Thursday, March 26, 2015

A Latvian Gambit on LSS

     The late Ken Smith advocated playing gambits. Writing about 'Romantic chess' he said, “Romantic chess starts with the opening or defense. For example: White will play the King's Gambit, Vienna Gambit, BDG, Scotch Gambit, Danish-Goring Gambit, or the Evan's Gambit, and there are more.
     Black will play the Albin-Counter Gambit, Henning-Schara Gambit, Englund Gambit, Latvian, Elephant and many more. The ones playing gambits are examining the Romantic side of chess beyond a closed door. HOW COULD YOU POSSIBLY LIVE YOUR CHESS LIFE LOOKING AT A DOOR AND NOT OPEN IT??”
     As I wrote in a previous post, any unsound gambit is likely to be a successful weapon at lower levels. But just because 99 percent of us aren’t good enough to refute them does that mean it’s a good idea to play such an opening? It’s argued that by playing them you can improve your tactics, but couldn’t you just as likely win a tactical battle in the Najdorf Sicilian or the Nimzo-Indian? If your opponent is weaker (or maybe even a little better) than you isn’t he just as likely to make a tactical mistake in the Ruy Lopez as in the Latvian Gambit? Are you less likely to blunder in the Queen’s Gambit Declined than you are in the Blackmar Gambit? The point is, as GM Alex Yermolinsky argues, that it's better to start out with a good, solid position in a mainline opening than it is to hand your opponent a material and, usually, a positional advantage right from the start then struggle in an inferior position hoping your opponent makes the first tactical error.
     Still, gambits have those who firmly believe in them and I have even experimented with some offbeat gambits on Lechenicher SchachServer where engines are used. That's probably not a good idea because engines aren't going to make any tactical mistakes, but the results have been somewhat surprising. However, that hasn't been do to the value of the gambit; it's the way my opponent's have selected their moves.
      The games have been more interesting than following 20-30 moves of theory in the Najdorf Sicilian, Ruy Lopez or Nimzo-Indian and the results have been no worse.  Does that mean the gambits are sounder than it is generally believed? Probably not but many CC players are playing a lot of games and don't take the time to do a very through analysis; they just play the engine's first recommendation after letting it think a minute or two and that will often lead to disaster.
     If you want to play those risky gambits you have to be prepared to do a lot of research and analysis and it's scary looking into a position where all the engines are telling you that you are at a significant disadvantage. It proves Yermilinsky is right.  A lot of these gambits do start you out in an inferior position. So, why play them in CC events where engine use is allowed?  It makes the games more interesting and sometimes you get to give a big sigh of relief when you gradually see the evaluation starting to shift your way. Sometimes too, actually quite often, opponents who are playing a lot of games and who rush their moves will choose a second or third best move and you find you are beginning to gain the upper hand, then suddenly your opponent realizes he has stumbled into a lost position...it's not a very scientific approach but it beats slogging through an 80 move ending! What I usually do before playing such a gambit is take a look at how many games my opponent has going.  If he has a whole lot of them I know it's a safe bet he isn't spending a lot of time on analysis so it's possible some offbeat gambit will work.
     While looking over some games played at LSS I discovered the following Latvian Gambit game. I don't know much about the Latvian gambit, but it's one of those gambits that has a long history of diehard adherents. Not knowing much about the Latvian, I don't know if the players were following 'theory,' but it seems that black was never quite able to equalize or even generate any serious threats, but playing it was a gutsy move on his part. Also, I don't know what engines were used or how much analysis went into the game, but a lot of the moves didn't agree with Stockfish 6 or Komodo 8 running on 2 cores and in the fairly cursory (a minute or two per move) analysis I did, both engines offered what may have been slight improvements.

Some interesting sites on the Latvian Gambit:  The Chess Website, Dan Heisman's SiteChessdotcom article #1 and Chessdotcom article #2.




No comments:

Post a Comment