Friday, September 14, 2007

Repeating the obvious (?)

Blue Devil gave us his abridged version of his chess thinking process:

1. Find candidate moves that generate threats and piece activity.
2. Play the candidate with the best consequences.

I commented the following:
This is what I used to do. But it is too meagre for complex positions.
Actually I hated myself for stressing the obvious again.

Reply from Blue Devil:
Tempo: what aspects of a complex (middlegame) position doesn't it cover?

At first I thought he was joking. But after a few hours I asked myself "what if he's not?". Is there a chance, after talking for months about exactly this very same subject, that there is someone out there who doesn't see the obvious? As Sherlock Holmes put it, "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?". After all, this is not the kind of joke Blue usually makes.

Since I never want to miss a chance to repeat myself I assume that, how improbable it might seem, I haven't explained myself clear:) Let's summarize what looks obvious to me:

I was always looking to a chessposition according to the the following two rules:

1. Find candidate moves that generate threats and piece activity.
2. Play the candidate with the best consequences.

Not the alpha nor the omega. . .
But I found that in the kind of position that causes me trouble this approach was to meagre. The search for candidate moves stimulates the trial and error approach. Since it starts at the trunk of the tree of analysis. So I started to look for another approach. Was it possible to start at the end of the line and work backwards to the trunk? No, that proved to be not possible, since the characteristics of the end of the line are not yet visible on the board.

But the bhèta.
It is possible though to start at the second stage and work your way back to the first stage or the trunk? The second stage is about identifying the invasion squares.
I can imagine that this sounds arbitrary for the casual reader.
It is not.
If you have a closer look at it you will see that it is inevitable. The highest strategy of the middle game is piece activity, but what is the ultimate goal of piece activity? Invasion!
After investigating complex middlegame positions for months I found this to be true in 100% of the cases. Identifying the invasion squares is the method nec plus ultra to find the second stage of the tree of analysis in any complex middlegame position. See for yourself!

From the bhèta back to the alpha.
What I'm trying to find now are the laws that govern the transition from stage one to stage two. This is where the men are separated from the boys. Even grandmasters fail in 25% of the cases, according to Rybka. This phase is still under construction.

A lot of my fellow bloggers seem to have no higher goal than to avoid the dropping of pieces. That is logical given their rating. But above 1700 the dropping of pieces is very rare and other methods have to be applied. The second stage of the tree of analysis lies already 5 to 10 ply in the future. Mastering the first two stages of the tree would mean a quantum leap ahead. For most of us, I suppose.


  1. Hello,

    I find your thoughts very interesting, but I don't know if I understand correctly what you are trying to reach by this process or how it is working. Are you trying to truncate the tree of variations? So that you have less candidate moves to look at? Or that there are not so many possible answers to the candidate moves? Or both?

    Could this not be reached by a refinement of the understanding of piece activity on one hand and increasing the number of plys ahead plus visualisation? Like: "This move looks active, but in fact isn't, because of ...", "now I can clearly see what the board would look like after these three moves, this looks ok, but wait, he has this continuation..." Maybe there is progress along these lines.

    I don't know if complex middlegame positions are always about invasion. How about positions where it is "just" about winning a piece, but in a very complicated manner? Isn't invasion more of a long term (strategic) goal?
    Undoubtedly, in those problems you look at, it's different. But are they representative in this aspect, say, of your own games? Certainly they are not of mine, but then, there aren't that many complex positions in mine (which I would recognize).

    I'm really drowning in the variations of these book-examples you give, they are over my head. In a game I would start looking at variations, and, as time allows, play the best I have found so far (trial and error). I would be happy, if there was another way, but I'm afraid there is none. I dont think the approach you first mention is too meagre, but my abilities to go through with it. Shortcuts would be nice, of course.

    Sorry for the long posting.

    kind regards,

  2. svensp,

    what I try to reach

    If I look at this last position, I see at least 9 moves that seem to deserve investigation. With my method 6 branches can be pruned.

    Could this not be reached by a refinement of the understanding of piece activity on one hand and increasing the number of plys ahead plus visualisation?

    Why to investigate 6 lines that lead to nowhere? Invasion is the refinement of piece activity. If you can't do these problems WITH a board, no visualisation skill is going to help you when you try it WITHOUT a board.

    I don't know if complex middlegame positions are always about invasion. How about positions where it is "just" about winning a piece, but in a very complicated manner?

    Give me an example that doesn't need an invasion square and I will agree with you. Until now it proved to be right in 100% of the cases. Why bother about that 0.0006%? But please don't belief me. Just see for yourself.

    But are they representative in this aspect, say, of your own games?

    Such positions play a role in about 70% of my games. Roughly estimated.

  3. Some FENs, which may illustrate either my point or some different understanding of when a position is really complex. I can only say: To me those were complex and I didnt see invasion squares as key here (admittedly, almost none of them is about winning a piece either, just complex middlegame):

    2k4r/ppp1rpp1/7p/3p1n2/3P4/bPPR1N2/1nKBNPPP/4R3 w - - 2 21

    it continued (to illustrate complexity (or what I regard as such), even though with wrong moves, it was fun):

    21.g4 Nxd3 22.Kxd3 Nd6 23.Ra1 c5 24.Nf4 c4+ 25.Kc2 Nb5 26.Nxd5 cxb3+ 27.Kxb3 Re2 28.Be3 Rb2+ 29.Kc4 f6 30.Kd3 Kd7 31.c4 Rb3+ 32.Kc2 Nxd4+ 33.Nxd4 Rb2+ 34.Kc3 b5 35.Rxa3

    r2q1rk1/1b2bppp/2n2n2/ppppp3/1P2P3/P1PP1N1P/1BBQ1PP1/RN2R1K1 w - - 0 16

    2rqk3/pp2ppbr/b2p2p1/8/2nNPN2/4B1Q1/PPP2RPP/4R2K w - - 6 28

    (invasion squares play a role here, but they are not the key (I dont know if there is any "key" here at all, just a lot of variations and threats from both sides), that's of course another topic: what if there is no "solution" like "mate in 5" or "white to move and win").

    r2qrk2/2p3pp/p1Pp1bb1/3Q1N2/2P5/2B1pP2/PP4PP/3RR1K1 b - - 0 22

    So I dont know about 0,0006%. Different people get different middle games, this may also contribute here.

    kind regards,

  4. Svensp,
    none of these positions has a clearcut solution. Of course I'm not talking of these. In these kind of positions your goal should be:
    1. Find candidate moves that generate threats and piece activity.
    2. Play the candidate with the best consequences.

