Hence serious tournaments do not use absolute time, both in go and chess.
In my opinion just don’t play absolute time, it tends to ruin games unfortunately.
7 Likes
I respect flagging as a legitimate part of the absolute time meta.
I disagree with Martin that it ruins the game – having to perform checkmate under time pressure is tactically demanding and, in my opinion, a significant part of the fun.
I do agree with Martin, though, that to avoid flagging one should simply not play absolute time games.
As an addendum, I’m a weak chess player of Lichess 1600 or so, which I’d say is something like 9k. For me, to even execute something like a KR vs K mate with 30s on the clock is an enjoyable challenge.
But yeah, I see accepting an short absolute time game of, say 5+0 and less to be a tacit admission that some level of flagging or, at least, attempt at stalemate scams by the losing player is OK.
1 Like
What is your opinion on theoretically drawn endgames like King + Rook vs. King + Knight? Where the player with more time can play on forever, trying to flag opponent or induce a mistake. Would you enjoy playing / watching such an endgame for 50 moves? In real life you’d also need to keep track of how many moves were played so you can claim draw by the 50 moves rule … ugh
In my opinion its even worse in go where players can keep playing useless moves in already claimed areas to drag out the game.
I feel the validity of flagging in chess depends on
- the level of both players
- the exact endgame
- the amount of time that is left on the clock
It’s not a simple question.
Also, I’m talking specifically about chess, which has a very different endgame and also the instawin condition of checkmate.
I think, as well, that it’s the right of the player with advantage to try to break through the drawn endgame. Again, any specific example depends heavily on player level and time, but there are many instances of “trying to break through” that are unrelated to time control.
In chess, we also have to distinguish theoretically drawn positions from genuinely drawn ones. It’s impossible to mate with KN vs K, but there exists a chance of mating with your example of KR vs KN, and the realism of that chance depends, as said, on player level and time availability.
A year or two, if memory serves me right, I saw Timur Gareyev (a very accomplished blindfold player) survive into a supposedly textbook “drawn endgame” against Fabiano Caruana; as it happened, Timur made a mistake under time pressure and his rook defence structure collapsed, and Caruana won. Was Caruana unsportsmanlike to press that endgame? I don’t think so.
2 Likes
A few times this year, the idea of the “backgammon rule” has been raised.
The “backgammon rule” is the option for a player to ask the opponent to resign; if the opponent refuses then the game will be played for double the rating points.
This would be a way to discourage long and useless endgames.
4 Likes
Good discussion.
For the record, I post this question mainly because I was bored on a road trip. 
2 Likes
It was a good question.
3 Likes
Discovering the chess world as a go player a couple of years ago, it was surprising to me how common absolute time control (and as a consequence, flagging) is in online chess. Having watched lots of chess content now, I understand it better. I’m still generally against absolute time controls, even in chess, but I also appreciate some of the skills involved in the resulting time scrambles.
But as others have noted, chess and go are very different in this respect, mostly due to the “instawin condition” as bugcat calls it. Since there is no such thing in go, absolute time controls don’t really work at all (unless you have some sort of “sportsmanship” agreement like gennan mentioned).
(or if we manage to introduce an instawin condition, like the GoQuest technical knockout I mentioned above, but I think that is a lot less useful on bigger boards)
I think playing with something like a 1-second increment gives almost the same experience as absolute time controls, without the drawbacks. It’s still not a good idea to get low on the clock during the game (because you probably don’t play very well with 1 second per move), but if you use up all your time to get a completely winning position, you will be able to convert it as long as you can play safe moves (respond to ataries, capture dead stones, remove ko threats) in less than 1 second. It takes some practice, but after having done this quite a few times on GoQuest I don’t mind it as part of the game. I’m not annoyed at my opponents for trying to flag me, but rather feel “oh cool, I get to practice/prove my skill in securing this win in the most fast, safe and efficient way”.
So in my opinion the best way to play absolute time controls is with a 1 second increment 
Of course OTB it takes longer to place the stone and press the clock, so something like a 2-second increment or a perhaps 3-second delay would be needed to get the same experience.
8 Likes
It is definitely allowed, though there are fewer “obviously drawn endgames” than one might think.
A game may be theoretically drawn, or even won, but that theory is usually dependant on perfect play. Playing it out to the last move gives the advantage player more chances to blunder.
I couldn’t tell you how many games I’ve won from a drawn position, or drawn from a lost position, or won on time from a losing position, just because I keep playing to the last possible move.
2 Likes
And then there is King + Bishop vs. King + Knight, which is not only theoretically drawn but also in a practical sense. The only way to get checkmated is for the player with the knight to position both their pieces in the corner of the same colour as the bishop, which never happens. But it is theoretically possible, and hence lichess does not end the game. Funny thing is, should the knight ever get captured, then lichess concludes the game in a draw due to insufficient material for a mate. Hence as long as the knight is on board, the side with the bishop can try to flag the opponent but has to avoid capturing the knight. I have absolutely no interest in such a game conclusion. 
5 Likes
Hilarious
1 Like
I honestly have very little new discussion to add, as others have already expressed my views.
