Sunday, March 26, 2017


How to proceed
I'm not quite sure how to proceed. So I will just give it a go, and see what I end up with. I'm a bit cautious with reacting to your comments, as you might have noticed, since usually it is easy to start a whole new series of posts on any subject you come up with. I want to stay focused on the faint thoughts that circle from high above through my head, like barely visible vultures. I don't want to let these thoughts fade away into oblivion by the bombardments of thoughts that reacting to all your comments would trigger.

Identify the sitting duck
Starting to look for the piece that is the most restricted in mobility is a good start. There is no doubt about that. It prunes a whole lot of branches from the tree of analysis from the very start. Sofar, I have never seen it failing. In the sense that I never have pruned a branch too many.

Some positions seem not suited
Some positions seem to be more suited for the plf-system than others. I called that the holes in my bucket. Every time I plugged a hole, I discovered that I just didn't apply the system rigorously enough. I just lacked the precision and the ausdauer needed. But never the system let me down. The final solution could always be explained within the boundaries of the system. Even the other motifs like promotion and assault can be described in terms of the plf-system.

While plugging along the holes in my bucket, I couldn't help to get the impression that the amount of  scenario's is fairly limited. Maybe the amount is vast, I don't know, but it seems to be finite.

Three types of moves
There are three types of immobility.
  • Space
  • Time
  • Function
The identification of the least mobile piece gives you a target to aim for. Yet it might not be clear from the beginning how to proceed in that direction. I found that asking the question "which piece is the weakest defender" could help to identify a subtarget to aim for.
But then again, sometimes it is easy, and sometimes it is difficult to envision the next move.

A move will always be designed to exploit the immobility of a target or defender. Since there are three times of immobility, it might go without saying that there are three categories of moves. Each type of immobility will have its own methods of exploitation. Its own set of scenario's.

I want to investigate this further, and I'm going to use a position that is provided by Takchess. So that I answer to at least one comment. Since I don't know what I will find beforehand, I cannot predict whether the position will be suitable for this investigation or not. Anyway, we will find out.

Diagram  1. White to move

r1b2rk1/qp3ppp/p3p1n1/8/4N3/1B5R/PP2Q1PP/3R3K w - - 10 23

Sitting duck: black king
Surplus attackers: 3. One to break the fortress by a sac, remaining 2 pieces to deliver mate
Potential king flee: Re8, Kf8, Ke7
We are going for mate, and are happy too when the opponent must give up a piece to prevent it.

points of pressure h7, f7
Of these, h7 is already under attack, and defended only once, while f7 is not under attack yet, and has two defenders.

Diagram 2. PoPs = yellow. White to move


lines of attack
  • h-file
  • f-file
  • d-file
  • diagonal b3-g8
The h-file is the most important line of attack.

Diagram 3. LoAs. White to move
The main defenders of the black fortress are f7,g7,h7, Ng6
The black knight looks a bit clumsy. Yet it defends h8 and the g-file. We might be able to play around the knight.

Weakest defender
After the initial inventory-taking of the main characteristics of the position, we need to identify the secondary target by asking the second question: which defender is the most vulnerable? (the first question, who is the sitting duck?, revealed already the main target, and the main intention of our attack (mate))

g7 is definitely the most important defender. It has two functions:
  • In order to keep the h-file closed, h6 must be played at some time. g7 becomes then the defender  of the h6 pawn, and is hence responsible for keeping the h-file closed.
  • g7 defends f6 against the white knight.
g7 might become overloaded. We should proceed our play with exploiting the immobility of the secondary target g7.

List of candidate moves
So far, we didn't need to calculate anything yet. The important characteristics of the positions are revealed to us by guiding our attention with the plf-method. The question now is: how to exploit the overloaded g7 pawn? Two moves spring to mind:
  • Qh5
  • Nf6+
Both moves are winning. There is another winning move, the immediate Rxh7. But that is not so strong as the previous moves, since it sacs a whole rook.

