Skype coaching session, 07 May 2017. 

Please note: Re-structuring of the sessions now offers the option for guests to bring their own games for group discussion. 

Recordings of some previous coaching sessions can be found at: 

http://www.open-aurec.com/Skype/PaulBenson/PaulBenson.htm 

This posting is divided into three sections: 
#1. Annotated game, 49 moves. 
#2. Endgame study arising from black move 32. 
#3. Unannotated game, 49 moves. 

Paul Benson. 

* * * 

Annotated game. 

White: Babu M. R. Lalith, 2526. 
Black: Rolando Nolte, 2453. 
Event: Kolkata Open 2012. 
Result: 1-0 in 49 moves. 
Opening: Gruenfeld Defence, Exchange Variation, Nadanian Attack, D85. 

1. d4 Nf6 
2. c4 g6 
3. Nc3 d5 
4. cxd5 Nxd5 
5. Na4 

| This move introduces the Nadanian Attack. It seems to break two fundamental rules of opening strategy. Firstly, a piece has been moved twice when perhaps other pieces or pawns could be offered that tempo. Secondly, the knight is now placed on the edge of the board meaning it influences considerably less squares than when sitting on the more central c3 square. There must be some gains or Grand Masters would not have it in their repertoire. Firstly, it is designed to deny black the option of the, Nxc3, capture which will be followed with pressure on the h8 - a1 dark square diagonal. Secondly, it controls the c5 square which makes the thematic black central challenge of, pawn c5, much more difficult to organise. Essentially the Nadanian Attack neatly sidesteps all those dynamic lines of the mainline where white takes a pawn centre and black tries to undermine it. Is it better than the mainline beginning with, 5. e4 Nxc3 6. bxc3 Bg7, with tension when, pawn c5 arrives? Wrong question. Perhaps players with the white pieces should ask themselves if those sharp mainline variations suit their style of play. If not, then here is a possible replacement where strategy more often than not overrules tactics. Do some homework and rewards are waiting to be reaped. || 

5. ... Bg7 

| Having said the Nadanian Attack is more often strategic than tactical, a special mention must be given to, 5. ... e5, which needs some study by the player with the white pieces if some short-term complexities are not going to cause confusion the first time they hit you. || 

6. e4 

| The less direct, 6. Nf3, is also quite popular here. || 

6. ... Nb6 

| The knights are again in conflict albeit on different squares. The white queen must now maintain a defence to the a4 knight, not a problem as such, but a point for those inexperienced in the system to bear in mind. || 

7. Nf3 

| Instead, 7. Be3, is the alternative means to protect the doubly-attacked d4 pawn. || 

7. ... Bg4 

| Playable, but much less popular than the logical, 7. ... O-O. || 

8. Be3 

| White can transpose the developing moves of, Nf3, and, Be3, if desired. Once white has a bishop on the e3 square, then the black queenside break of, pawn c5, is going to take much organising. || 

8. ... Bxf3 

| A very committal choice. Instead, 8. ... O-O 9. Nc5, tests the attitude of black. Some prefer the calm, 9. ... Qc8, while hardier individuals, will venture forth with, 9. ... Nc6 10. Nxb7 Qb8 11. Ba6 Nb4 12. Nc5 Nxa6 13. Nxa6 Qc8 14. Qe2, with a long tough struggle on the horizon of both players. Note the line just offered is not forcing, each side can choose from less explored variations yet to receive extensive testing. A less aggressive approach for black other than the game move could be, 8. ... Nc6 9. d5 Ne5 10. Be2 O-O, where chances seem equal. || 

9. gxf3 

| Clearly white is not interested in offering a pawn with, 9. Qxf3 Bxd4 10. Rd1 c5, when it is black who seems to have all the compensation for the material. || 

9. ... N8d7 

| My limited database gave only two other examples here. Choose from either, 9. ... e5, or, 9. ... O-O, with all to play for in both situations. || 

10. Nc3 

| The knight having don it's job of avoiding being exchanged on the c3 square, now returns to start to fight for the centre. Of course, this sidestep and retreat manoeuvre seems to have wasted two moves, and such a claim is difficult to deny. However, we must remember why white chooses, 5. Na4, in the first place. It is to avoid those sharp, heavily-analysed lines where black trades knights on c3 and then creates play by undermining the white centre. Choosing the Nadanian Attack makes it difficult for black to initiate low-risk aggression. || 

