I had a question about timeouts and if the system could be made better.
It seems that a player could just let all the games he is trailing in time out, without any effect on his ranking.
Similarly, the players who were leading these games don’t get a win credited to their ranking.
It seems to me a game should be ranked, as long as a certain number of moves (let’s say 10?) have been played.
There is the vacation feature to avoid timeouts, and perhaps there could be a pause button if both players agree and one has run out of vacation time.
It is unclear whether you are referring to timeouts in live games, or serial timeouts in correspondence games.
This is already implemented. (With a limit of 5 btw)
Hi, I meant timeouts in correspondence games. And you are correct, there is a pause button already. I am just wondering what effect a game should have on both players ranks once a timeout occurred. For example, if a weaker player is winning against a player 5 ranks higher, shouldn’t he be rewarded in his rank even if the other guy timed out?
An OGS site rule is that serial timeouts in correspondence games do not count after the first timeout in the series. The original purpose of this rule was to accommodate people who are incapacitated (hospitalized, broken computer, etc.). It has always been controversial, and it is now greatly abused by cheats who deliberately time out their correspondence games. In addition, the rule does not function properly anymore, as completed live games apparently fail to interrupt the series as they once did. See Annulment of Consecutive Timeouts in Correspondence Games
thanks for the explanation. i think the serial timeout rule should only go in effect after 3 weeks or so (all games are ranked for the first 3 weeks after the timeouts begin, the following ones are annulled). i think this would give some protection to incapacitated players, but also discourage cheating.
1 Like
Interesting idea. I don’t remember that being suggested before.
As correspondence games are (at least title tournaments and ladders) usually 7 days max time, the 3 week idea is equivalent to ditching the serial timeout rule. Would only make a difference for really long correspondence like that through the years tournament.
true. i don’t know the programming details, but there could be a player property “active” (yes/no) that is set to “yes” every time any move is made (in live or correspondence games). this would prevent somebody to just make moves in games he is winning.
2 Likes
Just to explain this a little more, the clock would still run down, but if the player has been active in other games, a timeout loss will count against his ranking and give credit to the winner
1 Like
I guess it’s a little more than just making moves in games that you’re winning - because you can’t control when your opponent is going to resign.
So you might stop making moves in losing games, one times out and it counts, then an opponent in a game you’re winning resigns, and then the next timeout counts and so on.
But I do see why it’s a problem.
but I don’t know the stats on how often it happens myself. Is it being abused by a lot of players?
I’m trying to understand why n =2 and not more like 5 maybe?
Is it because it will imply too much delay for the rating system?
I don’t see the issue. As long as you are active in the system, your timeout losses count towards the ranking. The measure for being active is making a move in any game. You could specify that a player is inactive when he doesn’t make any move within a week, and only then the “multiple timeout” rule would go into effect. But as soon as the player makes a move anywhere, he is active again. If he really tries to abuse the system by being inactive, he’d probably lose a large number of games that he is winning, too.