Kicking people out who are over time

[Jest Phulin]

Any decision about extending the time should be a conscious decision by all players, not an automatic adjustment.

Adjusting based on the average time remaining would actually discourage fast play, as players with a large amount of time remaining would have their clock reduced.

[Caffe Latte]

Jest
Not expelling the timeouted Player - making move instead - might count as a "decision" not to expel.

[eoc]

I can only suppose it is because you get a bigger ELO boost for "winning" a game at 99% progression than at 15%. But from my perspective, it would have been much better if he had kicked me from the game when I *first* ran out of time.

To discourage that kind of behavior, and assuming that most similar issues will arise in 2-player scenarios: how could this particular one be addressed?

Remembering when an opponent first ran out of time, and only counting that progression percentage for partial ELO (15% in your example) no matter at which point you decide to expel your opponent?
Introducing karma penalties if you expel your opponent after the game progression changed by a certain amount since they ran out of time?
Removing any personal ELO gain from games where you expelled your single opponent?
Ignoring / reducing the loss of ELO to an opponent who exceeded their time?
Simply reminding players they could also auto-abandon the game, losing no ELO that way? (Not everyone acts maliciously like you're suspecting for your particular case, some just don't know better)

Note that anything in this direction may encourage gaming the system just as it does now, but in different ways from now, so BGA will have to be careful with their solution(s). I would however oppose the suggestions of manipulating general playing clock behavior to solve such a specific problem. Doing so will cause many more issues down the line than it could possibly solve and most likely be perceived as rewarding "unfair" behavior. Which is not an improvement to the status quo

[Romain672]

Just a little idea : make an elo penalty (like 30%) if a winner goes out of time at one time on the game.

So if he won, he gain less elo (and you loose less), and if I won, he loose normal elo (and you gain normal elo).

Some problems with that idea :
- 3+ players game
- 1 second out of time is the same than 1 minute or 10 days
( - special elo game (hanabi, etc))

[Jest Phulin]

Jest
Not expelling the timeouted Player - making move instead - might count as a "decision" not to expel.

You are correct. I guess currently the default behavior (of not expelling a player) is a tacit agreement to let the person play.

[Caffe Latte]

Jest, I guess you read a story above, one from many examples used timeouting as a winning blocker. Not expel triggering timer reset would be much more fair. Either you use your absolute right to expel or lose this superpower till next timeout if any.

[Jest Phulin]

I can agree to that.

The situation seems to be that when a person is in the negative, they are in constant fear that someone will enforce the clock rule even if they didn't do it immediately. (I still say that with fast play they can get themselves out of this situation, but that's another discussion....)

So, a proposed change: When someone goes negative, the system automatically proposes a vote (with an obvious message that a vote is taking place) to extend the time of that player. If any person explicitly votes no, the out of time player will be kicked, just as if someone explicitly kicked them. If a player takes 2 turns, the system will treat it as a "yes" vote. (2 turns, because a person may want to make their move first while their clock is running, and think about the options.) If all players agree to let the time be extended, the clock will be set to the amount of time that is added per move. IE, if playing a 1 move per day game, their clock will be set to 24:00:00. If playing a game of 00:05:00 with 30 seconds added per turn, their clock would be set to 0:30.
If, during the time that the vote is taking place, the person is able to get back to a positive clock (through normal fast play), the vote will be discarded.

This will remove the fear, and allow a player who had an extended period of not making a move be able to maintain a positive clock if they play at the speed they agreed to. If they continue to play slower, they will again go into the negative and be subject to being kicked.

I don't think this will affect tournament play, as those clocks are a set time with 0:00 added each turn, so 1 second after the reset they'd be out of time again, and in the case of a game ending on total time the person who has the most time remaining is declared the winner. I could be wrong on this, though.

Issues that I see that will need to be addressed:
If a person constantly times out, but people want to allow them to play, will the obvious vote message become annoying?
How should Karma be addressed? If a player times out, gets the vote to be reset, and times out again should this be 1 out of time penalty or 2?
If more than one player goes negative while the vote is being taken, does a separate vote need to be taken?

Also, please never forget that the clock rule is a well-documented rule that each player is reminded of as they join every table. (Gee, your honor, I normally take highways when I travel, I took a turn onto a road with a clearly posted lower speed limit. The police officer following me didn't immediately pull me over for speeding, so I shouldn't have to pay this speeding ticket I got 10 miles later, right?)

-----------------
Side notes: The excuses I see of "I didn't realize the time limit for this game" and "I couldn't ask because of a language barrier" are flimsy at best. The time is easily stated, and Google Translate is a thing.

[eoc]

So, a proposed change: When someone goes negative, the system automatically proposes a vote (with an obvious message that a vote is taking place) to extend the time of that player.

Could extending the time of every player by that same amount be helpful? Both to encourage people to vote in favor (since they'll get something in return) and to combat perceived unfair systematic bias, I hope.

[Jest Phulin]

So, a proposed change: When someone goes negative, the system automatically proposes a vote (with an obvious message that a vote is taking place) to extend the time of that player.

Could extending the time of every player by that same amount be helpful? Both to encourage people to vote in favor (since they'll get something in return) and to combat perceived unfair systematic bias, I hope.

Not a bad idea, but it has its own problems. My idea was to raise a person's time to the standard thinking time, no matter how negative they were. If everyone's time is raised, would it be simply adding a single thinking time to everyone, or add the amount that was added to the negative player? (So, if they were in a 1 move per day game and were -36 hours, would people get 24 or 60 hours added to their time?)
What would happen if adding this time would take someone over the total thinking time?

Again, I'm just trying to see potential problems. More discussion and more questioning is helpful. Much better to go to the powers that be with "this specific implementation will fix issue A (and possibly create issues B and C, which have less impact)" rather than "y'know, something ought to be done about issue A."

[eoc]

If everyone's time is raised, would it be simply adding a single thinking time to everyone, or add the amount that was added to the negative player?

The latter, I was suggesting. Capping at the max. possible positive thinking time should not be as much of a problem I hope, since that by definition is the most time anyone on the table can get at once, and these players will still have more time than the others. I just think it's very important to maintain a sense of relative advantage: if you "force" players to change their plans because the table will now take longer than expected initially, give them a bit of extra time to accommodate.
That way I hope they're less incline to vote Expel when it comes to kicking out players who are slightly overtime. It's important that there isn't just one player benefitting a lot while the others don't receive anything for their decision in support of keeping the game alive (and may well time out themselves at a later point without the assurance of reciprocity).