10. ... O-O 

| Perhaps this was the last chance for black to fight in the centre with, 10. ... e5. || 

11. f4 

| So the possibility of, pawn e5, for black is ruled out for once and for all. Now black must organise a, pawn c5, central challenge if there is going to be any sort of a fight. || 

11. ... Rc8 
12. a4 

| Not content with a large centre, white snatches some queenside space. Black must make some sort of response immediately else the further advance of, pawn a5, will push the b6 knight onto the a8 square, after which a considerable investment of time will be needed to resuscitate it. || 

12. ... c5 

| At last, however the shuffling around since move 5 means a Gruenfeld Defence has morphed into something close to resembling a Benoni Defence. With that point in mind, just how should white respond to this long-delayed central challenge? The choices when pawns contact each other like this are, capture, push past, or sit tight. || 

13. d5 

| Push past every time, unless you have precise references giving advice on how to capture or sit tight. || 

13. ... a5 
14. e5 

| Staking such a claim in the centre needs careful strategic consideration. A simpler method here was, 14. Bg2, to be followed with, 15. O-O. || 

14. ... e6 

| The critical response. Black makes the standard challenge to the white d5 pawn, but the alternate plan of, Nb8 - Na6 - Nb4, seems rather slow and just what would the knight do on b4 anyway? || 

15. dxe6 

| White wants an open position with a weakness in the black centre. Ah, but hang on there. Surely an enemy unit is only weak if it can be attacked, right? || 

15. ... fxe6 
16. Qb3 

| The first attack is placed on the e6 pawn. It might be worth taking some time to calculate how many white attacks can bear down on the backward, isolated black e6 pawn. || 

16. ... Qe7 
17. Bh3 Rce8 

| So, white has two attacks while black is being tied down to two defences. Are there any more units available for white to increase the pressure on the black e6 pawn? Oh, and while we are discussing weaknesses, is there anywhere in the white position which could be counter-targetted by black? || 

18. O-O-O 

| A rather paradoxical opposite-wing castling middlegame is in progress. The usual strategy is to throw pawns at the opposing king, open lines with the odd sacrifice if need be, and then try to invade with as many pieces as possible. All this before your opponent does similar unto yourself. But here this apparent rule of mutual all-out attacking is not appropriate. The black 3 pawns against two pawns queenside majority cannot be mobilised, so black must resort to fighting it out in the centre / kingside. || 

18. ... Bh6 

| Black finds a white weakness and places a double attack on it. This passes an important decision back to white. White can either defend the vulnerable f4 pawn with, 19. Ne2, or instead increase the tension in the position with, 19. Rd6. Happy calculating! || 

19. Rd6 

| For better or worse, this is not the time to start defending. The black e6 pawn is now triply attacked and cannot be further defended. White should also be thinking in terms of potential discovered checks by the b3 queen as the tactics begin to roll. || 

19. ... c4 

| Black fears a potential white queen discovered check will prove extremely dangerous and so rules it out. Fine as far as it goes, but sometimes you prevent one disaster only to set up a different one. It is true that black has blocked that tender a2 - g8 diagonal, but it only seems to deflect the white queen to a square where she finds new strengths. || 

20. Qb5 

| On b5 the white queen has increased the pressure on the black d7 knight and also created an x-ray attack on the black e8 rook. || 

20. ... Bxf4 

| Setting white an interesting test. Should white hit immediately on e6 with tactics, or should there be an exchange of bishops on f4 first? || 

21. Bxf4 

| White needs to trade bishops first rather than let black play, Bxe3+. This is important, there is a weakness soon to be in need of protection which will not be possible if white has a pawn on the e3 square. || 

21. ... Rxf4 

| With this capture the black rook activates itself, places an attack on the unprotected white f2 pawn, but significantly weakens the black back rank. || 

22. Rxe6 

| So white removes the pressure on the black d7 knight, it might appear that black is going to come out of this in tact. Not so. Careful study here should reveal matters are actually very unhealthy for black. || 

22. ... Qxe6 

| A sad necessity. Instead, 22. ... Qf7, allows white to demonstrate how the black back rank is weak with, 23. Rxe8+ Qxe8, the black queen is drawn into an x-ray attack from the white queen. White now centralises with, 24. Rd1, triply attacking the black d7 knight. Black can try, 24. ... Rf7, but now, 25. e6, and black is losing copious amounts of material on the light squares. || 

