| From | Message | ||
|---|---|---|---|
|
 Planning Accurately and Calculating |
||
|
It is a pity... The thing about calculating moves is that you have as it were to play yourself blindfold when faced with any given position. The other thing is that you have to recognise when in-depth calculation is called for; it isn't always necessary, or even always possible to calculate move by move in any depth. Let us take a position from a recent annotated game: w White to play. In the lead up to this position, White had been intending to play 1.Nd6, forking the Queen and Rook and thereby forcing 1...Bxd6. But the situation demands more calculation than that, for Black would have had a swift riposte that would have won him a pawn: 1.Nd6? Bxd6 2.Rxd6 Bxh3! To see that this is good for Black, both sides would have had to calculate that 3.gxh3 is very bad for White: 3.gxh3 Re1ch 4.Kh2 Qh1ch 5.Kg3 Rg1ch 6.Kf4 (Kh4 leads to a swift mate) 6...Qh2ch 7.Kf3 Qxd6 8.Qc8ch Qf8 9.Qxa6 Qe8. Note that for much of its length, this is a forcing line - Black forcing White to play fairly specific moves in response to his own. Moves 8 and 9 show that White doesn't have a mate, and will enter the endgame the exchange (Bishop for Rook) behind. This sort of thing is easy to calculate, and the situation calls for it. The thing is to have the imagination to spot 2...Bxh3 in the course of your calculations. Now, White allowed that he missed this, but the fact is that having reached the diagram position, he found another idea - completely speculative, as he thought - that so intrigued him he put it to the test: 1.Nh6ch!! ... It may not be possible to calculate this exhaustively, but the first few moves at least are easy to see: 1... gxh6 As it transpires, 1...Kh8 scarcely comes into consideration: 2.Nf7ch Kg8 3.Ng5 g6 4.Qc3 2.Qc3 ... Threatening mate at h8. Only one defence again is available. 2... f6 3.Qxf6 Re6 Hoping to create a cover for the King once it has been forced into the centre, and to deprive the WQ of the f6-square. There doesn't seem to be any other worthwhile move... 4.Qh8ch Kf7 5.Qxh7ch Ke8 6.Qg8 ... So far, so forced. It was from here on that White had to find a way of continuing the attack. I suspect that having reached this position in his calculations, he probably figured that the resulting position was good enough for at least a draw and waited until it was reached (or close) before thinking about how to carry on. Sometimes positions are extremely complicated with all sorts of forced and unforced continuations possible. The whole thicket of possibilities is so dense that one can calculate only incompletely and hope not to miss anything vital. But there are positions in which lengthy calculation is not required at all. I'll choose a fair simple example - a straightforward pawn ending. DIAGRAM 2 w White to play, only one move comes into consideration; experienced players would play it almost without a second's thought: 1.g4! Why? There's (almost) no calculation required. White's 2 King-side pawns hold up Black's three: Black dare not advance as after capturing, White Queens first with 2 tempi in hand. That much is obvious. On the other flank, Black has to give ground 1...Ka6 or ...Kb6, after which White drives him back and back until he can abandon the a-pawn and go after Black' K-side. In effect, White is looking about 15 or 16 moves ahead, without really calculating anything! 1.g4 Ka6 2.Kb4 Kb6 3.a5ch Ka6 4.Ka4 Ka7 5.Kb5 Kb7 6.a6ch Ka7 7.Ka5 Ka8 8.Kb6 Kb8 9.a7ch Ka8 10.Kc6 Kxa7 11.Kd6 Kb6 12.Ke6 Kc6 13.Kxf6 Kd5 14.Kg6 Ke5 15.Kxh6 Kf6 (or ...Kf4) 16.Kh5 and Black's last pawn falls. Black can vary from this if he likes; White simply uses the a-pawn to force the status quo. If it were Black to play in Diagram 2, then, again, only one move comes into consideration. As g2-g4 is such a killer move for White, Black has to deal with that threat. So: 1...f5! |
||
|
and 1.....f5 would draw,of course if it was blacks move first in diagram #2. One recent game of mine also requires no further calculation as to anything but accepting or asking for a draw which any(or most) masters would recognize instantly...which sort of reminds me of your game ion,but in a different way.Instead of a pawn up in the game below,i'm up a "pawn value" in the game with a Rook vs a Bishop,but my 2 pawns vs my opponents 3 pawns.If anyone disagree's that this game would NOT draw...i'd be glad to play it out as black or white for another 50 moves and still say its a draw,regardless of any brainstorming,king moves or exchanges for white to play and try to win.Trust me,I never agree to a draw being "up"material in an endgame,unless I know its a draw for sure!In this case,all black has to do is brainlessly move his Bishop back and forth on his localized light squares and perhaps a short king move here and there,to avoid white from taking the game to any serious advantage.And thats exactly what I would do as black if I wanted to draw with the higher rated opponent to scoff up a few rating points!And thats exactly what my opponent has done in this game below.Whats interesting about this game below is that many players with lesser ratings would play the game another 50-100 moves thinking that they simply "MUST"win as white,with such material advantage as that powerful Rook vs blacks Bishop,but masters will almost always both throw in the towels in these types of endgame positions.  game |
