Saturday, April 4, 2015

Calculating Forcing Moves: Defensive Idea 5 - Counterattack to WIN!

We have finally come to the most exciting part of this defensive calculation blog series -- counterattacking to WIN!  The ideas in the next few posts generate the kinds of moves that annotators like to give one or two exclamation marks.  You'll see that the thought process and logic for finding these double-exclam moves is rather simple.  The hard part is calculating the actual lines.  That is because what you're doing is dangerous.   You're leaving one threat and creating a brand new one -- threats will be spreading across the board like wildfire.  There can be many branches of variations that run several ply deep.

Can you manage the chaos?  Will you get overwhelmed by the variations and the greater risks of miscalculating your idea (possibly even checkmate), and simply avoid counterattacking completely?  It is very simple to defend or move a piece that is being attacked.  The variations are short and easy to calculate, and there's often no major change to the status quo on the board.  There is little to no risk in making those moves, and they rarely ever get (or deserve) an exclam.

Counterattacking can be very upsetting to your opponent.  They expect you to defend or move, but you just ignore their threat!?  If you dare to venture into the wild lands of counterattacks, searching for those single or double exclam moves, then read on!

I titled this post "counterattack to WIN", however, when you begin calculating a counterattack you don't know where it is going to lead.  You will never know at the outset that you are calculating to win.  That said, you will always have some sort of expectation.  An easy example is when you have a big material advantage but your opponent has a strong threat.  If you can stop the threat, you'll win easily.  You might begin to falter if you fail to find a move that allows you to keep your material, and you just don't search (well enough) for defensive moves, or fail to find the forced draw.  The opposite situation can also be unsettling -- you're already down material and think you're losing, and are desperately seeking a way to survive the threat.  You may unintentionally exclude certain candidate moves that actually win.  The outcome of your calculations don't meet your expectations.

To avoid losing your way inside the counterattack variations, set your expectations aside going into the variations and always search for all three possible outcomes:  win, draw, or defend.

In the next few positions we'll look at counterattacks that result in a win.  In subsequent posts we'll look at counterattacking to draw, and counterattacking to defend.

As noted in previous posts every counterattack begins with an equal or greater threat.  That is the idea we'll use to generate specific candidate moves.  So let's get started.

White just played Rb7.  Black to move.

3r1k1r/pR4p1/6Q1/4pp2/P1Pq3P/8/5PB1/6K1 b - - 0 0

Let's evaluate the position and then find all of white's threats.  Black is up the exchange, and has a better pawn structure (three islands compared to white's four).  Black is playing for a win and has the simple plan of making equal piece trades, picking off white's weak pawns and promoting to a new queen, while white is playing for a draw and has OK drawing chances.

White is threatening mate four ways: 2. Qf7#; 2. Qxg7+ Ke8 (only) and then 3. Re7#, 3 Qf7#, or 3. Qe7#  He is also threatening black's pawns on f5 and a7.  The mate in one is white's greatest threat, so let's now look at the five defensive ideas to see what candidate moves we find, calculate each one, and then pick the best move.

Idea 1 (capture, pin, deflect):  none.
Idea 2 (block):  1...Qd7 (fails to 2. Rxd7).  1...Rd7
Idea 3 (move):  there are no escape squares for the king, and black can't create any.
Idea 4 (defend):  black can defend the g7 pawn with 1...Rg8, but that does nothing to defend against the mate in 1 so we won't calculate it.
Idea 5 (counterattack with an equal or greater threat to WIN, DRAW, DEFEND):  We begin by looking for equal or greater threats (EGT), which give us time to control the board for a while.  Look back at the EGT chart, and you will see that in response to mate in one threat there are zero equal threats, and greater threats must begin with a check, of which black has three:  1...Qxf2+, 1...Qd1+, and 1...Qa1+.