Qh5 threatens mate in 1, and forces black to push the h-pawn. Thus saddling the g7 pawn with its second function. The g-pawn is now overloaded, and can be exploited.
1.Qh5 h6
2.Nf6+ gxf6
3.Qxh6 Re8
Here went white astray.

What kind of move is Qh5?
 The Q is placed on the line of attack against the point of pressure h7. It is based on the fact that there are not enough pieces to perform the function of defending h7. In stead it must flea into the protection of a low valued defender g7. Adding an extra function to the pawn, and in doing so overloading the g pawn. It is the motif of encirclement.

In a way, the h-pawn seems to suffer from another type of immobility. It is not immobile by itself, but it is cut off from its potential defenders. Blockading pieces prevent defenders to protect h7. The defenders lack space. The queen move exploits that.

Diagram 4. White to move.

r1b1r1k1/qp3p2/p3ppnQ/8/8/1B5R/PP4PP/3R3K w - - 1 26

In order to proceed, we must draw the lines of attack again.

Diagram 5. White to move

There are two points of pressure (the yellow dots). Which one is the most favorable?
To conquer h7, the black knight must be eliminated. By Bb3 or by Rd1.
4.Bc2 and the black knight can be shielded from the bishop by f5.
Rd1-d3-g3-g6 is too slow.
So white must focus on the point of pressure f7.
f7 is attacked once (indirectly, by the white bishop), and defended once. How can white build up the pressure? White can triple the pressure on f7 by Qh7+ and bringing a rook to the f-file.

4.Qh7+ Kf8

Now blacks position starts to crumble. f6 is going to fall, f7 is going to be pinned which gives black the impossible task to defend both its knight and f7, and keep the f-file closed.

Yusupov failed to find the only winning move here (4.Qh7+) and played 4.Rg3. Which is giving away the initiative, and his best hopes now are to draw with an eternal check. Which happened in the game.There is no justification for 4.Rg3 in the plf-system.

From a description of Aox about a certain problem a few posts back, I concluded that his trial and error is much faster than mine, and that that is the reason why he is higher rated than me.

From the fact that a grandmaster could not find the win here, I conclude that he didn't rely on a kind of chess logic like I do here with the plf-system. That might imply that he uses a faster trial and error than I do, and not a superior chess logic. That is why he is grandmaster and I'm not.

When I become better in chess by applying the plf-system, will I be better in the same way as a grandmaster is?

When I wrote about improving in Troyis, I identified two ways to become better.
The first way is by frantically play the game without further thinking about it.
The second way is by inventing a new strategy for Troyis.

The first way was the inspiration for the salt mines. It didn't work out as hoped for, and in the biographies of grandmasters we can't find any indication of grandmasters that improved this way.
The second method, inventing a new strategy, is in fact my inspiration to develop a new strategy for tactics.

The rationale of my idea, is that it takes about 50 lessons of an hour to learn a complex motor skill like driving a car. Somehow, the unconscious needs very little time to work its magic behind the scenes.

Has a grandmaster indeed internalized chess logic at a young age, which he has forgotten he did? In that case the failure of Yusupov is due to the internalization of incomplete chess logic. In that case, there is no difference between my method for adults to become better at chess, and the youngsters method of chess improvement. Or are there two different methods, one for young people and one for adult people? In which case we have still not the slightest clue how young prodigies do it.

The plf-experiment will give us definite clues on this subject.

To be continued after the break.


  1. No sooner was I seeing this problem than my mind was noticing 1.Nf6+ followed by 2.Qh5, and thinking that that was the solution until I saw 2...Re8, then realized that if this were the answer it must be a deep solution, and not the type you'd find on chesstempo.

    Then I looked at 1.Rxh7, since that is the quickest attack, but realized it gives away too much material.