||
|
Yep: draw. w The general position is that White has Rook vs Bishop, for which Black has just the one pawn by way of recompense. What make life hard for White is that [a] all the pawns are on one side of the board, and [b] compounding this, all the pawns are contiguous from the h-file. These considerations alone would make the win technically challenging. Worse, the Black Bishop and f-pawn have formed a bond that is absolutely unbreakable with the means at White's disposal. He needs a pawn on the f-file, and there is no way to arrange that. The final kicker is that even if White were able to find a way to exchange his remaining two pawns for Black's three, the endgame would STILL be a draw as K+R can not (except in special circumstances) defeat a K+B. A position like the diagram's we would call a 'dead' draw: there is really no play left in it. Not all 'draw' positions are so devoid of play. Consider this endgame: b Black to play. If it were White to play, then 1.Ra3! ... would lead to a very likely win for White. But it's Black's move, and a very likely draw. But Black has to work for it. The two move sequence called for scarcely requires lengthy thought: 1...Rc1ch 2.Ke2 Ra1. Now that the rook is behind the passer, White's rook can protect the pawn only from in front, or the flank (Rd4), both having their inconveniences. Let's try... 3.Rd4 ... The pawn can not advance. But suppose White had tried 3.Rd8ch Kf7 4.Ra8, instead? Then 4...Kg6! and the threat of getting in amongst the enemy pawns on the K-side stops the White King coming to the aid of his Rook on the Q-side. Note that we are hardly calculating anything very much. Part of this is that we are really imagining future POSITIONS rather than future MOVES, without worrying about the specific moves that bring these positions about. OK then, what now for Black? 3...Ra3 Cut's off the White King from further advance, and attacks the pawn at g3. This isthe situation that I'll return to in my next posting. w |
||
|
A post game computer analysis will illustrate this.The "official"edge may appear to be a full 1 point advantage for white,but its closer to approx a half pawn value diminished due to the position....and reduces to about a +.12 with pawn exchanges.I actually rate the Rook in my position above to be even less,about a +4.6 due to its true lack of attacking capabilities,but thats my analysis opinion,not others. Many lesser rated players simply do not realize that even a and h file edge pawns have a theoretical less than a 1 point value as well(even in the opening!)just as a Knight would have when placed many times on an edge file.This knowledge alone,can help win games as you reposition your pawn or Knight moves in mid or endgame.In addition,doubled pawns in the opening such as a Ruy exchange or other exchanges do NOT necessarily diminish the value of those pawns(against popular belief),but doubled pawns in the mid or endgame,or centralized doubled pawns,can indeed,have a diminished point value for a host of reasons. |
||
|
Continuing from previous posting... If White plays 4.Rd5, Black might have to decide whether to take the a-pawn, leaving the remaining pawns all on one side of the board (1.Rd5 Rxa4 2,Rxf5 Ra3) or to take the g-pawn, breaking up White's K-side structure at the cost of allowing White to relocate his rook (1.Rd5 Rxg3 2.Rxf5 (or 2.a5 f5) 2... Ra3 3.Ra5. With his King cut off, he would have to bring it all the way over to the Q-side to chase off the enemy rook, whereat it will attack the f- and h-pawns successively and win one of them. All that leaves is White to do is to give up the g-pawn at once to bring his King over rapidly to the Q-side. 4.Kd2 ... In all this, although a certain amount of analysis has been done, none has been in any depth move by move, but we have considered in a very general way the sorts of positions that might come up several more moves down the track. 4...Kf7 Why take the g-pawn at once? Clearly White is going to abandon it, so why not improve our King's position, betimes? 