23. Bxe6+ Rxe6 

| So black has given up / lost the queen for a white rook and bishop. Judging who has the better prospects here depends on several factors. If the black pieces can flood into the white position then black is fine. If no such activity emerges, then black will have to start defending against the multi-directional threats created by the white queen in a fairly open position. || 

24. Re1 

| White had to choose to save either the e5 pawn or the f2 pawn. Why save the e-pawn, when surely preventing a potential black rook invasion onto the 7th rank made more sense? Simple. The e5 pawn is more advanced and so not only restricts the black pieces but can also become an attacking unit once a little re-grouping to evict the black e6 blockading rook takes place. || 

24. ... Rxf2 

| One black rook is active, the other black rook is taking up the unhappy role of blockader, but those black knights are not yet contributing to counterplay. White must find a good idea here or the advantage might begin to slip away. Clue: The tactics might appear to have fizzled out, perhaps true, or perhaps there are some neat strokes in need of discovery? || 

25. Nd5 

| Surely this knight has just blocked a vital defence from the white queen to her desirable e5 pawn? Yes, the e-pawn has been deliberately undefended. If this bold advance did not occur to you last move, then do not worry. Instead, start calculating the complexities of letting the e5 pawn be captured by black in either of two ways, and then the possibility of black ignoring the dainty potential morsel also needs examining. || 

25. ... Nxe5 

| The best try. The alternate capture of, 25. ... Rxe5, simply drops a piece to, 26. Rxe5 Nxe5 27. Ne7+ Kf7 28. Qxe5, and both the white queen and knight are well located while black is without either play or a plan. Declining the central capture with, 25. ... Rxh2, allows the remaining white forces to coordinate wonderfully with, 26. Nc7, forcing the black blockading rook to give way. There are two losing options for the rook: 
(A). If, 26. ... Re7 27. e6, deflecting a vital defender. After, 27. ... Nf8 28. Qxb6, as well as just having lost a piece black will be forced further back next move with, Nd5, and the passed e-pawn will then be unstoppable. 
(B). Or if, 26. ... Re7 27. e6 Nc8 28. exd7 Rxe1+ mate, is not going to happen, is it? White can improve on this horror. 
(C). Or if, 26. ... Re7 27. e6 Nc8, white has the startling, 28. Qxd7, when, 28. ... Rxd7 29. exd7, and the passed d7 pawn promotes. 
(D). Or if, 26. ... Rc6 27. e6 Rxc7 28. e7 Nf6 29. Qxb6 Rc8 30. Qxf6, too painful for words. 
(E). Or if, 26. ... Rc6 27. e6 Rxc7 28. e7 Nf6 29. Qxb6 Ne8 30. Qe6+ Kg7 31. Qf7+, wins the e8 knight with more booty to follow soon. 

White in the game must have something good against black's, Nxe5, which threatens, Nd3+, leading to a massive win of material. || 

26. Rxe5 Rxe5 

| So just who is x-ray attacking whom across the white 5th rank? || 

27. Nf6+ Rxf6 
28. Qxe5 

| Using pawn-count accountancy, it appears that black is actually a pawn ahead. Well, that old pawn-count method is a useful guideline to keep tally of captured material, but does not take positional matters into account. Those black queenside pawns are going to be very difficult to defend. Had black instead a pawn formation such as, pawn a7, pawn b6, pawn c5, then a simple rook placement on the black 2nd rank would protect everything, and black should be able to hold the draw. As it happens, in the game there is one simple drawing method for black just sitting in the position waiting to be discovered, the only problem then is how to arrange it, and white of course will be fighting to prevent it. 

