February 15, 2013 11:51

Nepomniachtchi wins Aeroflot blitz after 'failing clock incident'

Nepomniachtchi wins Aeroflot blitz after 'failing clock incident'

Ian Nepomniachtchi won the super strong blitz tournament held on Thursday in Moscow as part of the Aeroflot Chess Festival. The Russian grandmaster was leading by a full point with two rounds to go, but these were the double rounds against runner-up Peter Svidler. The first was won by Svidler, who lost the second game, in the final round, on time. But not before the DGT clock behaved somewhat peculiarly.

Ian Nepomniachtchi clinches the Aeroflot blitz title

Many Moscow chess players left their lovers on Valentine's Day to compete in a massive blitz event which had a prize fund of US $50,000. It was a Swiss event which lasted 9 double rounds with 3 minutes and 2 seconds increment on the clock. There were 275 participants and, as you can expect in the Russian capital, there was a huge number of titled players: 107 GMs and 33 IMs. Unfortunately the games were not broadcast online and probably not saved digitally either.

After four rounds there was a surprising leader: Alexandra Kosteniuk had beaten all of her opponents with a 2-0 score. She then lost 2-0 to Ian Nepomniachtchi, who took over the lead. When he beat Alexander Grischuk 2-0 in rounds 13 and 14 it became clear that he was really on fire.

Update: here's one of these games, won very quickly by Nepomniachtchi, filmed by Eugene Potemkin:

PGN string

"Nepo" then held Sergey Karjakin to a draw twice, after which he was one point ahead of Peter Svidler, his opponent in the last two rounds!

In the first game (round 17) Svidler won with White, but in the second the grandmaster from St. Petersburg got under some pressure. Time pressure, that is! You can see the final phase of this game in the Chess.TV video from 18:20:01. We have reconstructed the following game fragment:

PGN string

Inspired by Chess-news, we created a number of screen grabs of the video mentioned above, to point out something peculiar.

Nepomiachtchi and Svidler playing. The clock has 0:52 left for White, 0:04 (!) for Black.
Remember that each move adds 2 seconds to the clock.

0:26 versus 0:04, before Svidler will perform ...Rb7-b2.

After ...Rb7-b2 Nepomniachtchi has played Qa2-g8 and Svidler now has one second left.

Svidler starts the move Rb2-b1+ while the clock is still showing 0:01.

After performing Rb2-b1+ Svidler presses the clock...

...which shows 0:02 for him. This is strange, but might just be possible. However...

...after the rooks are traded, Svidler's clock now shows 0:01. This is clearly impossible, because the clock certainly cannot calculate from -0:01!

Nepomniachtchi plays Kg1-h2 and Svidler answers with Qb1-h7...

...when his clock finally shows 0.00.

Svidler resigns...

...and the players aren't really sure what happened.
(To be honest, it's not clear whether they were talking about the clock here.)

We contacted DGT, the manufacturer of the chess clock, and a spokesperson confirmed that it should be impossible for a clock to show 0:01 at this time control. They plan to do more research.

Nepomniachtchi finished on a splendid 15.5/18 score, one point more than Svidler. Alexander Grischuk came third with half a point less. Anton Korobov finished shared 4th together with Sergey Karjakin, Boris Savchenko, Ernesto Inarkiev and Rauf Mamedov. Two young talents did well: Ilya Nyzhnyk and Alexander Ipatov both finished on 13/18. Kosteniuk eventually finished as the best female player with 11/18.

Aeroflot 2013 | Blitz | Final standings (top 50)

