If "leaving a game without resigning can result in a temporary ban"...

@IbnHussein said in #20:

My point is that you don't need to, you just need to measure something highly correlated with winning/losing. Like StockFish's centipawn advantage.
But I had times I thought I am winning and stockfish laughed of my thought. Or that I was losing.
Just saying, other than power cuts I don't disconnect.

@EmY48 said in #21:

But I had times I thought I am winning and stockfish laughed of my thought. Or that I was losing.
Just saying, other than power cuts I don't disconnect.

Yes, but the point is just correlation. There are also times when you thought you were losing but stockfish knew the right moves. Statistically it all balances it out for the blameless disconnector. For the rage quitter, the pattern will be clear.

@IbnHussein

It's an interesting idea. But running stockfish every time someone doesn't move in time? Best break out your patron hat!

@IbnHussein said in #20:

My point is that you don't need to, you just need to measure something highly correlated with winning/losing. Like StockFish's centipawn advantage.
Which doesnt work in some cases

@IbnHussein said in #20:

My point is that you don't need to, you just need to measure something highly correlated with winning/losing. Like StockFish's centipawn advantage.

what do you mean?

Maybe it doesn't matter whether someone was winning or losing? Maybe what matters is just whether someone has a high disconnection rate or high number of disconnected games? Whether it occurs due to their internet being spotty or due to their actions when losing, it has the same effect on the opponent.

So perhaps members who exceed some threshold for disconnected games get a temporary ban, or maybe the interface can allow you to match against users based on disconnect rate (or completion rate) as well as rating range? Or maybe the interface can warn you when it matches you with someone with a high disconnect rate, so you can choose not to play that opponent?

I haven't played many games here yet but did have a game when I wasn't logged in (anonymous) where I got a winning position and the opponent simply stopped playing without resigning. It was annoying and since I like playing longer time controls (15 min+) it could lead to a lot of wasted time. Would be nice to be able to avoid repeat offenders.

@IbnHussein said in #15:

Uh, StockFish? This is just a heuristic anyway, doesn't need to be exact.
Stockfish analysis' are done by volunteers on their private computers, and the number of games you can require an analysis for is limited, because apparently there are not enough of these volunteers.
Your suggestion would require many more stockfish runs for a small improvement against game leavers.