Komodo Beats Erenburg In Time Odds Match

You asked for it, and Chess.com delivered! Followers of the Man Vs. Machine matches between Komodo and human grandmasters have long been asking for a match based on time odds. So be it. GM Sergey Erenburg waged battle against the world's top chess engine on June 25-26.

The match was played in GM Larry Kaufman's home in Maryland. Kaufman is, for lack of a better term, Komodo's mentor. Erenburg played on an actual board while Kaufman relayed the moves from the computer to the board and back.

In total, Erenburg received six different odds. These were as follows.

• Erenburg was granted 30-1 time odds. He had 90 minutes plus a 30-second increment; Komodo had three minutes plus a one-second increment.
• Komodo was limited to a single core. Komodo normally uses 24 cores.
• Komodo was not allowed to think on Erenburg's time.
• Komodo's opening book was limited to three moves.
• Erenburg had the white pieces in all games.
• Komodo did not have access to tablebases.

Sharp readers may wonder if it was possible to flag the computer thanks to time consumed in the relay. Sadly for Erenburg, this was not possible; Komodo's time was calculated independently by the engine. No time was taken off in the relay process.

Game 1

As the match got underway, things looked rough for Erenburg early on. Some advantage may have existed in the opening, but it was soon gone. 

Komodo demonstrated its ability to plan and played 19...Rfe8 and 20...e5 to break out of an initially passive Caro Kann position.

On move 23, Erenburg had to choose between 23.Nd4 and 23.Rf1. Nd4 accepted a weak isolated pawn. What about Rf1?!

GM hissha (Kaufman): Komodo expects Nd4, so something must be dubious about Rf1.

GM hissha: and now we see what!

What turned out to be 23...Re3!! You just can't turn your back on Komodo

Game 2

Game two was the kind of nasty slog that is so miserable for the human players to suffer through. Move by move and tenth of a pawn by tenth of a pawn, Erenburg's position worsened. 

It wasn't long before the chat's only question was if Erenburg could flag Komodo. Sadly for humanity, that wasn't possible as Komodo always had the one-second increment.

Game 3

The most interesting side of game three was the opening. Komodo, playing without an opening book, reproduced considerable theory in the Sveshnikov Sicilian. As Kaufman put it...

GM hissha: Komodo "invented" the Sveshnikov for this game! 

Perhaps the one mistake Komodo made in this mistake was 33...h6 which did not get commentator GM Alex Yermolisky's seal of approval. That move left Black with an awkward kingside structure that precluded pawn breaks on the kingside.

With the kingside off the table, Komodo opened the a-file instead, but Erenburg handled that well by pressuring Black's d-pawn. With neither side making progress, a draw was agreed on move 40.

Game 4

The final game much resembled game three for a long time. Erenburg played extremely solid chess and held off all attempts by Komodo to introduce tricks. 

Unfortunately for Erenburg, this time Komodo's maneuvers bore fruit, and on move 39, in a position that seemed almost unavoidably drawn, Erenburg slipped and allowed Komodo to win a pawn. The resulting queen ending was lost, and Erenburg soon capitulated.

Final Match Standings

Erenburg's final tally was disheartening,  but getting a draw against Komodo without material odds was quite an achievement. Kaufman postulated that at these odds, Carlsen might be a match for Komodo , and while that seems optimistic, it is conceivable that a top grandmaster might be able to hold his own against Komodo in this sort of format.

You can replay all games with commentary from grandmasters Alex Yermolinsky, Simon Williams, and Maxim Dlugy at Twitch.tv/chess.

Watch live video from Chess on www.twitch.tv

In addition to our special Komodo matches, be sure to check out IM Daniel Rensch in his #ChessMonday show Man vs Machine. Watch every other Monday at Chess.com/TV.

A complete Man vs. Machine historical archive is available here.