10

Let's say I totally dominate my opponent, but simply run out of time, then I lose the game right? But doesn't it make more sense to be given 51% of the rating adjustment (not 50/50 as for a draw).

The criteria would be simple, that is if a user is in a completely winning position by stockfish evaluation (mate in X) or they are up 5+ material points.

I feel like losing rating when you dominate a GM who is rated 500 points higher than you in a 1+0 bullet game just because they moved very fast (and randomly) at the end is a bit unfair. Is there a historical reasoning for not applying such rules?

  • 16
    I think at least part of it is that "completely winning position" is hard to define. For example, stockfish can be showing forced mate, but that mate is very difficult for a human to find, and otherwise the position is lost. Is this a completely winning position? – Akavall Nov 18 '20 at 4:17
  • 85
    You only got a winning position because you spent too much time. – RemcoGerlich Nov 18 '20 at 7:51
  • 8
    @lbragile Notice how your definition of 'simple determination' has already shifted, from your question stating "+5 material points as defined by Stockfish" down to "in a defined mating net with no other pieces". This should be a sign that "how much is too much" is really hard to define. As a corollary, note that in cases where you have insufficient material to mate with 100% perfect play (K+N, K+N+N, K+B), if you flag your opponent it is typically considered a draw. Even having a single blocked pawn is enough to prevent you from getting that 'mercy draw', however. – Grade 'Eh' Bacon Nov 18 '20 at 14:20
  • 17
    Your friend is being executed at midnight. You storm the castle with your army to save them. Fearing losses, you engage cautiously, taking your time and chipping away at your enemy's forces. By 11:59PM there's little left but a small garrison on the wall and a big gate - your army is overwhelming at this point. Feeling confident, you order them to charge, breaking down the gate with a ram... and you find your friend hanging from the gallows. Did you win the battle? Did your enemy? – J... Nov 18 '20 at 16:33
  • 4
    @lbragile "Very true, but if you are at a certain rating and have a Queen and a Rook vs your opponents "nothing", and are in a mating ladder when you run out of time, that clearly should be a win, no? " You lose on time and claim that you deserve the win. At the next table, your friend in the same position manages to keep playing and not lose on time. But they play so fast to avoid losing on time, that they make a mistake and allow their opponent to get a stalemate. Do you really deserve to win more than your friend who didn't lose on time? – Stef Nov 18 '20 at 23:30
54

I feel like you somewhat misunderstand the concept of a time control. The clock is a part of the game. If you are up a piece in a complicated position with five seconds left on the clock that may well be a losing position for you. Sure, the engine may show you +4, but the clock is an important part, especially in blitz.

Importantly, if you spent all your time getting a good position while the opponent played quickly to conserve their time, then they performed better, because they got to a still tricky position while using less thinking time.

If you don't like this flagging mechanic (I for one do not), increment exists for a reason. There is still some time management there, as playing on just the increment means you don't have time to think deeply. However, it allows you to keep playing instead of losing on time. If you play without increment, it is your decision, but then you have to be up to take the consequences.

  • 1
    I don't think this is quite an answer. I've seen things like lost on time because there wasn't enough time on the clock to play out the 50 moves to draw in an unbeatable fortress. Or lost on time where the opponent couldn't have checkmated against a frozen board despite technically having mating material (a couple of blocked pawns). – Joshua Nov 18 '20 at 20:56
  • 8
    Technically, if there is absolutely no way to mate, it would be a draw by FIDE rules – pulsar512b Nov 18 '20 at 22:23
  • 2
    @pulsar512b But that only applies if there is no way to mate against worst play from your opponent. Is that what Joshua is talking about? – Brian Drake Nov 19 '20 at 15:15
  • I was more looking at the "frozen board" instance- say a blocked pawn ending position. – pulsar512b Nov 20 '20 at 18:37
23

Let's look at the title question from the perspective of engine chess, which is more objective in the sense that you can get two engines to play tens of thousands of games against each other at whatever time control you desire.