Imagine everything on the queenside, except the white king, were removed from the board. If this can be achieved, then black has a draw by playing, Rf5, to be followed with a continual repetition of, Rh5, then, Rf5, then, Rh5, then, Rf5, and so on. The oscillating black rook is continually protected by the black g6 pawn, the white king cannot cross that black 4th rank, the white material advantage is unable to clinch the point. Once this oscillating black rook kingside defence is understood, then all black has to do is to find a method to eliminate all the white queenside pawns, even if it costs the black knight in the process. || 

28. ... Rf1+ 
29. Kd2 c3+ 

| A nice try which demands white defends precisely, flipping a four-sided coin for a defensive decision when the tactics are rolling will not suffice. Instead, 29. ... Nxa4 30. Qe8+ Kg7 31. Qxa4 Rf5 32. Qxc4, is almost the position which shall be examined in the supplementary endgame section. || 

30. Qxc3 

| The undesirable choices to be rejected are: 
(A). If, 30. bxc3 Nc4+, the knight invades on the vacated square with a royal fork, black is clearly winning. 
(B). Or if, 30. Kxc3 Nxa4+ 31. Kb3 Nxb2 32. Kxb2 Rf5, and black has a blockade and a draw. It would be inadvisable to try and make use of the queenside pawns, simply be grateful to escape with a hard-earned draw. 
(C). Or if, 30. Kc2 cxb2 31. Kxb2 Nxa4+ 32. Ka3 Nb6 33. Qe6+ Kg7 34. Qxb6 Rf5, and black has that blockade for a draw again. || 

30. ... Rf2+ 
31. Ke1 

| A good example of retreating to avoid disaster when the natural tendency would be to push forward. Advancing with the king loses as follows: 
(A). If, 31. Ke3 Nd5+, a royal fork costing white the queen and black wins. 
(B). Or if, 31. Kd3 Rf3+, a royal skewer this time and black wins yet again. || 

31. ... Rf5 

| The best square for the rook, black is aiming for the blockade but surely white is about to secure his queenside pawns, making it impossible for their removal? || 

32. b3 

| And there we go. The black rook and knight will not be able to combine against the white b3 pawn but the white queen has prospects of picking up the black b7 pawn and trying to make use of the 2 against 1 queenside majority to create a fast-running passed pawn. Black to play in the game might have a fascinating idea to find here, inspiration might be lurking in the quietest of positions. || 

32. ... Rh5 

| Black has planned for a blockade, unfortunately it cannot be achieved by force but instead relies on white being careless enough to permit it. Consult the supplementary endgame section given after this game for an in-depth study of a fascinating alternative for black which might have offered practical drawing chances. || 

33. Qc7 Nd5 

| White to play can capture either of the unprotected black queenside pawns. While all pawns start out "equal", as the game proceeds some become far more "equal" than others. Which pawn should white capture and which one should be left alone? || 

34. Qxa5 

| While captures of either black queenside pawn will give white a queenside pawn majority, the alternate snatch on b7 would only yield disappointment. After, 34. Qxb7 Nb4, black has set up an unbreakable blockade and would hold the draw. The f5/h5 rook protects the backward a5 pawn, the a5 pawn protects the b4 knight, and providing the knight sits still for the rest of the game, the white queenside pawns will be immobilised. The game capture, Qxa5, leaves the white queenside pawn majority capable of movement leading to the creation of a passed pawn. This is not yet an automatic win, white must ensure one last rabbit cannot be pulled from the hat by black to produce that rook against queen blockade. || 

34. ... b6 

| Black has set up a static defence, the rook protects the knight, the knight protects the b6 pawn, but this does nothing to halt the mobilisation of the white queenside pawn majority. || 

35. Qa7 Rf5 
36. Kd2 h5 

| International Master slightly worries Grand Rabbit. If all the queenside pawns are eliminated, say at the cost of the black knight, then black has an easy forcing blockade with the repetitious, Rh5 - Rf5 - Rh5. While this queenside clearance plan might not actually be available to black should white play precisely, at least keeping the simple blockade formation was surely sensible? As it happens, should there be a queenside clearance then the black rook can still create a blockade by repeated shuffling between f5 and f7, but why make life difficult for yourself? On a totally different matter, the plan of trying to create a kingside passed pawn for black will not succeed. The rook needs protection to support the centralised d5 knight, so any advance of the g-pawn will dismantle the defensive skeleton inevitably leading to material losses. Sometimes the best form of defence is attack. And sometimes the best form of defence is defence. || 

37. b4 Rf2+ 

| The white b-pawn is poisoned. If, 37. ... Nxb4 38. Qxb6, attacking the loose b4 knight and threatening a horrendously expensive, Qxg6+, courtesy of the over-enthusiastic black h-pawn. Push passive pawns at your peril. || 

38. Ke1 Rf5 
39. a5 

| White is now guaranteed a passed pawn, the only question is whether the black knight can sacrifice itself on it or will the white passer cost black a rook? || 