So that gives us four candidate moves to calculate (one block, and three checks): 1...Rd7, 1...Qxf2+, 1...Qd1+, and 1...Qa1+.  Let's start with the most forcing move, the capture-check 1...Qxf2+:

White has three replies -- one capture and two moves.  The most forcing (and likely refutation) is the capture 2. Kxf2.  Now white is still threatening mate, and black is running out of defensive options.  He must check!  So 2...Rd2+ 3. Kg1 Rd1+ 4. Bf1 (not 4. Kh2 because that allows black to bring in another piece with 4...Rxh4+), and now black is completely lost.  He can only block the checkmate with the now useless block (because it is undefended) 4...Rd7, or he can again try his last check 4...Rxf1+, but after 5. Kxf1 black has zero options to defend against checkmate.

This path was worth exploring to search for drawing resources.  In this position stalemate would not be an option because of black's multiple pawn moves, so repetition would be the only option.  If there was a forced draw, we would have made that variation our King of The Hill (KOTH), and then continued our search to look something better than a draw (wins).  But unfortunately for black, he can't force a draw with this capture check, and 1...Qxf2+ is easily refuted.  Next up is 1...Qd1+:

White has two replies -- one block, and one move: 2. Bf1 and 2. Kh2.  Neither are particularly forcing, so let's start with the block 2. Bf1.

Now white is still threatening mate and black's next move must address that threat.  The first check has not given black any new defensive options (Idea 1 - capture/pin, Idea 2 - block the attackers, Idea 3 - move the king, or Idea 4 - defend f7), but using Idea 5 we do have new checks to look at 2. Qxf1+ and 2. Qg4+.  The capture check is the most forcing but the other check also attacks white's queen, which is threatening mate so let's check that one out:  2...Qg4+:

White has four replies -- one capture, one block, and two moves -- but since white's queen is hanging 3. Qxg4 is forced.  After the simple recapture 3...fxg4 black is now winning!  He was able to use the time gained from the check to force an equal trade, which since he was up the exchange this was his basic plan to win the game.

But white had another defensive reply that might be better for him.  So let's back up and look what happens after 1...Qd1+ 2. Kh2:

2...Rxh4+ and white has one block and one move.  If he moves with 3. Kg3 Rg4+ wins the queen for the rook.  If he blocks instead 3. Bh3 Rxh3 4. Kxh3 (forced; if 4. Kg2 Qh1#) Qh1+ 5. Kg3 (only) Rd3+ 6. f3 (only) Rxf3#.  So 2. Kh2 is certainly not better for black!!

That means our current KOTH variation is 1...Qd1+ 2. Bf1 Qg4+ 3. Qxg4 fxg4, which leaves black with a winning material advantage.

Before making the move we need to calculate black's other two defenses (1...Qa1+ and 1...Rd7) to see if they are better than our KOTH.  First is the more forcing 1...Qa1+:

Here white has two blocks (2. Rb1 and 2. Bf1) and one move (2. Kh2), but 2. Rb1 just loses a rook for no compensation, and 2. Kh2 allows black's rook to join the attack with check.  After 2. Bf1, white is still threatening mate, has only one check that just loses the queen, and now has no options to defend against the mate in one (the block 2...Rd7 fails because it is undefended).  So we can reject 1...Qa1+ since it is not better for black than our KOTH.  Next up is the block 1...Rd7:

Due to the activity of white's pieces, he has lots of good options here.  If black can make equal trades (rook for rook, or queen for queen) then he should win.  I think easiest and most convincing for white begins with the forcing capture-check 2. Qxf5+:

Black has zero captures, one block (which leads to mate in one after 3. Qxf7#), and three king moves (2...Kg8, 2...Ke8, and 2...Ke7).  Since there isn't a most forcing variation, let's start with moves we can easily "trim".  Now 2...Kg8 appears to be the worst for black because of 3. Rb8+ Rd8 (only) 4. Rxd8+ Qxd8 (only) 5. Bd5+ Qxd5 (only) 6. cxd5 wins for white.  That leaves us with the two king moves and I think we can also trim 2...Ke8 because if 1...Rd7 2. Qxf5+ Ke8:

3. Qe6+ wins one of black's rook, because again black cannot capture the queen and his one block (3...Re7) leads to another mate in one.  If the black king moves away from defending the e7 rook (with 3...Kg8), then white simply takes black's e7 rook, and if the black king tries to guard the rook (with 3...Kd8) then the simple skewer 4. Rb8+ picks up black's h8 rook.  So we can reject 2...Ke8.

