Idea to mitigate cheaters and increase player satisfaction

Cheating is fun at first. But it's easy and gets boring very quickly I imagine. Because of this I imagine a general trend where most cheaters probably haven't got a lot of hours played.

I propose a change to the matchmaking system, whereby people are not only matched by elo but also by roughly the hours played on their account. E.g 1200elo[100 hours on record] vs 1210 elo [90 hours on record]. This would mean that the players who have invested the most time into lichess will have a lower probability of encountering cheaters, thus increasing player satisfaction which means more hours played overall.

This change would also increase revenue for lichess in the long term, as I imagine player's with more hours are more likely to donate.

Thoughts?

Thumbs down but no discussion. Why would it not work?

Assuming your guess about hours played for cheaters is true - how would this change have affected your experience when you were new to lichess?

@corvusmellori said in #3:

If new to lichess then you would play people just like you - people with no hours.

I see your point though, if my suggested change were implemented then newer players would meet more cheaters at the beginning.

If you imagine farther ahead though, new players would be pushed down into lower elo's where there would be less cheaters (cheaters presumbly have mid-higher elos but low hours). When these newer players grind to higher elos they would have more hours and so wouldnt be matched with new players. I don't think your argument devalues my suggestion, thanks for replying.

@Proficci said in #2:

Thumbs down but no discussion. Why would it not work?
It has been proposed multiple times before, perhaps people are tired of repeating the same arguments again and again.

@mkubecek said in #5:

It has been proposed multiple times before, perhaps people are tired of repeating the same arguments again and again.

If I was aware, I would not have asked - apologies. What was the main reason it wasn't accepted in the past then?

EDIT: Accidentally sent this on my alt (my blindfold account)

For me, the biggest problem is that if you let new users (with unreliable rating) play only against other new users (with unreliable rating), how does their rating become reliable and starts reflecting their real strength (relative to the whole pool of players)? You need results against players with relevant rating to be able to estimate someone's strength reliably.

Also, this assumption

Because of this I imagine a general trend where most cheaters probably haven't got a lot of hours played.
is rather questionable, IMHO. First, assuming that cheaters get bored soon is a bit too optimistic, I'm afraid. Second, you cannot assume that every cheater (or their majority) starts cheating right away. Many start later, sometimes even much later.

@mkubecek said in #7:

For me, the biggest problem is that if you let new users (with unreliable rating) play only against other new users (with unreliable rating), how does their rating become reliable and starts reflecting their real strength (relative to the whole pool of players)? You need results against players with relevant rating to be able to estimate someone's strength reliably.

Also, this assumption

is rather questionable, IMHO. First, assuming that cheaters get bored soon is a bit too optimistic, I'm afraid. Second, you cannot assume that every cheater (or their majority) starts cheating right away. Many start later, sometimes even much later.

That's a good point. By weighting matchmaking by hours played you would segment the elo pool. "groups" of elos would occur, based on hours played. Because people could only play against others of the same hours, someone could easily get to 2000 for example, because they're only playing new people. A 10 hour 2000 vs a 5000 hour 2000 would probably lose so what does the 2000 even mean. By weighting elo with anything you discredit it.

I didn't think my suggestion all the way through, thanks for helping me understand.