    Then I saw that on 1.Qh5 Re8, 2.Qhx7+ Kf8, 3.Rg3 does appear mating, and figured I'd probably just play 1.Qh5 in a game, but spaced out a bit because even then realized that after 1...h6, 2.Rg3 in this line would take a lot of work to calculate and might not be best, so I left this puzzle with a lot of work to still do.

    I really liked this problem, and the answer is quite long. I didn't think of the move 4.Rf3 (and really didn't try to think of anything here). If ...f5, 5.Rxf5 is nice to spot the mate on f7. However, it probably deserves a longer answer than this for completeness and verification's sake. So, 4.Rf3 (...Ke7, 5.Qg7 with Qf6+) b5, 5.Rxf6 Rb8, 6.QxNf6 (also a tough move to see from the starting position unless you are used to these tactics up a file)...Rc7 (material is still level!) 7.Qh6+ Ke7, 8.Qg5, and here the discovery 9.Rh6+ should lead to mate.

    I did spend quite a bit of time looking at lines with Ng5, but also realized that once you throw in some kind of rook sac, the knight is also more out of place here than in the starting position.

    I think that Yusupov was constrained by the time-limits of the game. It's very likely he looked deeper than I did in that winning line, but if he missed one thing, one variation in there, the Rg3 choice might have appealed to him more as a practical try where he wasn't risking the entire game on this one move. Also, it's hard not to notice that Black's position is pretty bad whether finding a mate here or not.

  2. I think this position is too complex for Plf. Most likely at a highest level of Ctart. Aagaard suggests a preperation move before the fireworks. Do you see what that might be ?

  3. I would like to share my opinion on that. I have NO PROBLEMS with finding the variation: 1.Qh5 h6 2.Nf6+ gxf6 3.Qxh6 Re8 4.Qh7+ Kf8, but after that I could not find the winning continuation. And that's the reason I would REJECT playing 2.Nf6+ as it loses a piece for "unsound attack".

    And after 4.Rf3 I would play 4...f5 and 5...b5. I still do not understand why white's attack is sound and how can white finish off Black's King.

    By the way: I think the majority of amateur players (let's say even 90%) has problems with so called 'quiet moves' (intermezzo is another name for the same term). Especially I have no idea when to play it and how to recognize wherever it works or not.

    PS. Do not get me wrong - your articles are simply superb, but most often they are over my head. Anyway I enjoy reading these as much as the discussion that goes with that (as a form of comments).

  4. The difference between stronger and weaker tactician is the speed. Speed of calculation, speed of memorising, speed of solving tactical puzzles...

    No tactician is using trial and error, we all look at the board and try to find "the idea" of the puzzle, try to understand the dependencies, try to see weaknesses. We all start concrete calculation only after getting some guess of whats going on.
    While we calculate we learn more about the puzzle, find new ideas and reorganise our calculations.

    we find first ideas by doing a type of partial/incomplete calculation and by pattern recognition along these incomplete calculations. As more pattern we know, as better and faster we recall such pattern as faster we can start with the final calculations.

    Grandmaster did start their chess in their youth and their saltmines where many hours of chess every single day. Most Grandmasters are known for their exceptional chess memory. I measured the effect of solving 4000+ easy mate problems at CT. After about 8 days not seeing one of these problems i was not faster by solving the exact same puzzle a second time again. Someone with a better memory would have been able to be faster after 8 or even more days. The pattern mate is not easy enough to generate new templates in my brain because i dont see enough related pattern within my memory/forgetting timeframe. I think / guess that kids can learn better with less repetition in a given timeframe.

    since almost 2 years i am using now a system related to your plf system ( chuzhakins system ) and it seem that i did not get any faster ( or better ) by that. I am just doing now conciously what i did unconciously before anyway.

    I think now, that improvement in chess is mostly about learning speed and duration of learned information ( = forgetting speed ). While you try to solve a tactical puzzle we are learning the puzzle. As soon we have learned it ( = gained the necessary understanding ), its solved.