Those two replies were easy to trim, and leaves us with 1...Rd7 2. Qxf5+ Ke7:

Here I like 3. Qf5+.  If possible white would like to push the black king to a light colored square so his bishop can join the fun, so checking with the queen on dark squares helps that while also controlling the d8 square.  Black cannot capture the queen and has no blocks.  The king has five moves.  Since none are particularly forcing, lets again trim the bad ones first.

Both king moves to the back rank fail to 4. Rb8+ -- if 3...Kf8 4. Rb8+ and since the queen is now covering d8 the block (4...Rd8) is simply not safe (two attackers to one defender), and if 3...Ke8 4. Rb8+ Rd8 and now the bishop gets to join with 5. Bc6+ Kf7 (only) 6. Rb7+, and now black has no captures, two blocks, and three moves.  Let's start by trimming:

if 6...Kg8 7. Qxg7#
if 6...Kf8 7. Qxg7#
if 6...Ke6 7. Qg6#
if 6...Rd7 7. Rxd7+ Qxd7 8. Bxd7 is winning for white, and
if 6...Qd7 7. Rxd7+ Rxd7 (forced, any king move leads to mate in 1) 8. Bxd7 is winning for white.

Was this line any better for black? Our KOTH is1...Qd1+ and then 2...Qg4+, leaves black up the exchange.  All other move one alternatives were worse for black because they were losing.  Not much of a comparison!

To recap, in this position we used our search for greater threats to discover defensive resources that allowed us to force an equal exchange of one of the pieces threatening mate.  Because black was already ahead material equal trades will result in a win for black...but he'll have to fight for it in the endgame.  If material was somehow equal in the beginning position than this post could have been called counterattacking to defend!

One of my favorite examples of counterattacking to WIN was included in an earlier post.  Counterattacking should not be something you only consider out of sheer desperation.  No, you should force yourself to use all five defensive ideas when dealing with a threat, and train yourself to use Idea 5 in particular and search for all equal or greater threats you can create.  You'll be finding some amazing resources once you do!

In the next blog we'll look at counterattacks using EGTs that allow the defender to force a draw by perpetual check or stalemate.


  1. I have a few suggestions to improve your actual post. For now there are three of them.


    You can use it in any place you wish - maybe above the "Let's look at the five defensive ideas to see what candidate moves we find, and then pick the best move" - instead your previous sentence (expand and use it or replace/correct this one: "White is threatening mate four ways: 2. Qf7#; 2. Qxg7+ Ke8 (only) and then 3. Re7#, 3 Qf7#, or 3. Qe7# He is also threatening black's pawns on f5 and a7.").

    There are two types of threats we are interested in. The first type are "check oriented" and the second ones "without check".

    A) with checks - 4 treats (with decreasing order of threats priority: the most important one at the top)
    1) 1.Qf7#
    2) 1.Qxg7+ Ke8 2.Qe7#
    3) 1.Rf7+ Ke8 2.Qe6# (if 1...Kg8 2.Qxg7#)
    4) 1.Qxf5+ Kg8 (if 1...Ke8 2.Qf7#) 2.Qf7+ Kh7 3.Qxg7

    B) without checks (gaining material or bringing the piece to attack)
    1) 1.Rxg7 (gaining a pawn)
    2) 1.Bd5 (interposing Q+R line and attacking a two diagonal squares arond the King [f7 and g8]

    As there are too many threats we cannot defend the position in a passive way. That's why it is important to look for the active moves ("with checks").


    The refutation/expanding of your defence (I mean - showing the way how it works)

    Idea 2 (block): 1...Qd7 (fails to 2. Rxd7 - as White obtains a decisive material advantage). And if 1...Rd7 then Black activates pieces and gain decisive material advantage due to open Black King; for example 1. ... Rd7 2. Rb8+ Rd8 3. Qxf5+ Ke7 4. Rb5 Qd1+ 5. Bf1 Rd6 6. Rxe5+ Kd8 (etc.,)


    An addition to the third way of counterattack (a short passage as a conclusion).

