Thursday, May 31, 2012

Openings Manifesto

I just re-read this PDF:  Training in chess:  A scientific approach.  It emphasizes things that should be familiar to most Knights Errant, and most chess students.  But it also argues persuasively against many chestnuts.  I'll probably refer to it from time to time, as I have a few different ideas about studying than the traditional advice that gets passed down through the generations.  Today I'm concerning myself with the openings, and choosing an opening.

The typical advice that you see spewed out by wise coaches and spelling-challenged commentators alike is:  "Don't waste your time learning an opening!  As soon as you get done with your eight moves, or your opponent deviates, you're stuck and lost!  Learn endgames first, study a general strategy book, and play over master games."

As soon as you get done with your eight moves, you are not lost.  You are presented with a position that you need to come to understand.  You will see this position later, when you review your game, when you practice your opening, and when an opponent presents you with it again.  It is inefficient to keep chucking your pieces out there unsystematically and presenting yourself with new positions time after time.  As much as I love and value Silman's strategy workbooks,  I am not Silman and shouldn't yet be trying to figure out plans by looking at an unfamiliar position.  I should already have a clue as to what my plan should be, and test to see if it's still applicable and why.  You are much better off presenting yourself with a Dutch Stonewall repeatedly, and learning how to get your knight onto e4.

The same applies to endgame study.  Beyond learning the common endgame techniques, you shouldn't be studying general endgames.  You should be studying endgames that can be reached from your openings.  You should learn to win the endgames you've encountered in your own games, and learn to win won endgames of master games in your openings.  Of course, I'm dubious about the value placed on studying endgames for beginners, but it does provide good tactical training and brute force training on how the pieces inter-operate, so it's probably good advice after all.

The same applies to playing over master games.  You shouldn't be studying Fischer's games if you're trying to master 1.d4 or 1..f5.

Pattern recognition is key to chess mastery.  You need to master your positions, and it all flows from the opening.  Obviously, this isn't black and white or absolute or maximizing of optimal joy, but yes, every adult chess student should pick a partial repertoire of openings they like and stick with it for a while.  Yes, they should learn some of the moves by rote.  They should understand the plans they need to execute when they get out of their short book.

So now we turn to me, because it's my blog.  It will become clear shortly that I have typical newbie openings angst, the type of angst that the "Don't learn openings!" crowd is kindly trying to protect me from. The situation is this:

  • I should learn aggressive openings and learn how to attack.
  • I'm uncomfortable playing in a gambit style and wonder if I should choose aggressive non-gambits.  It seems stressful.
  • But I do have visions of swashbuckling at the local coffee houses!
  • I do value learning plan-based openings, with consistent structural demands.  "Attack" might be too broad a plan.
  • I value offbeat, especially at low to mid-amateur levels.  However, I'm conservative by nature.
  • I'd like to get onto the business of choosing a long-term repertoire, and learning those patterns and strategies.
Okay, let's get to it!

1. e4.  Gambit-y.  If not 100% gambit-y, it's because 100% gambits seems a little exhausting and stressful.  Do I have the wrong attitude?
  • Evan's Gambit (Italian)
  • Smith-Morra Gambit (Sicilian) - The Wing Gambit (2.b4) seems interesting, because it's easy to get to and gets me started on the 2.a3 Sicilian.  But SM gets me started on the Alapin.
  • Cochrane Gambit (Petroff) - Every gambiteer probably has one or two favorites.  I love this one!  How come the whole world doesn't play this gambit?!  Don't answer that.
  • Panov-Bottvinik (CK) - Maybe. But there are some gambits I might like.
  • Alekhine-Chatard (French) - The Alapin comes sooner, however...