  5. @ Temposchlucker:

    An extremely minor quibble regarding language: the word "flee" (verb) should be used instead of "flea" (noun) - re: "Potential king flea: Re8, Kf8, Ke7". My apology in advance for mentioning this trivial point.

    I have some thoughts vis-a-vis your observation regarding the 50 hours of driver training but that is not sufficiently on-topic at this point; perhaps at some future point you might address it. The question is this:

    There is an assumption that youngsters and adults learn differently. Yet in the case of learning to drive a car, youngsters don't learn this at all; only adults (well, at least much older "children") learn to master driving skills. In this specific case, the adults acquire the requisite (unconscious) skills in approximately 50 hours, not the proverbial 10,000 hours.


    1. is there any racecar driver who did not learn driving as a child?

      " When Schumacher was four, his father modified his pedal kart by adding a small motorcycle engine. When Schumacher crashed it into a lamp post in Kerpen, his parents took him to the karting track at Kerpen-Horrem, where he became the youngest member of the karting club." see:

    2. @ Aox:

      Honestly, I don't know the answer to your question. Perhaps it is true that every racecar driver learned to drive (something) before the age of five, like Tiger Woods learned to play golf, and Mozart learned to compose). My question was not concerned with the elites (racecar drivers, or, by extension, golfers, musical maestros or chess grandmasters). It is about we ordinary mortals (the vast majority of drivers) who (somehow) manage to "master" (?) driving skills sufficiently to the point of driving being controlled (mostly) by the subconscious.

      Can we agree to disagree agreeably at this point, and save an in-depth discussion until later? I only wanted to record the question for future reference, to avoid taking the discussion away from Tempo's already stated direction of investigation.

  6. just had a chat with an american NM. Up to expert level he did mainly internet-play and reading chessbooks. He did read a book for 2-6 months. From Expert to Master he was doing endgame and mastergame studies for at least 2.5 hours a day

  7. Tempo , Well done, thanks for giving this a shot, it is a tough one. Hopefully, you will find the exercise worthwhile.

    What interests me is a correct move , B-c2, can be found without calculation by applying a global principle of activating all attackers before launching the attack.

    I would say in our defense that this complex position is not your typical position found on CTS with one set answer and totally forcing lines. The PLF found good targets and the LOF needed to shift to the b1-g7 diagonal. In most cases, for a tactical puzzle, the attackers are all in position and ready to strike. I think we might of come up with a correct answer if the bishop was sitting on c2 to begin with.

    This game is included in Aagaard’s Attacking Manual 1 in the chapter on Including all your Pieces in the Attack which entitled Bring All your Toys to the Nursery Party. He annotates this game beautifully in over 8 pages; and those that like this sort of thing should get the book !!

    In looking at the position he mentions that this is one of those games that is near impossible to calculate to the end but we can conclude that an attack is likely to be successful if it directed toward h7 while keeping an eye on f7. The Q can join the attack on h5, the rook on H3 is well placed. The e4 N is well placed to join in on an attack from either f6 or g6 . The d1 rook in time can join in but too slow . “This only leaves us with the bishop which is not attacking h7 but can quickly attack it from c2. It is tempting to say that one should play B-c3 without taking any other moves into consideration. “

    He still feels that at move 24 he could still play the Bc2 move even though he says there is a win on Nf6 check that the Bc2 moves would simplify things.
    Earlier in the chapter he states : I wish to convey that in a great majority of cases , we need to bring more pieces into the attack…..having lectured on and investigated this subject for almost a decade, I have come to the realization that it is exactly piece activity that most mistakes are committed in highly dynamic positions. (Us not reaching the correct plan , we are in good company. This game played under time pressure with elite players. As JA likes to say: Chess is simple but it not easy! )
    There are other technical things that could have been played for the win but are beyond the scope of my comments here.
    The following link shows a win with the Bishop on c2.

  8. Also same players. Yusupov Immortal Game.