    ... and if we play 1...Qa1+ White simply plays 2.Bf1 and we have no way to exchange queens or stop mating threat. That's why we cannot play this move to keep the position won.

    It is all for now. However I do not know what does it mean "Idea 1 (capture, pin, deflect): none." and "Idea 3 (move): there are no escape squares for the king, and black can't create any.". If you could explain your way of thinking (analysing) I would be able to understand what you meant. Thanks in advance for your help :)

    1. Tomasz -- thanks again for your feedback. Let me reply to your second point first, and then I'll do the first and third together.

      You're correct, I should have finished looking at the block 1...Rd7. I stopped calculating because we found a good move for black that wins but as the old saying goes "when you see a good move look for a better one". I'll add a line about 1...Rd7 to see if it is in fact any better. We won't know until we calculate it!!

    2. Ok, I finished adding variations for black's other move 1 defensive options. Regarding your first and summary questions/points...

      On your first point, I do think it is very helpful to find and prioritize all threats. That alone will helps beginners make huge strides in chess. But I don't understand the point of organizing your opponent's threats by check/no-check. I'm not sure what that step gains you. What is more important (and is the thrust of my blog series), is that you find the greatest threat, and look at all five defensive ideas. The point is that you prioritize YOUR defensive replies based on the severity of the threat.

      For counterattacking ideas, a counterattack must be equal to or greater than the biggest threat against you. In this position white's greatest threat was the mate in 1, and the only way a counterattack could begin is with a greater threat (a check).

      That leads me to your last question about what is meant by the "ideas". This is the goal of my entire blog series on this so it is very important to me that you understand!! I'm not communicating this well if you -- my most avid read -- aren't getting it!! So let me try it this way:

      In calculation you must find candidate moves. But the key question how do you find candidate moves? Most books on calculation don't really address that.

      The answer is that you must first have an "idea" about the kinds of moves you should be looking to find. Jacob Aagaard wrote this in his "excelling at chess". For example in the opening, you know you have to accomplish certain things and the opening "ideas" like develop, castle, etc, lead you to search for specific moves that accomplish those ideas. You must have some sort of idea of what you are trying to accomplish in order to make any move in chess!!

      Ok, that's the opening but what about tactical positions? On offense how do you find good attacking candidates? How do you get ideas? I'm planning an entire blog series on this question too, but the short answer is training patterns, learning to see things like pieces in a line or loose pieces, etc. There are visual clues on the board that give you "ideas". Those general ideas help you search for specific candidate moves, that you then calculate.

      The point of my blog series is on defensive "ideas". What does that mean? A simple example is that when you are in check, there are only three defensive ideas: capture, block, move. Those three ideas guide your search for specific candidate moves that you then calculate.

      When there is any other type of threat on the board there are FIVE ideas you should always consider. So in this position when I wrote "Idea 1 (capture, pin, deflect): none", that meant that I used defensive Idea 1 to search for specific candidate moves to prevent the mate in one, but there were none.

      Does that make sense now?

  2. Thanks for you answer. Now my turn ;)

    1. prioritize all threats is the KEY - without it understand chess will not make any sense.
    2. organizing your opponent's threats by check/no-check - it is just an IDEA. If you are in check you have VERY limited possibilities - and most often when your opponent make a move without check - you can do whatever you wish (I do not say about the consequences, but the legal moves). You can use this idea in later stages of your articles, but if you do not need to do this - you can just leave it. No problem.
    3. You have explained the ideas quite well, but probably I did not ask a proper question. I meant - why you wrote the specific answers. Let me clarify what I mean:

    a) Idea 3 (move): there are no escape squares for the king, and black can't create any.
    What about playing Kg8? It is ONE square to escape (from Qf7# threat) and after playing this move - White mates in two moves. Did you omit it due to the "additional" mate scenario? Do I think in a correct way that "move - means moving the King"?
    b) the rest of "Ideas move" I understand (more or less) - however let me explain if I understand it correctly, ok?

    - Idea 1 (capture, pin, deflect):
    CAPTURE - there are no captures or no captures that stops mate in one move?
    PIN - there are NO pinning to the white pieces (this one is quite easy to understand)
    DEFLECT - there is no way to deflect the pieces (which ones)?

