The game Lerner-Dorfman (1980) is well known in the endgame literature. It has been recently extensively dealt with also in Micawber`s excellent tutorial (see the thread “Bacrot-Robson; a few explanations”,  D. 30 and 31 as well as the attached file). 2011 I made some corrections in the analysis of this ending which showed that White’s winning resources after 58…Ra4+? are greater than one thought before. 



Lerner - Dorfman
URS-ch FL48 Tashkent, 1980

58…Ra4? 59.Kd5 Ra3 60.a6 Kf6

This move suggested by Anikaev in Informator 30 is stronger than the actual game continuation 60...Rxf3. 
 


2003 the endgame theory has correctly established that 60…Kf6 also loses and that the decisive mistake of Black was 58…Ra4? In 2011 I have proved that 

1) in the line 61.Kc6 Kf5 62.Kb5 White wins not only by 63.Rf8 (Steckner) but also by 63.Rc8.
2) in the line 61.Kc5 (Vulfson)  61...Kf5 62.Kb4 Ra1 63.Kb5 Rb1+ 64.Kc6 Rc1+ the move 65.Kb7 does not throw away the win. 

The established theory assumes that after 61.Kc6 Rxf3 as well as after 61.Kc5 Rxf3 White throws away the win if he plays 62.Rb8. This opinion was presumably based on the initial analysis of Anikaev who after 61.Kc6 Rxf3 62.Rb8 Ra3 continued 63.Rb6?  which indeed lets the win slip – 63…Kf5 64.Kb7 Kg4 65.a7 Ra7+ 66.Ka7 Kg3 67.Rb4 f6 (Averbakh) followed by 68…g5=. But why not to play 63.Kb6 instead of 63.Rb6?  As I shall demonstrate this obvious move secures the win. 

61.Kc5

Or 61.Kc6. The small drawback of 61.Kc6 is that after 61…Kf5 White can no longer hit at the rook from b4 as in the line 61.Kc5 Kf5 62.Kb4. 

61…Rxf3 



62.Rb8

The theory condemns this move and suggests only 62.Rd8+-. 

62…Ra3



63.Kb6!

This is my improvement over Anikaev`s 63.Rb6? 
63.Kb7? would be a fatal mistake because in the ensuing rook versus pawn position the White king is one square further away from the king side – 63…Kf5 64.a7 Kg4 65.a8Q Rxa8 66.Rxa8 Kxg3 67.Ra4 f6=. Not  67...f5? (as suggested in the tutorial) in view of 68.Kc6 f4 69.Kd5+-.  

62...Ra3 63.Kb6 Rb3+

63...Kf5 64.a7 Kg4 65.a8Q Rxa8 66.Rxa8 Kxg3 67.Kc5 Kxh4 68.Kd4+-. 

64.Ka7



We have now transposed to a position which has been already analysed by K. Mueller in his comments to Bacrot-Robson (the line 73.Ka7! Rb3 74.Rb8). It has been thoroughly dealt with in the tutorial of Micawber D. 33 after 2.Rb8 and in the attached pgn-file.

64...Rxg3 65.Rb5 Re3 66.Kb6 Re6+ 67.Ka5 Re7 68.Rc5 Ke6 69.Kb6 Re8 70.a7 f5 71.Ra5 f4 72.a8Q Rxa8 73.Rxa8 Kf5 74.Kc5 Kg4 75.Kd4 Kxh4 76.Ke4 Kg3 77.Rg8+- (K. Mueller).