    But still it should be of help when you know what the goal of piece activity is so your pieces can converge to the same invasion squares in stead of all pointing in another direction.

  5. Well, I'm sure focused on avoiding dropping pieces. But I still don't quite understand what you are saying.

    To summarize your thesis:
    TT--for all complex middlegame positions the main goal is to identify invasion squares.

    My hunch is that TT (Tempo's thesis) just describes one of many possibly useful long-term middlegame plans. But for now let's assume TT is right.

    Finding an invasion square is to find a particular threat that you can work toward. This fits in fine with the abridged process.

    Perhaps you are too focused on the fact that it says 'find candidate moves' as if that means you just look at moves (the trunk of the tree) and then see what they can do. That would indeed be a superficial way to play chess. But that's not really what it's saying (or at least not what I mean).

    In step 1, for instance, you can have big threats in mind (mate, invasion square, knight fork), and then look for moves to achieve them. That seems to be all you are talking about doing.

    Also, step two leaves plenty of room: one of the consequences that may be best is that I get a higher probability of having an invasion square. So I like those consequences so I'll play that candidate move.

    The abridged thought process is like Creationism. Unfalsifiable.

  6. Note I look at piece coordination as a special case of piece activity (this is likely b/c active pieces are the opposite of passive pieces, and coordinated pieces are anything but passive). Perhaps that might be one little semantic thing that makes you not understand why my system is so brilliant.

    :) Hee hee.

    I could say that and get away with it if my rating were higher than the temperature.

  7. Blue,
    don't you think Creationism is a too meagre explanation for the Universe?

    I never realized that the inevitability of my point is so hard too see. Sorry for my nagging, I didn't want to improve the world.
    To all: feel free to think what you like.

  8. So you stand by TT, and don't think finding an invasion square is reducible to some mixture of threats and piece activity.

    But you don't want to say why.

    If it's a topic you brought up, it's not nagging. In English it's called "finishing a discussion."


  9. That was a joke, the last line, but it was in the tradition of Tempo's unintentionally offensive lines with a grain of truth. I hope I haven't offended. (the last line being a DK-ism, which shows I must really be going off the deep end, acting like Tempo and DK in one thread!).

  10. "none of these positions has a clearcut solution. Of course I'm not talking of these."

    Ok, but how do you know in a game if it is one with a clearcut solution... :)

    If it is a complex situation I bet most of the time it cant be said before the solution is known. But if it is known there is of course no more need to investigate the position. Unless there is a gut feeling that something must be in there, but not what. Maybe I lack that feeling.

    Anyway, thanks a lot for your answers and explanations, you are right: if it cant be done with the help of the board, it cant be done without. And if your method does indeed cut 6 out of 9 possibilities it is obviously pretty helpful for you.

    kind regards,

  11. Blue,
    I was looking for a way to encourage you guys to have a second look at it since it is important. In doing so I tried to find a way that would not trigger any ego by saying nothing offensive or provocative.. But it seems that violating my character doesn't work either. Anyway, shoot after you understand what I mean in stead of before. In that way it is advantageous for me too.:)

  12. Tempo: I'm very confused. I asked some very simple questions in my second to last comment. I have followed your posts about this the past month. I just don't see why finding an invasion square, even if it is extremely important for transitioning toward the end of a game, isn't a special case of threats plus activity (as I've previously defined activity).

    Dialogue with my chess coach....

    Me: I had the plan of attacking f7, invading his territory to destroy his pawn structure, expose his king, or he will lose material.

    Coach: Ah, the invasion square. That is the most important idea in complex middlegame positions. That is what they all aim toward.

    Me: I thought we aimed toward mate.

    Coach: yes, but in games where people aren't dropping pieces left and right, you must pry open his position, penetrate his camp, to give yourself a chance of this.

    Me: so, you mean it is a particularly powerful threat.

    Coach: NO! It is not a threat! I thought I was clear, that this would be obvious but your mind is enfeebled as you are still trapped in the phase of thinking of material gains.

    Me: OK, so the goal is mate.

    Coach: Yes.

    Me: So, isn't that a threat?

    Coach: Yes, but the invasion square is a particularly potent way to get into the enemy position. It is the only way you will get mate against a worthy opponent, so it becomes an almost surrogate for mate.

    Me: OK, so because play between good players gets to very complex middlegames since they aren't dropping pieces, you need to find some way to force weaknesses in his position.

    Coach: Yes, if you must.

    Me: What types of weaknesses?

    Coach: The types that result from finding invasion squares!

    Me: so, what weaknesses are these? Decreased material, activity, pawn structure, king safety?

    Coach: you have the narrow perspective of a patzer!

    Me: You're fired.

  13. Blue,
    LOL. I probably hadn't read your comments quite well so I hadn't noticed the urge in your questions. I will try to address them in the next post.

  14. hard to comment from affar or not following along in detail preoccupied with my own study, but only to note that i continue to visit your blog, and appreciated.