Okay, this is where the crazy starts.  My Black choices are still in committee.  The bipolar nature of my choices reflects this.  Generally, these are the competing ideas: 

a) I think I might like the Dutch, starting with the Stonewall.  The plan is clear enough that even I might be able to get a handle on it.  I like the idea of sometimes having something different to learn about chess than gambits. I like the idea of taking a break from the gambits.  I also like the idea of having at least one "solid" opening I might like to keep for the long haul.

b) But the Leningrad looks good too and is tactical.  Aesthetically, I like the opening setup and the e5 plan.  However, my opening book says that the Stonewall position is crucial to understand even for the Leningrad.

c) I believe in the power of offbeat openings.

d)  Maybe I should just gambit it up as Black as well. The Dutch seems like a jarring departure from my themes.

Anyway, enough preamble.  These are the current choices:
  • Against offbeats: Just wing it for now.
  • Against 1. d4 and friends:  Dutch Stonewall, possibly starting with 1..e6 because...
  • Against 1. e4:
    • Balogh Defense (1.e4 d6 2.d4 f5), or 
    • Kingston Defense (1.e4 e6 2.d4 f5), or
    • Kingston Defense, Franco-Hiva Gambit (1.e4 e6 2.d4 f5 3.exf5 Nf6?)
Both the Balogh and the Kingston are Staunton Gambits Declined, and the move order avoids some of the anti-Dutches (I think).  I could learn the Staunton Gambit Accepted, but I don't like having to defend against a gambit when I'm the gambiteer!  Why should I let my opponent have that "advantage"?  I prefer the idea of turning the tables.  If I learn the Balogh or Kingston, and even play it as my e4 defense, then I will be very comfortable when I decide to face the Staunton Gambit.

The idea is also that learning an unknown, somewhat weak opening is a nice placeholder to cut down on the theory I have to learn initially, a placeholder that will still have practical value against the Staunton Gambit when I start to fill in my 1.e4 defense repertoire.

The computer engines don't think much of the Kingston, and they hate the Franco-Hiva.  However, the Balogh is somewhat weak, but not horrible.  Either would give me home turf advantage, and move order flexibility.

Or, alternatively as a Black defense, I could
  • Learn the Scandinavian gambits against e4 (Marshall, Portugese, Icelandic)
  • Learn the Soller Gambit against d4
The Soller and the Icelandic look cool.  But I feel obligated to learn a lot of theory for all these gambits.  And it all seems so "nerve-wracking", as Temposchlucker concluded after seven years as a gambiteer.  In fact, I'd say that Temposchlucker's repertoire decisions are very close to mine, except I'm several years behind him.  Non-risky by nature, we realized we needed to learn tactics and how to attack.

It's way too soon to decide, but I imagine myself doing something in the Bird/Dutch/Polar Bear.  I find the idea of consistent pawn structures and plans very appealing.  I like the Dutch, and I like that it's less popular.  But who knows.  I have a lot to explore.

It's still Tactics, Tactics, Tactics.  This is a lot of words for very little upcoming study.  But I'm getting ready!


  1. The truth here is that you have to improvise and be flexible. Openings often have typical plans, but untypical plans are required just as often. There are no easy answers.

    Dan Heisman recently quoted GM Maurice Ashley as saying that he tells his students not to do any detailed opening study until they are 2300. Dan said that was a bit extreme, which sounds a fair assessment to me. Nonetheless, there is no point in learning GM openings if you cannot play the resulting positions, so basically he is right.

    Opening moves are a compromise between:

    * Instructive value.
    * Objective value (different for you and a GM).
    * Surprise value.
    * Amount of work.
    * Nice explanations in books.

    You can just improvise, select moves from MCO or use books like the Starting Out series. Improvisation is best at the lower levels.

    As regard to specifics, the usual advice is just play 1.e4, reply to 1.e4 with 1...e5 and 1.d4 with 1...d5.

  2. Thank you for the advice. You're right; I suppose it's easy to arrive at uncommon positions in your own common openings, so I should be careful not to overthink it.

    The internet also is a good alternative to books for lower players.

  3. I do not know what level you are at. Up to about 1400, you should find all you need to know about openings in books that start off by explaining the moves. These books should not be underestimated! Woolum suggests 50% on tactics, 25% on endgames and 25% on everything else.