39. ... bxa5 

| Instead, 39. ... Nxb4 40. Qxb6, again hitting the unprotected b4 knight and g6 pawn. Evasion with, 40. ... Nd3+, fails to, 41. Kd2, the black pawn and knight are again unprotected. || 

40. bxa5 Re5+ 

| Black aims for another defensive formation but all it can do is slow down the progress of the white passed a-pawn. The black knight needed to be on the c5 square in order to sacrifice itself on the advancing white a-pawn, isn't 20 - 20 hindsight wonderful? It would then be up to the black rook to shuffle between f5 and f7 to keep the white king from entering the kingside. These rook versus queen blockades get everywhere. || 

41. Kd2 Re8 

| White's moves are now easy to find and understand. All it requires is avoidance of a royal fork from the black knight. || 

42. Qd7 Nf6 
43. Qc7 Ra8 
44. a6 Ne4+ 

| Clearly, 44. ... Rxa6 45. Qc8+, forking the black king and rook is not going to happen. Nothing can stop the white a-pawn now, black is going to lose a rook in the transactions. || 

45. Kc2 Nf6 
46. a7 g5 
47. Qb8+ Rxb8 
48. axb8=Q+ Kf7 
49. Kd3 
Black resigns, 1-0 

* * * 

## Endgame study arising from black move 32. 

Position with black to play move 32: 
White: Ke1, Qc3, Pa4, Pb3, Ph2. 
Black: Kg8, Rf5, Nb6, Pa5, Pb7, Pg6, Ph7. 

Here is a sample of rambling shambling thoughts yer average Grand Rabbit might construct during play in order to understand a tricky ending. 
All ideas produced in this comment are bio-organic, no silicon beasties have been consulted, they would almost certainly disagree either in general or at specific points. 
During the session an idea of immediately sacrificing the black knight to achieve a blockade caused much excitement. 
It all starts with the sacrificial, 32. ... Nxa4, handing the awkward decision back to white. 
There are two options for white to consider and then black has two options within white's better choice, three plans in all. 

Plan A. 

White accepts with the forcing sequence of, 33. Qc8+ Kg7 34. Qxb7+ Kg8 35. bxa4, and black is given that rook blockade unless white tries some inspirio-desperation. 
The general idea is for white to perhaps retreat the queen to somewhere like g3, advance the king to c4, and then get down to some hard work. 
White must calculate a potential king and pawn ending arising after a queen offer of, Qc5, and, Rxc5+. 
Note that if black declines an immediate capture of the queen, then the white king marches to b6, which would be followed by, Qxa5, so black might as well grab the queen at the first offer. 
Upon the recapture of, Kxc5, the black kingside pawns can organise a promotion in 7 moves. 
That is, pawn g5, pawn g4, pawn h5, pawn h4, then pawn g3, and whether white captures or not, black promotes in 2 more moves. 
Note that if white does not capture on g3, then black will have the option of promoting on either file. 
As it happens, white is recommended to capture with, hxg3, forcing the black promotion plan onto the g-file, this is necessary as if black is allowed, gxh2, then the new black queen guards the white promotion square of a8, whoops! 
Incidentally, when engaging in move counts on one side of the board, the white capture of, hxg3, followed by black replying, hxg3, does not alter the moves count. 
Each capture cancels the other out as far as counting moves to reach promotion is concerned. 
Think of the kingside play as black making 8 moves and white making 1, leaving the overall promotion count being 7. 
Meanwhile over on the queenside in the white part of the promotion race we count the moves of, Kb5, Kxa5, Kb5, and then 4 pawn moves to promote, total 7. 
So, the promotion race seems to be 7 moves each, black promotes first to be immediately followed by a white promotion. 
We can conclude that after, 32. ... Nxa4, and white captures with the b3 pawn, white cannot win and the black inspirational knight sacrifice receives the reward of a draw. 

