defeats. Wouldn't this just create more of the same current "problem"? Instead of the occasional new player getting lucky, or someone creating a new account, you essentially have THE ENTIRE PLAYERBASE going back to provisional 1500 rating, and so you have more chance of "getting lucky" (e.g. an 1100 beats a 900, but their "rating" is both 1500 since it got reset and suddenly the 1100 is a 2000) and muddying the rating waters much further than the occasional smurf.

i hope this rant/explanation satisfies you and if not i'm more than willing to continue this conversation :) thank you for reading

Inactive accounts don't influence the ratings of others, because they don't play. And they don't influence the rank, because only players active in the last week are relevant for it.

> Yeah, but its only for 10 games or so. Need a larger number of games, and the rank reset to clear the rating pool of inactive accounts.

inactive accounts? accounts that are inactive for a while are already removed from the leaderboards (for example hUdSonZiNho before he got banned) and they're not queueing games very often if they're inactive, so i don't think they would be a problem

@jdwhite42 said in #15:
> @midnightmusicnetwork the "problem" you are describing is solved by provisional rating, showing everyone that the player has not played enough games to get a solid ranking. If they somehow "get lucky" for 10+ games and stay in the 2000 range, then i guess they got really really lucky and the matchmaking system failed because:
>
> let's say player A is around 1750 blitz and creates a new account. Their first game is vs a 1500, they win, get 500 points. Their next game is vs a 1950 and they get lucky and win, earning about 200 more points. Now they're at 2200, and it's almost physically impossible to get lucky in 8 more games against players 500+ rating higher than you. It's highly, highly improbable, and their rating WILL balance out to or very near to their true rating
>
> Now consider this: wouldn't a rank reset every January 1st be even worse than the current "problem"? Suddenly, if you're a 1500, you have a higher chance of playing someone above your rating than at your rating. 2500 rating players are going to be crushing everybody in their path, giving them easy victories and the <1500s unfair defeats. Wouldn't this just create more of the same current "problem"? Instead of the occasional new player getting lucky, or someone creating a new account, you essentially have THE ENTIRE PLAYERBASE going back to provisional 1500 rating, and so you have more chance of "getting lucky" (e.g. an 1100 beats a 900, but their "rating" is both 1500 since it got reset and suddenly the 1100 is a 2000) and muddying the rating waters much further than the occasional smurf.
>
> i hope this rant/explanation satisfies you and if not i'm more than willing to continue this conversation :) thank you for reading

It would iron itself out in a month or sooner. You would not see the 2500s at 1500 for very long.

@jdwhite42 said in #17:
> inactive accounts? accounts that are inactive for a while are already removed from the leaderboards (for example hUdSonZiNho before he got banned) and they're not queueing games very often if they're inactive, so i don't think they would be a problem

Not just the leaderboards, but from the rating pool to ensure the games are fair. Also, for this metric "You are better than "80%" of blitz players"". That's simply not true, because many of them are inactive. Remove the inactive players from the metric by a rank reset. Plus, the added bonus of huge influx of people playing to grind to their rank again. There would be a large pool of people to play, and chess would become very much alive again.

@midnightmusicnetwork said in #18:
> It would iron itself out in a month or sooner. You would not see the 2500s at 1500 for very long.

This is exactly what I'm saying!!! there's no need to implement a new solution when the current "problem" irons itself out after 10-20 games. Having a rank reset would make more people in the wrong ranks. Having a hard rank reset is not a good solution, and it makes it less fun for the lower rated players who are going to be playing inappropriately ranked opponents for "a month or sooner"

>Not just the leaderboards, but from the rating pool to ensure the games are fair. Also, for this metric "You are better than >"80%" of blitz players"". That's simply not true, because many of them are inactive. Remove the inactive players from the metric >by a rank reset. Plus, the added bonus of huge influx of people playing to grind to their rank again. There would be a large pool >of people to play, and chess would become very much alive again.

the games aren't determined by who is in the rating pool overall, they're determined by what rank you are and what rank others are that are queueing the same time control you are. (or who picks up your challenge in the lobby) inactive players aren't going to be queueing.

i don't know if it counts inactive players in the "80% of blitz players" but i know it doesn't count inactive players (of that time control/variant) on the leaderboards, so it seems more logical that they're removed from the ENTIRE rating list until they play more games. They lose their trophy as well if they're in the top 100 but are inactive.

"The added bonus of... people playing to grind to their rank again." Sure, that could be fun for those willing to invest time into getting back to 2100 or whatever. But what about 1100s who are playing 1500s and 1600s instead of the 1100s they should be playing? They can't "grind" to anything because they're starting at 1500. Unless we're starting EVERYBODY back at 600 (the minimum) then not everyone will be able to "grind back to their rank." and even if you start everyone back at 600, anyone who loses will be stuck at 600 and playing people not of their rank.
If all MY hard work to get to 2000 atomic was erased at the start of every year, in all honesty i might consider switching to chess.com.

"a large pool of people to play." there already is a large pool of people to play. If we had a hard rank reset, sure there would be more people to play but a good portion of them (probably 80%+) would be higher or lower than your rank, giving you a game that's not up to standard.

besides, it would also cause a lot of confusion, and there's no real need to implement a solution like this.