If you look at the latest season of the unofficial world computer chess championship, you'll find that the strongest engine is Stockfish, and the weakest is Bagatur/A0lite. A match between the two engines on equal hardware & time control won't be much of a match - in a 100-game match I'd be surprised if Stockfish doesn't score 95% or more. The gap is just that large. Let that sink in for a moment: in an equal match, Stockfish is expected to win with both colors, again and again and again.

However, if you allowed Stockfish only 1 second a move while giving Bagatur/A0lite an hour a move, then it's no longer so certain that Stockfish will beat Bagatur/A0lite. In fact, it might very well lose. See the result of similar testing in the past, where Stockfish had only 0.3% of its opponent's time.

You can see the absurdity of it all. Let's say I sit down to play a blitz match with Bagatur against Stockfish at time controls of 5 minutes per game. Stockfish plays a move. I let Bagatur think for two days, then come back and make a move. Stockfish says I lost on time. I counter with "I flagged, but flagging isn't considered a win". If Stockfish were a human, how do you think it would respond?

tl; dr: the clock is part of the game. If you get a good position, but have to spend too much time to reach that good position, you might not be winning.

18

Because not having it would result in some perverse incentives. In particular, in any game where a draw is an acceptable outcome, the optimal strategy without flag drops being losses would be to never make a move and simply wait for the flag.

With any partial approach like you propose, the optimal strategy would instead be to try to force such a situation (far easier than actually winning - your "+5 on material" suggestion is especially bad for this, because you end up with situations in which the stronger player has a forced mate that involves, say, a rook sacrifice that takes them to -6 on material, at which point the weaker player can simply let that happen, then wait for the flag. Your rule, in particular, would make a fair number of winning positions into losing positions.

It is also worth noting that there already is a partial form of this rule in place: if there is no sequence of legal moves by which the player with time left on their clock can checkmate the player whose flag dropped, then the game is a draw. This avoids the above issue, because in such a situation, a draw is already your worst possible outcome, so the perverse incentive is avoided.

12

Since you specifically asked for historical reasons: Think about your two proposed criteria in the context of the year 1800 or 1900.

  • Using engine evaluation is obviously out of the question. So is "asking a better player", because you may often not have one at hand (physically, due to limited means of telecommunication!).
  • Using material count is at least possible. But keep in mind that back in the day, people generally liked to play chess in a much more "romantic" way than today. In other words, less fear of sacrifices. Determining the outcome of games by mere material count probably had very few fans when the rules started to be formalized (and as any strong player will confirm, the romantics wouldn't be entirely wrong, as material is only one of many factors of the evaluation of a position).
  • 2
    "So is "asking a better player", because you may often not have one at hand (physically, due to limited means of telecommunication!)." Also, the rule should work for all games. What happens if the best player (human or otherwise) in the world is one of the players? – user3482749 Nov 18 '20 at 17:58
  • 4
    This is the tournament for the Title! And this is the final game. The best player in the world is whoever wins this game. Now, if only they agreed... – Ángel Nov 19 '20 at 0:25
9

Except for exceptional circumstances, only three results are considered valid for a chess game (and the corresponding rating change): win, draw or loss.

Your suggestion would come with a huge problem: what exactly counts as flagging? How big should your advantage have been for your opponent's victory to not count as a "real" victory? Who should evaluate that advantage? Would it be fair to take 20 seconds per move on a 3-minute game and still claim you dominated the match?

As other answers have pointed out, the flag is a part of the game. If you don't like losing on time, then you probably shouldn't be playing 1+0

You won't lose much rating for losing to a GM, whether by flagging or not, by the way.

