Popular

Ozymandia on the Importance of Blunders

By Juan Molina aka Ozymandia (28th April, 2014)

Let’s get one thing out of the way before anything else, this was not a good tournament for me. Not only did I lose a game, but I failed to repeatedly beat the players at the bottom of the table. If there’s one thing you can learn by simply looking at the final cross-table, is that to be at the top you have to really abuse the weaker players.

That said, there’s something other reports have failed to show, and that’s the importance of blunders. Everyone has been praising the chess played here as of the highest quality, and while that has been the case 95% of the time, the influence of the other 5% can’t be overlooked. People missing games, or playing as stand-alone engine instead of centaur, or making result-altering mistakes... all that has an impact on the final standings. Once taken into account, there’s no longer a big gap in performance between the first two players and the third (20 and 30 ELO points). Incidentally, that third player would be Mig29.

Let’s see those blunders. First those which clearly change the outcome of the game:

 

Ocirema - LordSirKnight

A mouse slip (Qg4 instead of Qxh4) and black gives the game away. White is a pawn up, but the Knight on e5 should be enough compensation to hold the draw.


Akhtar - jobboy

A mistaken input (Qc4) allows mate on g7. This should’ve been a draw.



Newton296 - Paul
A mouse slip (Qa1+ instead of Qb1+) and black can’t force a threefold.

Frauholle - mateinfour

Engine bug (probably something to do with the hashtables). Rc8 is played directly, instead of 27.Nxg4 Bxg4 28.Bxg4 Rxg4 29.Qh3 Qg5 30.Bxh4 Rxh4 31.Qe6+ Kh8 and only then Rc8. Looks like a perpetual for white can’t be avoided.

Olivierevan - Regina_H.Mllch

Tablebase draw.

(White lost due to the blunder Ke5??. Cf. Black's tour report. A.N.)

Now the games where one side has some sort of advantage, but it’s not clearly decisive. As in any engine test, where crashed games are removed before computing performance, I also deleted the next ones before proceeding.

Ocirema - Victorious

A mistaken input (Re7 instead of Rxe2) allows white to keep his Queen and the full point. After 31.Rd1 Rxd2 32.Rxd2 Bc4 33.Rxc6 Ne7 34.Rxc7 Nxf5 white still has material advant-age, but decisive?

Akhtar - Olivierevan

Black leaves the board with a slight advantage for white.

Deepthroat - sunnytown

Black resigns. Slight white advantage.

(Black ran scared under time pressure, assuming he had blundered. A.N.)

EtaoinShrdlu - Sun_Zhu

A mistaken input allows white to take on g4.

Funvillage - Intagrand

An engine bug. White doesn’t take the pawn on d6, there’s a clear black advantage nonetheless.

Newton296 - Olivierevan

Mistaken input. Na5 (instead of a5) allows black to shift the evaluation, but just four moves later, black returns the favor.

Olivierevan - Victorious

After 43.Rb3 Bxe4 44.Qf8 black has the upper hand, but after the immediate Qf8, black takes the Rook with check.

Sunnytown - Thomas_A_Anderson

Black has the exchange, but things don’t look so gloom for white, as to be forced to surrender.

 

Donkasand - Stabiloboss

Barely out of book, black commits positional suicide with ...c5.

(Black was a Stockfish engine, but the mistake ...c5 could not be reproduced. A.N.)

Jobboy - EtaoinShrdlu

Mistaken input. White has a clear advantage even without taking the on f5, but there was no reason to give away the pawn.

On a final note, the game Paul - Regina_H.Mllch could be considered the April’s fool of the tournament (it was played, after all, on April 1st). In the final position, after Qd4, a draw is agreed.

Note by the Tournament Director:

Actually Regina_H.Milch had offered a draw just before Qd4 was played, and Qd4 was in fact intended as a joke by Paul, who made use of a "server bug", as he knew that the draw offer was still valid after his move.

For explanation: The "server bug" is related to engine games. Here games are often so fast that players are unable to accept a draw offer. So the application allows players for a certain period of time to accept a draw offer even after further moves have been played. This method should of course not apply for human games and freestyle games, and I expect this soon be resolved, if it isn't done yet. A.N.