Play Like Grob?
Monday, February 21, 2011 at 4:35PM Over on the ChessBase site there's a curious article by IM Andrew Martin. It's largely a tribute to IM Henri Grob, and there's nothing to scratch one's head about there. Indeed, it's nice to see articles praising the interesting figures of the past, especially lesser-known ones who are usually neglected. What is curious, or strange, or at least not clearly defined is the broader lesson Martin seems to want us to take from Grob.
The headline of Martin's essay is "Anyone up for the Grob?", which refers not to the man himself but the opening that starts with 1.g4. Martin becomes the article like this:
The difference between a professional and the other 99.99% of chessplayers is not always appreciated. Funnily enough, the refusal to accept that one is not a professional can form a barrier to improvement. For the amateur it is completely useless to try to play like a grandmaster, yet how often do we see this among our chess-playing friends. They keep up with the latest theory, assiduously study games from the latest GM tournament (or they think they do) and fail to make a single step forward in their playing strength.
The article concludes similarly:
My main point this time is that you should think very carefully before trying to emulate the randmasters. You don't have the time, you don't have the talent and you probably don't have the energy, unless you are under 20 years old. But what you do have is discretion and experience and those are invaluable tools which can be used to help you to get stronger at chess. Know your limitations and play to your strengths. That's what Grob did!
In between, the bulk of the article presents three of Grob's games, two where he's White and wins with 1.g4, and one where he wins with Black using the Englund Gambit (1.d4 e5 2.dxe5 Nc6 3.Nf3 Qe7 etc.). But is Martin suggesting that we use these openings? At one point, he says of the Grob that "1 g4 is not easy to meet, especially as there is a tendency to overreact when confronted with nonsense like this."
Now, this really isn't borne out by statistics. In ChessBase's online database, 1.g4 scores a dismal 47.7%, and an even lower 46.1 in Mega2011. If one only includes games with players 2300 and up, White's score falls to a catastrophic 26.1%. One might be inclined to say "Aha, that proves his point! Strong players have no trouble with it, but that's precisely not his audience! Filter out the games with serious players, and then White scores well." There's probably something to this, but less than you might think. There are only 69 games involving 2300+ players, while Mega has a total of 1593 games when the filter is off while the online database has 3049. The overwhelming majority of the games feature players in the rating range you'll find at a medium-sized chess club.
Additionally, Martin offers no fewer than four lines for Black that at least seem to be worthwhile from the way he speaks about them. And when we turn to the Englund Gambit, an interpretation that suggests Martin is in any way endorsing these openings becomes even harder to sustain. Theory has long had it in for the Englund, and he presents the known prescription and freely states that it "seems to place Black's whole idea into question."
Well, okay, but then what is the point of the article? Is it nothing more substantive than to avoid copying the (elite) GMs who spend hours and hours working on absolute main lines? That point could have been made without referring the avant-garde Grob (both the man and the opening). Or is it that we should, like Grob, jump off the deep end in our openings? If so, more of an argument should be supplied in its justification, for reasons I give in the next two paragraphs.
Perhaps the reason it's foolish for amateurs to copy GMs in the opening, assuming that it really is foolish, is not that there's something wrong with the opening but that they're spending their time on openings the wrong way, or spending too much time on openings, period. If that's the case, there isn't even the beginning of an argument for playing something like 1.g4.
Suppose instead that the, or at least a, reason why it's wrong to copy GM openings is that they are too conceptually rich for amateurs. Better they should know their limitations and play something simpler. Ok, suppose that's true. (I don't think that follows either, but let's pretend.) How does one get from "Don't emulate super-GM openings" to "Play gimmicky, positionally unfounded openings that score poorly"? In short, the connection between "Don't Copy GMs!" and "Play the Grob!" (or openings of that ilk) is somewhere between tenuous and non-existent.
Maybe Martin just means that we shouldn't spend, or even try to spend, the overwhelming majority of the time we have for chess trying to develop and memorize theory in the opening lines chosen by elite players. (Unless it's what we like to do for fun, of course.) If so, I would agree with him there, especially as one goes down the ratings & experience scale. But nothing follows from this about what openings one should choose! Someone might read Martin's article and think "Ah, he's right! I'm going to spend my time working on the Latvian (or the slightly better 1.f3 + 2.g4 Attack)!" This looks like the exact same error, except that one is choosing inferior and less rich openings to spend their time on.