Rk. Title Name FED Rtg Pts. TB1 TB2
1 GM Nepomniachtchi Ian RUS 2723 15.5 206.0 172.75
2 GM Svidler Peter RUS 2697 14.5 194.0 153.75
3 GM Grischuk Alexander RUS 2860 14.0 207.0 150.25
4 GM Korobov Anton UKR 2700 13.5 207.0 151.50
5 GM Karjakin Sergey RUS 2901 13.5 204.0 144.50
6 GM Savchenko Boris RUS 2601 13.5 202.0 138.75
7 GM Inarkiev Ernesto RUS 2652 13.5 200.0 148.50
8 GM Mamedov Rauf AZE 2651 13.5 200.0 144.00
9 GM Fressinet Laurent FRA 2726 13.0 200.0 140.00
10 GM Nyzhnyk Illya UKR 2619 13.0 177.0 126.75
11 GM Ipatov Alexander TUR 2594 13.0 175.0 126.75
12 GM Mamedyarov Shakhriyar AZE 2734 12.5 210.0 138.50
13 GM Chadaev Nikolai RUS 2716 12.5 205.0 139.00
14 GM Khismatullin Denis RUS 2635 12.5 205.0 133.75
15 GM Andriasian Zaven ARM 2667 12.5 201.0 136.25
16 GM Kokarev Dmitry RUS 2635 12.5 196.0 133.00
17 GM Dubov Daniil RUS 2585 12.5 195.0 127.25
18 GM Sjugirov Sanan RUS 2627 12.5 187.0 126.75
19 GM Efimenko Zahar UKR 2719 12.5 187.0 125.25
20 GM Khusnutdinov Rustam KAZ 2617 12.5 183.0 123.75
21 GM Andreikin Dmitry RUS 2877 12.0 206.0 128.50
22 GM Sasikiran Krishnan IND 2652 12.0 198.0 124.25
23 GM Guseinov Gadir AZE 2554 12.0 197.0 126.25
24 GM Smirin Ilia ISR 2654 12.0 189.0 120.00
25 GM Morozevich Alexander RUS 2804 12.0 188.0 120.00
26 GM Timofeev Artyom RUS 2619 12.0 184.0 122.25
27 GM Popov Ivan RUS 2656 12.0 184.0 118.25
28 GM Dlugy Maxim USA 2602 12.0 184.0 117.75
29 GM Volkov Sergey RUS 2566 12.0 184.0 115.75
30 GM Durarbayli Vasif AZE 2598 12.0 182.0 115.25
31 GM Bukavshin Ivan RUS 2697 12.0 180.0 116.75
32 GM Rublevsky Sergei RUS 2630 12.0 180.0 115.50
33 IM Eraschenkov Denis RUS 2402 12.0 175.0 105.50
34 GM Stupak Kirill BLR 2615 12.0 157.0 104.00
35 GM Kamsky Gata USA 2745 11.5 210.0 126.00
36 GM Maletin Pavel RUS 2603 11.5 196.0 115.75
37 GM Jakovenko Dmitry RUS 2719 11.5 195.0 119.75
38 GM Eljanov Pavel UKR 2657 11.5 190.0 116.25
39 GM Kazhgaleyev Murtas KAZ 2670 11.5 190.0 115.50
40 GM Levin Evgeny A. RUS 2533 11.5 190.0 110.75
41 GM Motylev Alexander RUS 2659 11.5 182.0 112.75
42 FM Seliverstov Vladimir RUS 2549 11.5 182.0 107.00
43 GM Potkin Vladimir RUS 2665 11.5 180.0 109.50
44 GM Shimanov Aleksandr RUS 2705 11.5 172.0 100.00
45 IM Reshetnikov Alexey RUS 2448 11.5 169.0 99.75
46 GM Smirnov Pavel RUS 2569 11.5 168.0 105.00
47 GM Tkachiev Vladislav FRA 2681 11.0 199.0 116.00
48 GM Kosteniuk Alexandra RUS 2581 11.0 196.0 109.50
49 GM Wang Hao CHN 2724 11.0 192.0 110.75
50 IM Mesropov Konstantin RUS 2531 11.0 189.0 107.25

(Full final standings here)

Peter Doggers's picture
Author: Peter Doggers


achtie's picture

Would have been nice if you would post Nepo his win vs Grishuk in less then 15 moves

Peter Doggers's picture

Well, perhaps you can help? Are the moves available somewhere, or was this game broadcast in full in the video?

achtie's picture

:-) sorry, I hoped you would have it available in pgn.
I did see it on the live transmission (it was a caro kann) I will check in a bit to see at what time it was played.

Peter Doggers's picture

Found a video of this game online and added it to the article! :-)

achtie's picture

Awesome, you rock Peter!

Guillaume's picture

There is no time increment for Nepomniachtchi either during the last 2 or 3 moves (when he is down to 17 or 16 seconds).

Guillaume's picture

There is no time increment for Nepomniachtchi either during the last 2 or 3 moves (when he is down to 17 or 16 seconds).

Remco G's picture

My guess is that the clock just wasn't set correctly.

Anonymous's picture

Guillaume is correct - and if Nepo had taken 28 seconds to make his final couple of moves, then he would have lost the game.

Guillaume's picture

I think Svidler flagged when he played Rb2-b1+ with about one second left on the clock. The display after this move shows a little flag on the left side of Svidler's time. After that, there's no increment for either players.

Michel83's picture

If that is the case how did the clock go from 1s (4th picture) to 2s (5th picture) on his clock *without* increment?

Guillaume's picture

It was the last move to be incremented. It was also on that move that the "flag" appeared on the display left to Svidler's time. It doesn't look like a flag on the video, just a blurry mark, but check the manual of the DGT XL clock on page 2 Fig. 2.

Clearly, the clock seriously malfunctioned at this point: flagging Svidler, incrementing his time (possibly from a negative value) and letting the clocks run from then on without increments.

Niima's picture

I have heard from several people that despite being the official FIDE clock, DGT does not live up to high standards.

strana's picture

If Nepo's game were more consistent he would be the world best player, not only in blitz but also in classic. It is a big "if",ok, but his wins in Aeroflot when he was just 17 and his wins over Carlsen show how talented he is. Also his russian and european titles are a prove of it.