Looks good, does it? 
There is a very serious hole in the above plan. 
The white idea of using the queen as a shield to permit the white king to reach b6 and support, Qxa5, is badly flawed. 
Consider the white queen on c5 with the white king sitting on b5 as part of the plan to reach b6, and it is black to move. 
Here black can take full advantage of the skewered white royalty, the big idea is to get the promotion race started a move earlier with, pawn h5, while the white queen is in a pin. 
White must continue with, Kb6, but now, Rxc5, and when white plays, Kxc5, 
black will now win the promotion race as described above on move6, while the white pawn in reply to the black coronation only manages to land on a7 the move after. 
A promotion is no longer being followed with an immediate counter-promotion, a consequence of black hitting out with a timely, pawn h5. 
The new black queen then flicks herself to h1 and the white a-pawn promotion on the a8 square is denied, black actually wins! 
Tricky things endings, aren't they? 

Plan B. 

After the inspirational, 32. Nxa4, white shuffles to capture the knight with the queen, 33. Qc4+ Kg7 34. Qxa4. 
This now passes a decision back to black, which pawn does black choose to keep, a5, or, b7? 
For Plan B, we will have black keeping the b-pawn, keeping the a-pawn is discussed in Plan C later on. 
So we have the moves, 32. Nxa4 33. Qc4+ Kg7 34. Qxa4 b5, letting the a5 pawn go. 
This permits the white king to infiltrate into the black position by using the squares, b4, a5, b6, and then march toward the black king. 
Black would be advised to keep the h-pawn on h7, those kingside pawns should not march up the board. 
Their advances would only give the white queen space in which to infiltrate and then shuffle the black king toward the centre until something probably gives. 
If allowing the white king to reach squares like e6 or e7 is a win for white it will demand good technique. 
Is this a better plan for black than the one chosen in the game? 
It would definitely take more time, more moves, and the more moves you force your opponent to make, then the greater their opportunity they have to mishandle the position. 
This philosophy of course applies to a defender, but if one is lost anyway, why not force the opponent down the most difficult route to victory? 

There is a different plan available for white other than infiltration into the black centre / kingside to be considered. 
After the sequence of, 32. ... Nxa4 33. Qc4+ Kg7 34. Qxa4 b5 35. Qxa5, and the white king marches to b4, there is a possibility of white simply playing, Qxb5, leading to pawn promotion races similar to those discussed in Plan A above. 
When white recaptures with, Kxb5, the opposite-wing races start with black making the first move and this again is 7 moves to promote. 
White needs to move the king out of the way with, Kc4, and then follows 5 pawn moves to promote making a total of 6 moves. 
This means white promotes first and black promotes immediately after, it is a draw, so perhaps the white king should slip around the black b5 pawn to try for a squeeze in the centre / kingside after all. 

There is another important method in Plan B to investigate. 
Again the moves are, 32. ... Nxa4 33. Qc4+ Kg7 34. Qxa4 b5 35. Qxa5, the white king marches to b4 and white enters into the queen for black rook and b-pawn trade. 
However this time white is going to run the king back to the kingside to try and prevent black from promoting. 
When the black pawn advances have reached, g4, and, h4, the sideways sprinting white king can reply by landing on f4. 
Black can liquidate all the kingside pawns with, pawn g3, a pair of captures, and the white king snatches the remaining pawn with, Kxg3. 
Now comes the crunch. 
Remember when the white king was marching to the b4 square to set this liquidation process in motion? 
We did not have any moves for black, not fair, time to recompense. 
Let us assume the undisclosed moves were spent by shuffling the black king, perhaps between the g7 and f7 squares? 
We would now have a position with the white king on g3, a white pawn on b3, and the black king either on g7, or, f7, with black to make a move. 
If the black king is on f7 when the sprint begins, then the black king can reach a drawn position by travelling via, e6, d5, c5, to land on the b4 square, a textbook draw. 
However, if the black king is on the less favourable g7 square then some subtlety is required. 
An important feature to note here is that the kings are in distant opposition to each other. 
The key for black is to run along the black 2nd rank and white in turn will run the king along the white 3rd rank, and as such white keep retaining the distant opposition. 
This might appear to be favouring white, but this is only temporary. 
When black plays, Kb7, white cannot claim the distant opposition with, Kb3,because there is a pawn on the b3 square. 
Consequently white must concede the distant opposition and it will soon be a draw. 
If white ever tries stepping onto the 4th rank with, Kc4, black must ensure there is always the necessary reply of, Kc6, finally taking the opposition. 
If white ever tries, Kb4, black must be able to reply with, Kb6, opposing again. 
Lastly if the black king ever reaches the b5 square it is easily drawn. 
Going back to where the two kings were on, g3, and, g7, white might respond differently when the black king to move slides along his 2nd rank. 
If black plays, Kf7, and white tries, Kf4, then black must change strategy and similarly advance to the centre with, Kf6, claiming the opposition and a draw will again result. 
In this way of stepping onto the same file as the white king, the black king always ensures his catching up move across the board will gain the opposition. 
If black was really clever when making waiting moves while the white royalty were setting up the liquidation on the black b5 pawn, then the king would be placed on the f7 square and the rook would shuffle, Rh5, Rf5, Rh5, Rf5. 
Not only are endings tricky, they are also very brain-draining. 