Finally, unless Martin believes that IMs are untalented hacks, at least relative to GMs, there's something absurd about his suggesting that we not emulate GMs in all of their glorious talent, but should emulate IM Henri Grob! What happened to knowing our limitations? A common sense response is that we should follow Grob in some respects and not in others. Fine, but are we forbidden from using common sense to follow GMs in relevant ways, too?
In my experience, encouraging amateurs to leave GM openings behind doesn't usually help them improve. It just encourages them to get addicted to quirky openings, and to replace one set of growth-impeding habits with another. (It does make chess book publishers happy, though.*)
* Note: I have nothing against quirky openings. Some of my best friends are quirky openings. I DO have a problem with claiming that amateurs ought to play quirky openings rather than traditional ones. Occasional dabbling in weird, even dubious lines is fine for a little variety, to be forced to improvise and to expand one's sense of what's possible, most definitely. But as a staple of one's "diet"? No way.
Grob in Commentary, Openings
Reader Comments (7)
I remember an eye-opening experience at the 1988 British Championship. Andy Martin was commentator. He was discussing one game in which white (a "mere" 2200) had employed a variation of the Nimzo-Indian, that resulted in the traditional doubled pawn complex, against a GM. Martin "chastised" white for playing that line because it would be easy for the 2200 to go wrong and for the GM to neutralise white's play and cash in on the weak pawns, which the GM duly did. Verdict: wrong opening choice, despite the good theoretical reputation of white's line.
There is something similar in the classic "Chess For Tigers" by the late Simon Webb. Webb wrote something along the lines of: When choosing an opening line, it's all very well if it ends in a white += to you; but if the person who had white and went for that was Vassily Smyslov, you could find yourself in trouble because very few of us can play like Smyslov.
Lajos Portisch, in his essay in the book "How to Open a Chess Game" (RHM Press, 1974) writes: "It is illogical for a player who has not earned his master title to ape the complicated opening variations played by, say, a World Champion....Your only task in the opening is to reach a playable middlegame."
Lastly, my play improved immensely as a teen when I followed Korchnoi and Spassky's old trainer Zak's advice. I dropped the Sicilian Najdorf and went into the 1.e4 e5 Open Games. Zak's argument was that a study of the Open Games would teach one more about chess than some obscure tactics in the Najdorf.
All of the above complements what you've said and is probably the gist of what Martin was trying to say. He just didn't seem to do it well. The Grob is, in the words of John Watson, probably the weakest opening white can play!
I am still waiting for an article on Oostende 1937, where Grob beat Keres, Fine and Tartakower. How many can boast on that achievement?
Andrew Martin mentions the message of his article clearly only in one the game annotations. He says, "Whatever your level, it's important to have a style all your own. Slavishly following the Grandmaster arena and trying to copy what is there brings you nothing in the end. Of course this does not mean you should rush out and start playing 1 g4! ".
I think I agree with him, but as you said, his main article does not really convey this message clearly.
Another great blog entry from Dennis. That's why I read him every day!
There's a difference between "quirky" openings and just plain junk. I research lines that can be deemed quirky only if they seem to be unfairly dismissed as not worthy of main line status. (The Bird Ruy comes to mind, and the Evans Gambit!) Martin's article seems to be one of those cases of: write article, get paid. If I actually starting seeing Martin games in the database featuring 1 g4, I'll pay attention!
The problem with Martin's approach, such as it is, is that it ignores fundamentals that have to be learned. To me, it's not such much a question of 'which openings should I study?' but 'which order should I study openings?' Better yet, don't study opening variations, study 'the opening' and it's historical development. Why did the main lines become the main lines, while others lie dormant? In this context, learning all the classical stuff first gives one the foundation to expand one's study to hypermodern fare.