    Albert suggests 25% on the opening. This might equate to 25% opening, 25% tactics, 25% strategy and 25% endgames. It is a mistake to think that you can spend 50% of your time on Starting Out opening books and learn about strategy from the illustrative games.

    It is easiest to learn strategy from old games where one plan is pursued from start to finish. Modern GM games (and indeed most practical examples) are more complicated. Usually, your opponent has some compensation and is pursuing his own plan, or stops your plan at the expense of incurring a new disadvantage, which requires a new plan.

  4. My skill: Ten years ago, I played six games in the USCF. I lost them all! That was fine; I expected it. I was completely outplayed in five of them. In the sixth, I felt comfortable against a 1200. I picked up a piece for a pawn in a simple tactic, then a few moves later I left a bishop en prise after a long think. :D That pretty much sums up my skill. I think of myself as a well-read novice. On a superficial level, I know about the important strategic themes, the different openings and their flavor, etc. But I'm not very experienced, and haven't actually put much of it into practice.

    Six years ago I idled away some time on CTS and maxed at 1370. However, I was playing for points. When I play at CTS now I go for accuracy and I can't do much above 1220. My brain is in glue; I need more than three seconds. Humbling, but it will come.

    Right now my goal is just to improve at chess tactical puzzles. I'm 90% tactics, 10% everything else. I'm not even playing again yet. I do enjoy mapping out my future repertoire (which changes weekly!), but I've promised myself not to practice it for a long while. Today's whim is to scrap most of it for a long while and just wing it with the Center Game and the Scandi.

    Thank you for taking the time to help steer me in the right direction. I look forward to being apart of this chess blogging community.

  5. 90% tactics is too much. Keep tactics down to 50%. Avoiding adverse tactics, e.g. by defending your pieces, is just as important - as is getting a position where the tactics work for you. Do not bother with opening books. Good books: Logical Chess Move by Move, Winning Chess Strategy for Kids, Comprehensive Chess Course Vol II (according to Heisman), A World Champion's Guide to Chess (great reviews, and you seem to like puzzles), Right Way to Play Chess (big following). Less is more though. Learning and forgeting does not help.

    1. No, no, I have books like that. I've done the novice/beginner "How do I play?" thing. I'm not trying to learn how to drive a car for the first time. I have an old beat-up, smoke-belching car that can at least get me on to the road. The problem is the wheels are extremely wobbly. I'm careening all over and often barrel into trees.

      I tried to avoid tactics before by hiding behind Nf3, g3, Bg2, O-O. I know now I can't avoid tactics. I don't love puzzles particularly; but tactical training involves puzzles. I can't follow a plan or read a board. Tactics (or at least Board Vision) is the roadway of chess that I need to convert from gravel to asphalt, or at least rough concrete. :)

      But so many people adore Logical Chess, and my local library has a copy. I've browsed it and it seemed great. I will read it for my own enjoyment, and out of respect for the wonderful help you've given me.

      I do have to figure out when I'm going to ease off the tactical training, and incorporate the rest more methodically. I should be done with Chessimo's first module in a few weeks. That might be as good a time as any. We'll see.

  6. If you completely master the novice/beginner stuff you should be about 1800, not about 1200. I believe the Albert book claims that you will be 1800 if you completely master its contents. I have not seen the book, and I expect that there is be a bit of sales hype there, but I do not believe that you need anything more advanced. You may need other similar books for extra examples, different explanations etc., but that is another matter.

    Advanced books will not make you an advanced player. Fully mastering basic books will. Being widely read does not help. Being narrowly read in the right things, and mastering them fully does. Imagine that you have to pass an exam with a 99% pass rate, and resit it every week. The beginner/novice books contain all the most important stuff, and have the most detailed explanations. If you read more complicated books before you have mastered the basics, you will just confuse yourself.

