How to beat a Grandmaster

I feel uniquely qualified to provide this advice, not that I’ve ever even taken a GM all the way to time control, but I do have 4 IM draws notched on the gunstock, so really it’s just a matter of time, cough cough, another 300 ELO and I’m so there. But why wait? Here is your crystal-clear, utterly failproof step-by-step instruction manual for beating a Grandmaster.

1. Play him in, like, 150 rated games. Eventually he will tire of flogging you in his favorite openings, his offbeat openings, in tactical flurries, positional squeezes, and every imaginable endgame. And his boredom will eventually coincide with him eating a bad burrito for dinner, plus having a spat with his significant other and/or a falling-out with one of his children or some other relationship calamity, plus a fender-bender on the way to the club, and in this bored and nauseous and emotionally unhinged state he will blunder and give you the opportunity to pounce.

That’s it. I’m telling you, it’ll work. Give it time.

(Plan B – as you take time-scramble notation, poke him in the eye with your pen.)

Advertisements

15 thoughts on “How to beat a Grandmaster

  1. According to this link

    http://www.ascotti.org/programming/chess/elo.htm

    someone of your caliber, Derek, at about 2150 USCF, should beat a 2550 GM about 8% of the time. This means you need only 13 games to do it.

    Why does this not seem right, intuitively? This result does not seem to happen, though — not just for you, but in general. This means you should nick a draw about once every 25 games from a GM, but does it really happen this often?

    Howard

  2. This win% correlating with ELO difference I’ve always heard, but wonder how reliable it is. Without having access to data, my instinct says the percentages vary based on actual rating.

    Take the match 800 vs. 400. How comfortable would you feel betting money on the 800? There doesn’t seem to be a great deal of actual skill difference at this level. At the very least, just about anything could happen.

    A 1600 vs. 1200? The 1600, while by no means great, has a certain level of experience and/or talent. He is competent enough to usually to defeat the weaker player. I’d feel comfortable betting small money on the 1600.

    Now the GM vs Derek? The GM is very experienced and talented, and is likely life-dedicated to the game. While Derek is also very good, he won’t find the GM dropping a Knight (barring his plan A working out once every 150 games). The blunders are a greater possibility in lower levels of play, which gives the lower-rated players a better % to win.

    Hopefully Derek is OK with me referring to him in third-person on his own blog.

  3. Re: 3rd person — He actually kinda likes it.

    I think the stats probably reflect accurately what happens across the whole universe, and have little predictive value for betting on a single game.

    Re: Blunders, of course GMs and IMs blunder. Actually IM Foygel (2500ish) has made clearcut, significant tactical oversights in a few of our games. (Let me count. Um, I can think of two in games I’ve lost anyway, plus at least one such mistake in each of the four draws.) Of course I’ve made clearcut significant tactical blunders in, let’s see, at least a dozen of our games.

    The key to blunders is pressure. If you can put pressure on your opponent — positional pressure plus time pressure — the possibility of a catastrophic mistake arises. The problem is that it’s really tough to put these guys under pressure. They just play so many darn good moves. It takes a master to put consistent pressure on other master-level players.

    Of the four draws, in three of them I’ve sacrificed material for initiative/development/attack – one pawn (once) or two pawns (twice). I can’t say that’s a recipe for everyone, but those are the positions I tend to play better. Of course in the vast majority of our games I don’t generate any such play; he’s the one turning the screws.

    Side note as I think of it: Of the six “blunders” I can count by Foygel, exactly one of them occurred in a position where I had no pressure on him. He was grinding me to dust quite effortlessly and I think went into autopilot mode for one move, allowing a trick that got me into an almost equal ending. Which he then won in eight or ten more moves.

  4. I’ve played just three rated games against IMs (I have two draws, one loss), and I have only played one grandmaster in a tournament game. While I did lose the game to the grandmaster, transitioning into the endgame I missed a tactic that would have eventually forced the win of a pawn. The extra pawn, together with a much better remaining pawn struture in a rook and minor piece ending, should have been enough for a relatively simple win.

    Titled players definitely make mistakes. I feel like when I get a chance to play against these players, I play better because I really try to bring the strongest game I can. Also, when you’re playing someone who is “supposed” to beat you 96% of the time, the pressure is off.

    That said, a car accident or relationship blowup in the titled player’s life would really help on gameday. You could even get proactive about it. Stop off and cut their brake lines on your way to the tournament site. Hit on their significant other (for this approach you’ll want to chose your grandmaster target wisely…in the even you have to follow through with something, you might wind up thanking yourself).