A better way to leave the beaten path without getting too deep into the weeds is to search out lines like, for example, anti-Sicilians 1 e4 c5 2 Nf3 Nc6 3 Bb5 or 1 e4 c5 2 Nf3 d6 3 d4 cd4 (3 ... Nf6!?) 4 Nxd4 Nf6 5 f3!? Anti-Grunfeld: 1 d4 Nf6 2 c4 g6 3 f3!? With Black, say you like the Open Ruy, 1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Nxe4 6 d4 b5 7 Bb3 d5 8 de5 Be6, but dislike the Exchange, 4 Bxc6. In that case, there's the order 3 ... Nf6 4 0-0 Nxe4 5 d4 a6!? Now 6 Ba4 b5 7 Bb3 d5 8 de5 Be6 returns us to the Open while dodging the Exchange. This assumes that you then find 5 ... a6 6 Bxc6 dc6 7 Qe2 Bf5 acceptable for Black. One more example, versus 1 d4: say you like the QGD Cambridge Springs but don't care for facing the Exchange Variation in its pure form, e.g., 1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cd5 ed5 5 Bg5 Nbd7 6 e3 c6 7 Qc2 Be7 8 Bd3. Now 8 ... 0-0 leads to the well worn established lines, but 8 ... Nh5!? is interesting try, leading to 9 Bxe7 Qxe7 followed by ... g6, ... Nb6, preparing queenside castling. Everything's different now! These are just are just a couple of random rare ideas, given to demonstrate that the best new ideas are probably to be found in the main lines but just under the radar.
Above a certain rating level opening study is comprehensive and time-consuming whether one plays main lines or not. The problem with junky sidelines is that they have no shelf life, especially in this age of the databases. If you can't play the same bad line against the same good player twice in rated games, it's a bad line! John Nunn once wrote that moving from one sideline to the next is actually more work than maintaining the same set of main lines.
Capablanca once wrote that the Ruy Lopez was the cornerstone of opening understanding, or something to that effect. I'm not sure that studying the Grob contributes to anything other than learning damage control after a weakening and wasted first move with White. Martin's article falls short because it wasn't really well thought out, and the caveats contained therein were certainly inadequate to the larger subject at hand.
Strong contender for worst-ever article on chessbase?
Last year a local player here won a 6 round weekender 1600-2000 Elo category all of his games featuring unusual off-beat openings - all of which he new very well. I am not sure how he will get on at this year's event (actually next weekend - Bunratty Master) when he completes in the Master's section which incidentally includes Svidler and Short!
At a certain level there is no problem playing 1. g4 (or "the Spike") especially if you enjoy chess - guaranteed to provide an interesting games. I think every young player has to have a go at playing all of these kinds of openings especially against weaker opponents - unfortunately they don't work against stronger opposition. Many years ago member at our local club used to push all of the young players to play the Scotch and Goring gambits - of course we all got thrashed but we all learned got several lessons in tactics!
Mr Martin is trying to make a living so "of course" 1.g4 is what he advocates. Next week it will be St. Georges. defence (1. e4 a6!?) with which Miles beat Karpov no less!
[DM: The thing is, though, that he doesn't exactly advocate it (and definitely doesn't commend the Englund Gambit), but something more nebulous like playing in Grob's spirit. (The man, not the opening.)]
To my mind, the most crucial statistic to evaluate Martin's argument is: when White is ranked 400+ points below Black, does the Grob perform better (or even comparably) to mainstream openings? If the Grob scores below 50% overall, but even slightly improves White's odds when the rating differential is huge, then his point stands.
Simon Webb's advice comes to mind, i.e. that when playing higher-ranked opposition, the best thing to do is to make the position as messy and unfamiliar as possible. After all, a huge part of what makes a strong player strong is pattern recognition. If the weaker player can short-circuit that, even at the cost of a somewhat dubious position, then it's a bit like taking a strong tennis player out of his rhythm -- it means that technical and psychological weaknesses come into play. It's the "winning ugly" approach popularized by Brad Gilbert: play what will make your opponent annoyed, frustrated, even insulted (à la Karpov-Miles). Which means that if we ever meet OTB, Dennis, I'll be tempted to play the Latvian. :)
When playing against peers of the same strength, it's a somewhat different story. There I personally think the best strategy is to adopt either trappy openings with a basically sound foundation -- like, say, the Max Lange Attack -- or unbalanced, slightly dubious lines which are annoying to meet and require very accurate play to punish, like the Caro-Kann/Modern hybrid with ...c6 and ...g6, or the Four Knights Sicilian with 6. Ndb5 Bb4 and ...d5.
[DM: The problem is that there isn't anything "trappy" about 1.g4, as long as the player doesn't rush to take on g4 - and even that's okay as long as he doesn't panic after Qb3. It's weird, not tricky or trappy. Something like the Max Lange is a much better way for a lower-rated player to go after a strong opponent. If Black knows everything, then White is still fine; if he doesn't, he can perish quickly.
Looking forward to 1.e4 e5 2.Nf3 f5 and ...1-0!]