    I am sorry to trouble you, but I REALLY want to understand how you are evaluating the position and what is your way of thinking and looking for a defence. I thank you in advance for explaing these things to me. They are quite hard to understand, but I will try if you help me :).

    1. It is no trouble at all! I am very glad to talk about these things!!

      Yes, I completely agree on the need to identify all threats and prioritize them, but I prefer to prioritize them by evaluation instead of by check/no-check. For example there might be a double attack that starts with a check but only wins a pawn, while there might be a "silent" first move that is not a check, but threatens to mate in one move. Thus if you base your calculations off of the first move instead of the evaluation of the line, you might mis-prioritize the threats and defend against the wrong thing!!

      Okay, now on the ideas. You are correct on all counts!

      a) Idea 3 (move): there are no moves you can make with the king that allow him to run away from the checkmates (all of them, not just the mate in 1)

      b) Idea 1 (capture, pin, deflect) -- that is indeed specific to the threat on the board, not all captures, pins, deflections. Take a check for example. The only captures you can even consider are those that capture the piece giving check. The same logic applies here. In this position white's queen and rook are threatening mate, so defensive "Idea 1" is to capture either the rook or the queen. There might be other captures on the board but we don't care!! Any other captures would have zero impact on the threat and we can ignore them!! So using the idea of capturing/pinning the rook and queen (creating the mate threat) we search the board for specific moves...and find there are none. So we move on to the next defensive "Idea 2" which is to block the pieces creating the threat, and we search for specific moves that block either the rook or the queen from getting to their mating squares. We find one specific move that does that, so we add it to our candidate list and continue looking at the rest of the five ideas to generate as many defensive candidate moves as are reasonable for the position. By "reasonable" I again mean only those defensive replies that address the specific threat. We do NOT need to calculate all captures, checks, and threats. Only those that address the mate threat.

      I hope that helps, but do not hesitate to ask anything you want!!

  3. Thanks for explaining these things. I will have to repeat the previous posts and read it again to comprehend it up to next level :).

    In the future - you may think about using the system of variations at the interactive diagram. It is really hard NOT to lose in these bunch of variations since they are very similar and quite complex ones (to see it clearly without a board).

    And of course you are right - there is no need to divide moves into "check-non-check", but according to the importance of threats (priorities as a main goal is the key!).

    I am still not fully satisfied with the quality of this post, but for now - let's leave it as it is. Maybe when I see it in a new perspective and after some time (and new posts) passes by... I could suggest some better kind of improvement :).

  4. Let me know if you have any specific ideas for writing out the variations. I didn't really like the embed tool, but perhaps I could include more graphics at key positions? Have you have seen any other blogs that you think do a good job of writing out long variations? Of course the ideal is to do all of this in your head!! :)


  5. Look here:

    Please focus especially on "Online preview" chapter. I hope it helps you a bit ;) :)

    (Quote): Pgn4Web

    Pgn4Web is yet another promising viewer. Apart from the viewer itself, author provides plugins for WordPress, Joomla and MediaWiki and hints for using the viewer on many other blogs and forums.

    Pgn4Web website provides also online generator of diagrams to make preparing and publishing them easier. It is also possible to publish games using scripts hosted on their website, without installing anything.

    The viewer is open source (GPL) so usable in most contexts.

    This on-line viewer may be one of the best solution :)

    BTW. You can google "pgn chess game viewer" and check out most of these viewers. There are about 8-10 are just a few of these are really well designed and work great. I hope you suceed! :)

  6. Last minute: You can look at the blog here as see how the game with comments is done. I hope you like it:

    As far as I know - you can use the positions as much as the whole games (with commentaries, variations, symbols, etc.). Let me know what do you think about this!

    1. I'm not a big fan of the embedded viewers, since all of the analysis is in the tiny windows. I'm going to try adding more graphics of key transition points, as well as final positions, which hopefully will make this easier to read (and understand). Thanks again for the suggestions!

    2. Ok, I added a chart!! Let me know what you think. It is my first try at it but it covers the key ideas and process.