Anonymous's picture

nepo got some training from working with carlsen- but anand is champ b'cos of his wonderful hair

Anonymous's picture

it's the other way round. Carlsen is paying Nepo to improve

Anonymous's picture

could be ! he wants 2 nepam! the opposition - but anand's hair .. is hard to beat ;-)

ChessFan's picture

We need Apple to come out with the iClock and take over the market!

Anonymous's picture

Very nice video work by Eugene Potemkin!

RdC's picture

I believe clocks have settings for what happens when time runs out. It might seem logical for the clock to stop and a flag to indicate which flag fell first, but it may not be programmed that way by the organisers. Some arbiters cling to the long standing rule, still current, necessary for mechanical clocks, that if both flags fall, then the game is a draw. The DGTs have settings that will support this interpretation and continue to run even after zero.

Anonymous's picture

Don't be ridiculous. You can CLEARLY see that Svidler's flag falls @12:23:08 in the video, and Nepo never gets another increment following the 2 next moves.

There is ABSOLUTELY NO long standing rule that states if one player's flag falls, then the other does not receive his due increments.

Video: http://video.russiachess.org/view/1642

RdC's picture

Laws of chess 6.11
If both flags have fallen and it is impossible to establish which flag fell first then:
b.the game is drawn if it happens in the period of a game, in which all remaining moves must be completed.

So there is a law of chess which says that the game is drawn if both flags are down. Some arbiters interpret this to mean that the clock should continue to run and go negative and the game doesn't end until the players notice the clock. There's a setting on the DGTs which supports this. I don't know exactly how this is supposed to interact with increments.

Is this a plausible hypothesis?
Svidler's flag falls but his clock continues to run, reaching minus one second. He then presses the clock, adding two seconds, so the clock goes back to plus one second.

Anonymous's picture

Here's what happened in the video for those without time to view it -

1. With 1 sec on his clock Svidler takes 1 second to move Rb2-b1+.
A - http://oi47.tinypic.com/2vumq8j.jpg

2. Since the clock runs out, it flags Svidler and also awards the final 2 second increment to him as well - you can see the fallen flag symbol.
B - http://oi50.tinypic.com/1z57uc9.jpg
(Svidler's flag up-close - http://oi47.tinypic.com/2dhi349.jpg)

3. Svidler's flagged clock shows 1 second with his hand hovering slightly above to stop it, just as Nepo readies Kg1-h2...
C - http://oi48.tinypic.com/opbeqs.jpg

4. ...and Svidler's clock still shows 1 second as soon as it's pushed and Nepo moves the King.
D - http://oi49.tinypic.com/1ggj6c.jpg

The time between photos 3 and 4 could not have possibly taken 3 seconds, so the idea that his clock continued running to minus 1 second is wrong. So what happened?

The answer is that after Svidler's flag fell, the clock ran without adding players' increments, which means that it would have been possible for Nepo to lose the game if he had taken longer than 27 seconds for his final two moves.

Wouldn't that have been a h00t?

This is not the fault of the arbiters setting the clock, nor is it due to their interpretation of the laws of chess. The fault is with DGT engineering alone, and puts players at risk as long as the clocks are involved in FIDE events.

RdC's picture

DGT's have a setting where you can tell them whether or not to stop when a flag falls. If they stop the display freezes and starts flashing. It's evident that they have an alternative setting where the clock continues to operate but doesn't add increments.

As to whether Nepo could have also run out of time, then he could and the arbiters would then have had to award a draw unless they were watching or used the video evidence. DGT Clocks are capable of stopping on flag fall because I've seen games end that way. The Moscow ones were not set to do this. It's my view that the Law of Chess which proscribes a draw when both flags are down is obsolete for digital clocks and the clock should stop and clearly signal the end of the game when someone oversteps.

Anonymous's picture

You have been completely missing the point.

1. There is no rule that allows the arbiter to select a setting such that the clock will flag a player for running out of time, and then award that same player more time.

"You're out of time Svidler, and according to the Laws of chess...here's two more seconds."

2. There is no rule that prevents the 2nd player from receiving time increments if the interpretation of the rule is to let the clocks run after the fallen flag.

"Congratulations, Nepo, you've played the entire game so far with almost 30 seconds to spare, as FIDE arbiters, we must now take away your increments."

RdC, there is no need to defend them. You must understand there exists an engineering flaw in these DGTs that may alter the outcome of games, and this directly puts the player's livelihood at risk.

Anonymous's picture

DGT clocks are of course faulty, everyone who used them a lot knows that. It is sad how little anyone cares to fix this. Its just another chess mess. Probably the players are to blame for accepting to play chess under any conditions as long as there is chance to win money, but then again what can they do about it?? Hope DGT will take the matter seriously.

RdC's picture

DGTs have a setting that makes them freeze when a player runs out of time. I've seen this used at the 90 30 rate. Why don't arbiters use this setting at a 3 2 rate?

st32's picture

Nepo is smiling as if he f*rted in a closed room and is watching everyones reaction

Latest articles