  5. I’ve played just three rated games against IMs (I have two draws, one loss), and I have only played one grandmaster in a tournament game. While I did lose the game to the grandmaster, transitioning into the endgame I missed a tactic that would have eventually forced the win of a pawn. The extra pawn, together with a much better remaining pawn struture in a rook and minor piece ending, should have been enough for a relatively simple win.

    Titled players definitely make mistakes. I feel like when I get a chance to play against these players, I play better because I really try to bring the strongest game I can. Also, when you’re playing someone who is “supposed” to beat you 96% of the time, the pressure is off.

    That said, a car accident or relationship blowup in the titled player’s life would really help on gameday. You could even get proactive about it. Stop off and cut their brake lines on your way to the tournament site. Hit on their significant other (for this approach you’ll want to chose your grandmaster target wisely…in the even you have to follow through with something, you might wind up thanking yourself).

  6. Besides taking your experience into consideration (the masters are human, too!), another element I did not consider is for a given level of play, blunders might be defined differently.

    Leaving a piece en prise for nothing is a blunder at any level of play. As one moves up the ladder, start throwing in forks, combinations, position, etc., and defining an “obvious” mistake changes.

    The GM won’t be hanging pieces as often as the 800 guy, but that doesn’t preclude other types of mistakes. The 2150 has a reasonable chance to see a 2550 blunder which may be missed by the 1600 or 800. The ELO table applying to all rating sets seems more reasonable with this reasoning. (Hope it’s clear; it’s a stream of consciousness more than anything.)

    Still, maybe some weekend I’ll research actual game data (or more likely, someone’s done it already) and see how the table holds up.

  7. My boy will become the best of all chess players in the world (and a GM, too). So far, he’s learnt by himself and has already won many trophies but next year I will look for a trainer.
    Do you also play online?

  8. I’d start dressing up like the GM and mock him while playing.

    If that doesn’t work, you could accelerate the “bad day prior to the club” scenario with a little “social engineering”. Start asking his wife out for coffee and lie. Tell her you find her more fascinating than a silly old game of chess. See where it goes from there. :)

    Yes, the dark side runs deep in this one.

  9. I had one win over a 2200–I’m pretty sure he’d been drinking. While I did not contribute to his situation, this would seem to offer another avenue for a bit of gamesmanship; pay the bartender to serve you colored water while your opponent gets plastered

    Given the suspicious nature of chess masters, though, they’d probably smell it out.

  10. Robert, Blunderprone, Pete: very nice ideas. Remind me not to play poker with you. Or chess for that matter.

    Egg: Nice link. Now THAT’s a blunder.

  11. Hey – just stumbled on this blog while looking for US Amateur Team results.

    I’ve beaten GMs, so let me add a couple of thoughts here. I’m a weak FM, semi-retired from tournament play now.

    My first rated game vs. a GM, I won with an easy sacrificial attack. It was round 1 of the World Open (many years ago), and he was probably tired from traveling. I felt like a rookie homering in his first major league at-bat.

    My next three GM games went loss, draw, win (1st round of New York Open, probably same reason). I started to think, “Hey, these guys aren’t so tough!”

    Then they beat me the next 17 games in a row (before I groveled an endgame draw with a bored Anatoly Lein at the USATE). I gained a whole lotta respect for them.

    Genuine tips for success vs. GMs:

    1. Play main-line openings. Yes, of course they know them better than you … but there’s a _reason_ those are the main lines: because they’re the best moves. (I tried some goofy b3 system vs. Dmitry Gurevich’s Sicilian once, and when we went over the game [which he won easily], he said, “After this move it is like _I_ am white and _you_ are black.” Words of wisdom.)

    2. Don’t be afraid to sac a little something when a good opportunity arises. Both of my wins came when a pawn sac forced my opponent to switch to the defensive and he erred. (But pick your moment – give away a pawn without a good reason and you’ll be painfully ground down in an ending.)

    Good luck (or as my dad used to say as I left for a tournament, good skill).

  12. Randomo – cool stories, thanks :)

    My teammate Denys (2475) was talking about his relatively poor score against GMs. I said something like “well hopefully you’ll get another shot this weekend” and he replied “it’s certainly a good opportunity since they show up hungover” (you may have noticed a prominent example or two at USATE over the years)

Comments are closed.