In a double time scramble in the above position, white plays 1.Rxe5 Rxe5 2.Nxe5.
Black quickly asks the arbiter to declare a draw. The arbiter replies, “No, at least not yet. Play a few more moves.” Black panics, with just a second or two remaining on his clock, and notices that he has only one move that does not lose the knight. He quickly plays that move, 2…Ng8. Of course white responds 3.Nf7 mate.
The game is played under FIDE rules, where K+N vs K+N is not an automatic draw. There is a FIDE rule that says an arbiter can declare a draw if neither player is trying to win “by normal means”.
Should the arbiter have declared a draw after 2.Nxe5, as black requested?
This is not a quiz, it is a discussion question. Opinions may legitimately vary here. Have at it!
Since K+N vs K+N is not automatically a draw, I don’t see how an arbiter can reasonably rule that a player isn’t trying to win by normal means immediately upon that material situation occurring on the board. Don’t they have to play a few moves before you can ascertain what they are trying to do?
Black should have just given away the Knight. King & Knight versus King is a dead position [i.e. no possible mate], whereas King & Knight versus King & Knight can lead to a mate [i.e. as in the game].
To be honest, my first impulse would have been to call it a draw, but given the above explanations, I think the arbiter acted correctly. I guess that’s one reason why I’m not an arbiter.
The c7 pawn could eventually underpromote to N/R/B as part of a helpmate. Would you then rule differently?
Observations:
When combined with zero-increment time controls, existing FIDE rules violate the spirit of the game. But even a two-second increment (or delay!) should solve the problem.
I generally prefer FIDE rules to US Chess rules. But Bill Smythe is pointing out an area in which FIDE rules (ahem) suck. It’s not fair to saddle the arbiter with determining what “normal means” are. E.g., KR vs KB (bishop not in bad corner) is an edge case…didn’t Kramnik flag someone in an exhibition match circa 2000?
Here’s an example from a recent game between two patzers.
Blitz game (let’s say three minutes with two seconds increment)
White flags after having played Kh5-g4. What is your ruling?
-under FIDE rules?
-under US Chess rules?
(If you happen to know the case in question, I agree with the ruling, although I hate it. Edit: Firouzja is appealing the ruling, and the penultimate round is being held up.)
Tweaking the position a bit. (This tweak was suggested by Peter Leko and is itself the subject of debate by the commentators.) Again, White to play flags. What is your ruling?
-under FIDE rules?
-under US Chess rules?
Proposal: mates constructed by hypothetical underpromotion should not be taken into consideration.
Both are losses for White under FIDE rules since White can underpromote in the second case and block himself into a dark-squared corner. I would agree that a helpmate that requires underpromotion (aside from being much harder to verify for a TD) means that a player is harmed by having extra pawns, though it would be somewhat unfortunate if #1 would be a loss for White (i.e. having an extra Bishop is also bad) while #2 wouldn’t.
Unfortunately, the US Chess Rule blitz rules uses the phrase “insufficient mating material” which no longer is the term of art in the main rules. I assume the intention is “Insufficient Material to Win on Time”, which would mean that both are draws under US Chess rules.
In position #1, at least one can imagine White putting bishop on h7 & king on h8 in a vain attempt to queen the g-pawn. Black puts king on f8 and plays …Bc3 #.
In position #2, White’s selfmates require underpromotion. I think it would be a very reasonable tweak to FIDE rules to eliminate this type of suimate, which is a bit more basic than “don’t put your king into a corner.”
The best advice, of course, is “don’t flag.” Poor Peter Leko was driven to distraction by Firouzja’s clock handling…There is really no reason to flag in this position with a two-second increment. Ba2-b3-c4-d5 could have been Firouzja’s next four moves: four “clockside” moves would net at least two seconds, if not more. As a wise person observed upthread , the increment really should eliminate the danger of flagging.
Video here (the incident is in the first two minutes). Instead of playing Kh5-g4, White could have forced resignation in a couple moves with the cute overload g5-g6.
Great results for Carlsen, Nakamura, and Kramnik (!). Those of you following Twitter or Reddit may already know that Firouzja is no longer playing under the Iranian flag, and is contemplating a switch to France or the USA. He still has fans in Iran (IRNA reports that Magnus Carlsen is Belgian).
One of the unfortunate things about the way the FIDE rule works is that it can be really hard in time pressure to properly determine whether you are actually in danger of loss. For instance, holding R v N (no pawns), there are a fair number of positions where the position is theoretically won or at least requires accurate defense, but R v N is subject to helpmate so you would have to be very careful with your time. OTOH, R v B isn’t subject to helpmate, so you can play up to flag in a position that is almost certainly brainlessly drawn.
And …Nf7 forces a knight off. Either white plays NxN+ with an automatic K+NvK draw or white moves the knight and black plays Ne5+ where if white doesn’t capture the black knight for a draw then black will capture the white knight for a draw.
It is always easy to see what a blitz player should have played when you are not the one in time pressure.
White should refuse to take the knight and the game goes on forever as both sides have mating material (K and N)! Somewhere around move 500 repetition of position becomes increasingly likely, but the observant player will move his king to a new square about every 30th move while the knight goes around in circles. Increment allows players to play the game out for days and weeks.
The arbiters can call it after 75 moves.
Continuing would require either both players wanting to continue or black failing to see that a knight can be forced off by Black’s third move.
Even if both sides claim they are playing for a win, the arbiter can still call it a draw after 75 moves have been made by each player without a capture or a pawn move.
No argument from this patzer: the win hinges on the extra doubled pawn, and the two-file-separation-rule-with-many-exceptions has been drilled into our heads. Carlsen (who does not have a future in poker) certainly acted as if he were busted.
It should be noted that Mr. Smythe’s reference to the rule allowing the arbiter to declare a draw based on a player not trying to win by normal means is somewhat dated. Some time ago, that rule did appear in the Laws of Chess. Recently, the rules on “quickplay finishes” (what US Chess calls a “sudden death” time control period) have been moved to the end of the Laws of Chess and have been demoted to a “guideline.” Specifically, Guideline III is titled “Games without increment including Quickplay Finishes”. (FIDE tends to be somewhat allergic to “delay”. It does accept delay time controls, but the Laws of Chess don’t mention delay. Basically, US readers should think of “increment” in the Laws of Chess as “increment or delay”.)
Anyway, the situation with Guideline III is very similar to US Chess’s rule 14H (insufficient losing chances.) Article III.2.1 states:
(Yes, variation-ism is creeping into the FIDE Laws of Chess. YUK and BLETCH!)
Article III.4 states that a player with less than two minutes on the clock may request that a five second increment may be added for both players. (Guideline III only applies to games without increment or delay.)
Article III.5 states that a player with less than two minutes on the clock may claim a draw based on the opponent not trying to win by normal means.
Essentially, the whole mess is avoided by having an increment or delay. If the time control has neither, the whole mess is optional and does not apply unless the organizer specifically states that it does. I will withhold my opinion of organizers who are crackers enough to do this.
, the increment really should eliminate the danger of flagging.