For some context, I started playing internet chess long before I started playing Go. Initially, I definitely played more games without increment (aka “absolute time control”), and I just took that to be the norm. However, as time has gone on, I started favouring increment (aka Fischer time control) more and more, until now the vast majority of my internet chess games are played with increment (3 minutes + 2 seconds per move is my main time control now).
I simply don’t personally enjoy some of the time-strategies that arise from the use of absolute time controls in chess. But that’s just my preference, and I have nothing against the many players (including many strong grandmasters) who enjoy playing this way online. I just don’t feel good about personally playing moves such as:
- Deliberately throwing in useless checks and piece sacrifices to force the opponent to spend time responding
- Trying to “guess” the opponent’s premove and playing an objectively bad move that only works because the opponent has already committed to their move before I make mine
- “Miai time draining” - when a position is easily drawn, but the defender’s response is dependent on seeing what the attacker will do (the defender can’t autopilot all of their moves)
Serious chess tournaments (such as national championships etc.) should always have increment though. The following video (for the title of USA women’s chess champion in 2008) is an example of something I do not wish to see in a serious tournament (be it Go or chess): GM Irina Krush throws king when getting flagged - YouTube
I also agree that small increments are not a substitute for good time management during a game. If I’m ‘living off the increment’ when the game is still complicated, there is no way I can avoid blundering. Similarly, there is no way I can evaluate a multi-group fight with only one 30s byoyomi period. I’ve lost count of the number of games I have lost where I simply played worse because I had used up too much time, and in each case, it is entirely my own fault.
Finally, I see no issue in “testing” the opponent’s skills, even though you know you are lost with best play. In fact, I would go as far as to say it’s OK to test the opponent’s skills, even if you know how to win if the players switched colours at that exact moment - though I sense this might be a bit more controversial.
My opinion is that it’s all relative - what would be obviously disrespectful in a game between two dans, might not be between two DDKs. But for example, I think a lot of players around my rank (myself included) have poor reading skills. So I have absolutely no issue with trying a speculative cut or dodgy-looking invasion, or starting a capturing race from a disadvantageous position - because I know that if such a thing were tried against me, I would occasionally fumble too - which is why I’m not a higher rank than what I am!
5 Likes
I would say that is actually the BEST time for testing. Western players seem to eschew endgame training in favor of opening memorization and middlegame tactics. So just cause someone can gain an advantage doesn’t mean they really know what to do with it.
1 Like
If you have a chess opponent of the keep playing to the last possible move school, it’s fun to see for how long you can make them keep playing in incremented time control.
Consider this position, KR vs R. White declines to implement a rook box mate.
-
Ka8–>b8–>b1–>c1–>c8–>d8–>d1–>e1–>e8–>f8–>f1
Toggle rook G1<–>G8 when attacked. -
Kf8–>e8–>e1–>d1–>d8–>c8–>c1–>b1–>b8–>a8–>a1
Toggle rook G2<–>G7 when attacked. -
Ka8–>b8–>b1–>c1–>c8–>d8–>d1–>e1–>e8
Toggle rook F1<–>F8 when attacked. -
Ke1–>d1–>d8–>c8–>c1–>b1–>b8–>a8–>a1
Toggle rook F2<–>F7 when attacked.
etc. etc. etc.
It was with setups like this that I discovered that Lichess calls an automatic draw at 300 moves.
There is the fifty-move rule, though, so in practice you want to have a store of other pieces that you can force-feed Black every forty or so moves.
eg. it’s quite easy to promote spare pawns to rooks and then play them into the H file with check, such that Black is obliged to capture and reset the count.
And don’t forget that pushing pawns resets the count as well! Do a forty-move king lap, push the pawn, another lap, another push…
1 Like
Now that’s just evil lol
2 Likes
Not that this topic really needs revival but I did play quite a silly 9x9 game on BadukPop.
As an example of why certain time settings are a hassle, depending on the server of course.
The standard settings on BadukPop’s automatch are 3m+1x30s.
While you can’t see how long moves take, what happened is that the opponent played on far beyond what was reasonable with pointless moves. Now They had a higher rating than me but a lower rank (because badukpop’s rank is linked to tsumego and beating the ai among other things, although one can skip to 1d I think if one gets a high enough rating I believe.), so for a while I wasn’t sure if maybe they were still a bit new, so playing on is fine. Then after a while they started using up their entire 30s just to play another pointless move, so maybe they just wanted to time me out since I was also in byo-yomi.
BadukPop has no call moderator button during the game, so one just has to wait it out, like I did where they timed out, or resign. They do however have a report button after the game and an option for this scenario, and also a button to say something like “don’t match me with this player again”, which I wouldn’t normally think of doing, but in this case I don’t want to play a player like that again.
Anyway while it’s not an example of absolute time, it is an example of the nonsense of trying to win on that, that OGS and I presume BadukPop have rules and moderators dealing with.
7 Likes
the only solution is to play live games, then you can just suck it up and let your opponent run out of moves. 
1 Like
It was a live game 
2 Likes