5.Kc2 Kg6 Black threatens again to get amongst the pawns with his King - something we mentioned as a possibility last time. 6.Kb2 Rxg3 The rook can not maintain itself on the a-file, so grabs the pawn. Here Black would have to do some calculating, as he has to be certain he can stop the White a-pawn in time. Check it out: 7.a5! Kh5! 8.a6! Rg6! The previous King move was to make room for the rook! 9.a7 ... White might instead try 9.Ra4 (getting the rook in behind the a-pawn before Black can). Black can still stop the pawn by 9...Rb6ch 10.Kf3 Rb8 11.a7 Ra8. Now, here, the only way to dig out the Rook from a8 is with the help of the King. But while that's happening, Black wipes out the White K-side, and advances a pawn. By the time Black has to give up his rook for the a-pawn, he will have a K-side pawn far enough advanced, with the White King so distant, that White will have to give up his own rook to save the draw. We'll continue this line to show what I mean: 12.Kd4 Kxh4 13.Ke5 Kg4 14.Ra1 ... (White plans to relocate his rook to g7 without losing the a-pawn - a cunning plan) 14...g5! 15.Rg1ch Kf6 and White probably dare not bring the rook back to a1, but has to go on: 16.Rxg5 Rxa7 17.Rxf5 h5, say. 9...Ra3 10.Rd7 g6! This illustrates just how much play remains in this simple looking position. Were Black to play 10...Kxh4, then 11.Rxg7 would leave the King cut off and inhibits his h-pawn's advance. Instead, 10...Kg4 is playable, intending to take the f-pawn. 11.Rxh7ch Kg4 12.Kb3 ... Instead 12.Rg7 doesn't really attack the g-pawn (12...Kxf4 13.Rxg6 Rxa7). 12... Ra1 13.Kb4 Kxf4 14.Kb5 Kg3 15.Kb6 f4 16.h5 gxh5 17.Rxh5 f3 The draw is easy to see now... 18.Rg5ch Kh2! 19.Ra5! ... Or is it? How is White to stop the a-pawn? He doesn't. 20...Rxa5 21.Kxa5 f2 22.a8=Q f1=Q and there is no win to be had. Draw! This little exercise was included to show that although the situation was at the beginning of this ending, as given in the second diagram of my previous posting, a 'book draw', it would have been worth the attacking player testing his opponent's knowledge. There is plenty of play in the position. The final position - no. w Material is dead level, and if White stops checking, Black will begin (Mind you, if it happened to be Black to move in this final position, what would the result be? ) |
||
|
BTW |
||
|
That game... Speaking of which, in a recent game between jstevens1 and a much more highly rated opponent, the game ended here when the players agreed to share the honours. b White had just played 1.Kd2 and made the offer duly accepted. It doesn't take long, without going into calculation, that there is no breakthough that can be effected by Black King or Rook. Note that there is no question of White looking for a win - Joanne is much less favorably placed: a pawn behind, the only open file occupied by the enemy rook, less space, and breakthrough chances even slenderer than Black's. The only way Black can break the deadlock is by a pawn move, ...d5 or ...f5. As either would hand White, on a platter surrounded with watercress, a protected passed pawn, one feels that maybe there is no point in calculating further. Either move certainly carries a considerable risk of a loss, especially if the rooks came off. Such an attitude would be entirely reasonable, and yet here is one situation in which I would be inclined to go in for some 'hard' calculation. First of all, I don't like 1...f5 as after 2.exf5 White might be able to make good use of the e4-square. But 1...d5 might well be worth at least a thorough look. If White plays 2.cxd5, Black's c-pawn becomes mobile, threatening to push down ro c3, and the Black King has a road into the White position. If 2.exd5, Black might even try 2...f5!? 3.gxf5 Rxc4 4.Rxg5 Rf4. If Black wants to shove some dynamite into the position, 1...d5 would blow it up all right! The slight glance I've given the position here would just be the beginning. More accurately to assess the risks, one would have to look deeper into the alternatives available the White, and what their consequences are likely to be. In the end, you might decide that Black ends up in a life and death struggle merely to hold the draw, in which case, you take the offer that has been made. Black has no other plan available that I can see. There are risks, but if they can be correctly judged as minor compared with the winning chances that are brought forth, then I'd go for it. 'Correctly judged' - that's the trick, eh? But that in my view is what this thread is all about. |
||
|
Many thanks, Ion! I am glad you looked deeply into my Expert draw game with the rook ending and put it up on this forum. Again, I am very pleased that I successfully held off a 2100+ player. Again, many thanks. Joanne |
Log in 
