ELO ratings should decrease over time if you are inactive

If I recall, it was in the mid 1800's when the Pawn was changed to codify some of the rules about how promotion worked.

 

And it was the late 2000s when Pawns became stars. I hear it happened in the Las Vegas area.

 

My idea to solve it would be week/month rewards.
For example:

• if u get 40 wins during 1 week you get a golden chest
• if 60- 2 goldens
• 80 - 1x purple
• 100 - porsche
• etc etc

Of course these rewards would have another timer to reset and would be nice addition to everyday rewards.

 

Did some more thinking on this and came up with this scenario:

a) So you take the median amount of games played by the average player per day lets say that number is 10.

b) Each day you must play the median amount of games (this number can be hidden it doesn't matter) or suffer ELO decay.

c) You work out the average amount of ELO that is gained/lost from a single match and that is deducted from your score times the amount of games you are short of playing from the median amount.

Example. (using the above hypothetical)

Johnny has an ELO of 1000, He has played only 5 games today so he is short 5 games, The average ELO gain/lost that day was 20, Johnny's score is adjusted down by 100 points to a score of 900.

 

If at least half the accounts are averaging less than 1 game per day (which is likely), then the median will be 0.

Oh wait, that's the median of the average number of games per day for each player. I don't know how to define "average player".

 

Obvious sarcasm aside, it would have to be calculated from active players OBVIOUSLY.

 

The deal-breaker here is probably the fact that a person has to play every single day or suffer ELO decay. If people go nuts after I suggested a small ELO decay after one week of inactivity, their head would probably explode if what you suggest was made a reality. Even if it made sense both logically and mathematically, the simple fact that you are "forced" to play every day would be seen as a bad thing.

My first reaction to your idea is that there is a lot of decay. I mean losing 10% of your rating in one day, even though you played during that day? Way too much for my liking. I don't think it should ever drop down more than 25-50 per week. And those numbers should require a lot of inactivity.

 

although i dont think the numbers would be that high, i guess you could use the above method on a per week basis or even have an adjusting multiplier or divider at any given tier.

Say your in the top tier you could have a divider of say 5 and your elo decay is divided by that number.