    Reading a book once does not do the trick, and reading just one book is often not enough. I suggest the one volume introductions, because then you do not have to manage your time. The green Coakley is a good book, because it shows you that the devil is always in the detail. Strategy is impossible without tactics. Put a sheet of paper over his examples and guess the moves. Keep going through the book until you get most of the moves right. If you cannot do that the book is too hard. (Notoriously, Coakley’s books are too hard for most kids.)

    For tactics, Bain is a good start. Keep working through it until you can solve all the problems very quickly, and repeat as often as you need to maintain the skill.

    Shredder Classic is good. Set it to 1200 and play lots of games against it. You need to learn to make practical use of what you have learned.

  7. As you may have seen on my blog, I have some fairly strong opinions about the (positive) value of (proper) opening study for improving Class-level players. Regardless, it's always good to see fellow players doing serious thinking about these issues and coming to their own conclusions as to what's valuable for their own training programs.

    Rather than offer lots of advice on openings etc., I'll share a couple hopefully useful observations based on your post.

    Gambits: the point of playing one is to make your opponent stressed, not you. The fundamental idea is that your advantages in time and piece development will cause your opponent to squirm and be under significant pressure at least into the early middlegame (ideally longer). I like your choices as White for that. Stretching your mental envelope is a good thing, if you can successfully embrace the necessary swashbuckling attitude. For those days where you're not mentally prepared for whatever reason to play like that, have a non-gambit opening choice available. As Black, gambits are much harder to pull off, with the lack of the move advantage.

    Reversed opening structures: I would be careful with this one, one size does not fit all. Notoriously, playing 1. f4 is not necessarily similar to playing the Dutch Defense as Black. Usually there are concrete reasons for why a reversed opening can miss the intended mark. In the above example, there is nothing forcing Black to play 1...d5 in response and other moves can take the sting out of White's setup. From my own experience, playing 1. c4 (the English) is not really anything like playing the Sicilian Defense, with only a few lines leading to a reversed-type position.

    1. Yes, it's my understanding that to some extent you share Robert Pearson's opinion - that moderate and directed opening prep is good. I suppose I have some work to do to form a completely informed opinion, but I agree. The "don't study openings until x" seems to me to be overkill in the opposite direction, designed to combat the amateur's typical over-emphasis on openings. But I'm an adult, and I can succeed in not being stupid. I want to present myself with familiar positions, learn to handle their strengths and weaknesses, and grow with them.

      But I'm not studying openings yet, and I've dropped the gambits for now. I'm just thinking about and preparing a repertoire right now. And it changes every week, so it's a darn good thing I'm not studying yet! The Bird is probably a couple years away; it's too pawn-y for me right now. But I might go with the Grand Prix, the Austrian Attack, etc., for transpositional purposes, and to ease a possible cutover to the Bird. Of course, by then I might have no interest in the Bird, but that's the way it goes.

      What surprised me is that you play the CK and Slav, yet you rate yourself an F in endgames! I had assumed these openings, especially the CK, were designed for endgamers! Reach an equal endgame and technique 'em. Once you get that F to a C, I foresee good things for you.

    2. Heh, in terms of the self-rated "performance inventory" you reference, by now I may be all the way up to a D- on endgames, but it's still obviously a weak point. However, as a general consideration, when you play opponents at the Class level with a similar (lack of) endgame knowledge, it doesn't matter so much.

      Re: endgames and my opening choices, the Caro-Kann and Slav normally result in solid positions, so if a balanced endgame is reached, it's hard to screw up (although I've done it). Conversely, if the endgame is significantly unbalanced due to material or positional factors, then it's not so hard to win.

      Overall, at the Class level, wins are much more likely to result from opponents' errors than ground out by superior technique. As I move up the ladder, endgame study will become more important; as a result, it's towards the top of my medium-term to do list.

      You may want to consider posting some of your thinking and criteria involved with openings selection and any "test drive" games you play with various openings. I'm sure that would help benefit players in similar situations and would also be of general interest.