Plan C. 

This occurs after, 32. ... Nxa4 33. Qc4+ Kg7 35. Qxa4, and black simply allows white to win the b7 pawn whenever desired but keeps a strong hold on the a5 pawn. 
This now takes us back to similar strategy of Plan A, that is white marching the king across to the a4 square, with the idea of combining with the queen to capture the black a5 pawn. 
There is however a big difference from Plan A, there is no longer a white pawn occupying the a4 square. 
So after white organises, Qxa5, and the black reply, Rxa5+, and then white, Kxa5, the promotion races begin again. 
Black always needs 7 moves on the kingside, but now the white king is excellently placed on the a-file, that is it is not blocking the path of the b-pawn, white can now promote the b-pawn in only 5 moves. 
When the white promotion occurs the black pawn can reply with either, pawn g2, or, gxh2, but in both cases, the white queen can prevent promotion with, Qb1, and then swing kingside to mop up. 

In both Plans B and C there is also the possibility of the black king trying to race into the queenside to prevent the sprinting white passed pawn. 
In these instances, black is not making progress on the kingside, so white can similarly make a king move on the queenside to either prevent the black king reaching a vital square, or more likely, the white king advances to the 7th rank to control the promotion square. 

Conclusion on, 32. ... Nxa4. 
Plan A: 33. Qc8+ Kg7 34. Qxb7+ Kg8 35. bxa4, black has at least a draw, and a win if white plays too ambitiously. 
Plan B: 33. Qc4+ Kg7 34. Qxa4 b5, 35. Qxa5, white might be able to squeeze a win by marching the king into the heart of the black position via the a5 square, difficult to assess. 
Plan C: 33. Qc4+ Kg7 34. Qxa4, and black keeps hold of the a5 pawn and lets the b-pawn go, white seems to have a win by combining against the black a5 pawn. 

* * * 

## Unannotated game. 

White: Babu M. R. Lalith, 2526. 
Black: Rolando Nolte, 2453. 
Event: Kolkata Open 2012. 
Result: 1-0 in 49 moves. 
Opening: Gruenfeld Defence, Exchange Variation, Nadanian Attack, D85. 

1. d4 Nf6 
2. c4 g6 
3. Nc3 d5 
4. cxd5 Nxd5 
5. Na4 Bg7 
6. e4 Nb6 
7. Nf3 Bg4 
8. Be3 Bxf3 
9. gxf3 N8d7 
10. Nc3 O-O 
11. f4 Rc8 
12. a4 c5 
13. d5 a5 
14. e5 e6 
15. dxe6 fxe6 
16. Qb3 Qe7 
17. Bh3 Rce8 
18. O-O-O Bh6 
19. Rd6 c4 
20. Qb5 Bxf4 
21. Bxf4 Rxf4 
22. Rxe6 Qxe6 
23. Bxe6+ Rxe6 
24. Re1 Rxf2 
25. Nd5 Nxe5 
26. Rxe5 Rxe5 
27. Nf6+ Rxf6 
28. Qxe5 Rf1+ 
29. Kd2 c3+ 
30. Qxc3 Rf2+ 
31. Ke1 Rf5 
32. b3 Rh5 
33. Qc7 Nd5 
34. Qxa5 b6 
35. Qa7 Rf5 
36. Kd2 h5 
37. b4 Rf2+ 
38. Ke1 Rf5 
39. a5 bxa5 
40. bxa5 Re5+ 
41. Kd2 Re8 
42. Qd7 Nf6 
43. Qc7 Ra8 
44. a6 Ne4+ 
45. Kc2 Nf6 
46. a7 g5 

47. Qb8+ Rxb8 
48. axb8=Q+ Kf7 
49. Kd3 
Black resigns, 1-0 

* * *