  • 1
    How about a rule which would specify that if a player that runs out of time and doesn't concede, the player with time gets to move pieces for both sides until either a player is checkmated (and thus loses), or other circumstances result in a draw? In 99.9% of situations, good sportsmanship would demand that a player who runs out time immediately concede, but such a rule would avoid any need to have an arbiter try to evaluate whether a position should be considered "dead drawn". A player whose opponent runs out of time but can't manage to secure a checkmate, doesn't deserve a win. – supercat Nov 18 '20 at 23:31
  • 1
    @supercat the player with time left, now moving for both sides, would be encouraged to play bad moves for his opponent, making his side win. – Ángel Nov 19 '20 at 0:22
  • 1
    @Ángel: If the player who ran out of time could have lost by making any sequence of sufficiently bad moves, and the opponent can demonstrate that by demonstrating such a sequence, the opponent should win. The question would be whether the opponent should deserve a win without being able to demonstrate that the person who ran out of time could have made losing moves. – supercat Nov 19 '20 at 5:51
  • 4
    @supercat There is indeed already an equivalent rule. If a player runs out of time but there's no sequence of legal moves that would lead to the opponent winning, the game would be declared a draw – David Nov 19 '20 at 9:15
  • 1
    @David: I don't think there's a procedure given for deciding whether a legal sequence of moves exists that would give checkmate, and I don't think the rules take into account whether any such checkmate could be achieved before a draw based on move count or repetition (e.g. if a player's flag falls at a moment when the player's only legal move would be a third repetition of a position). – supercat Nov 19 '20 at 14:43
7

It's considered a win because that is what it is by definition from the rules of timed games.

When you enter a game you commit to its rules; in this case one of those rules is that the time you have to force mate or other result is constrained by the clock, and you must adapt your strategy to meet that constraint.

In the example you have given, you have not met that constraint and been outplayed by the opponent who has dominated on time.

The simple solution is to adapt your game strategy to the timed format or play another format instead.

3

Why does a football team lose when the clock runs out?

Time is a key factor in playing chess. Else you would be back in the olde days when players could take days or weeks or longer before they moved.

Even postal chess has time limits.

Time limits is fair for both sides. If you played faster thereby not dominating you would not have won. Why should you be special and not have the clock run out on your team just because you did not get to kick that final field goal?

  • A football team only loses if the time ran out and the ball is “dead”. If they are in the process of scoring a touchdown (no matter how unlikely) and time runs out, they still can win if they score the touchdown and it brings them over the score of their opponents. This is kind of why I asked the question as it seems that in chess, depending on the rating, a player should be allowed to checkmate after their time runs out - as a final stance kind of deal. Maybe they can only be given X+1 more moves for a mate in X position? – lbragile Nov 18 '20 at 19:29
  • 2
    Then your adjudication requires that the position be evaluable as mate in N. What if the position is winning, but the forced mate is in 314 plies? Should you get to play another 150+ moves despite running out of time because an engine miraculously found a forced mate? Besides, your football retort is weak- it's like every move is a quarter, more than like every game is. You may be allowed to finish a move if you're already holding the piece ("running a play"), but otherwise you're saying "I should get more moves because I didn't get a fair chance to come back and win!" See? Not how it works. – Lieutenant Zipp Nov 18 '20 at 21:23
  • 2
    @lbragile So, are you saying that you want time increments after your time runs out? Well have I good news for you! Those exist and are common practice :) – Stef Nov 18 '20 at 23:37
  • 1
    @Lieutenant Zipp fair enough, I guess it is simply much more difficult to have a concrete rule for this than I originally thought. – lbragile Nov 19 '20 at 0:13
0

There is simply no other practical "punishment" for overstepping the time limit. You cannot remove a piece or something like that. In fact, the only punishment for other infractions which are not grave enough to warrant declaring the game lost is a time penalty, which clearly doesn't apply here.

It's not that clear-cut though: In order to win, you must have enough material left to mate your opponent. If you are left with the bare King, even when your opponent oversteps the time limit all you get is a draw.

  • In Weiqi(Go) a variation of rule exists, when a player running out of time, their land occupied will be deducted at the end of the game instead of losing immediately. – LIU Qingyuan Nov 20 '20 at 7:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.