Smit-Pytel 8.Rc7 - pgn-file

4 versus 3 Endgames with the d-Pawn (Part 1)

The second edition of the Encyclopedia of Chess Endings (ECE II: Rook Endings, Part 1, Goran Arsović, Zoran Petronijević, Branko Tadić, Šahovski Informator, Belgrade 2014) includes several 4 vs. 3 rook endings with the d-pawn (nos. 1346-1347 and 1349). All of them already appeared in the first edition (1985) as nos. 1181-1183.

While many analytical errors from the first edition were corrected by Petronijević, the 2014 edition still contains major evaluation mistakes - especially in positions 1346 and 1347. In addition, the second edition analyzes the endgame from Carlsen-Wang Hao 2013 (no. 1348), where a critical mis-evaluation has been missed by all commentators up to now.

In this part 1, I examine the endings Ghitescu–Rajković (no. 1347, ECE II) and Smit - Pytel (no. 1346, ECE II). The key difference between the two endings lies in the position of the f-pawn: it stands on f2 in Ghitescu - Rajković and on f4 in Smit - Pytel.

As we will see, this seemingly small difference becomes crucial whenever White tries to support the passed pawn from the queenside while Black defends passively.

Ghitescu-Rajković 1984

ECE II (no. 1347) now correctly evaluates the position as drawn, but the analysis of the line with 8.Rb2-d2 still contains a decisive mistake.

D.1



White moved the rook from the b-file to d2 - 8.Rd2, intending to support the pawn from behind. After 8…Ke7, White can penetrate on the kingside if he begins with 9.d6!

Instead of 8…Ke7, nearly 25 years ago, Müller and Lamprecht recommended 8…Ra8 (Fundamental Chess Endings, 2001, p. 232), claiming that this also draws.

D.1-1



However, White wins very simply: he retreats the rook to the b-file, shields his king from checks, and then executes the f3-g4 break. This plan works because the pawn stands on f2. With the pawn on f4, the position is drawn.

9.Kd4! Ke7 10.Rb2 Ra4+ 11.Kc5 Ra5+ 12.Rb5 Ra7

D.1-2



13.Rb6 Kd7 14.f3 Ra5+ 15.Rb5 Ra7 16.g4 Rc7+ 17.Kd4 Ke7 18.gxh5 gxh5 19.Ke5 f6+ 20.Kf5 Rc3 21.f4 Rh3 22.Rb6 Rxh4 23.Re6+ Kd7 24.Rxf6+-.

The placement of the f-pawn is the decisive factor: with f2, White wins, since permits the f3–g4 break; with f4, the position is drawn (see D.1-3, D.2-2).

The line 9.Kd4! Ke7 10.Rb2 shows that 8.Rb2-d2 is simply a waste of time. The b-file or c-file is optimal. White can immediately play 8.Kd4, invading on the queenside and supporting the pawn.

8.Kd4 Ke7 9.Rb7+ Kf6 10.Kc5 Ra5+ 11.Kc6 Ra2 12.Rb4 Ra6+ 13.Rb6 Ra7 14.Rb2 Ra6+ 15.Kb5 Ra1 16.Re2 Kg7 17.d6 Rd1 18.Kc6 Kf8 19.d7 Rc1+ 20.Kd6 Rd1+ 21.Kc7 Rc1+ 22.Kd8 Rc6 23.Ra2 f6 24.Ra8 Kf7 25.Rc8+–.

The overall conclusion from these lines is that after 7...Kf6? Black's position is lost. Instead of 7...Kf6? correct defenses are 7…Kd6 or 6…Ke7.

Smit-Pytel 1972 (Correspondence)

D.2



Game continuation: 33.Rb3? Ke7 34.Kd3 h5 35.Kc4 Ra1 36.Rb7+ Kf6 37.d5 Rc1+? 38.Kd4 Rd1+ 39.Kc5 Rd3 40.Kc6 Rxg3 41.d6? Rc3+ 42.Kd7 Kf5? 43.Rb5+ 1-0 (all marks are mine).

This endgame has been analyzed numerous times:

•      Marić (Informator 13)
•      Smit (ECE I)
•      Petronijević (ECE II)
•      Kantorovich (1989)
•      Beliavsky & Mikhalchishin (Winning endgame technique, 1995)
•      Mikhalchishin (Mastering Essential Rook Endgames, 2018)

All sources except Kantorovich claim that after 33.Rb3 Ke7 34.Kd3 h5, White is winning. Even Kantorovich erroneously thought Black loses later, after 38…Rd1+.

When I first analyzed this in 2013, I shared that view; Petronijević reached the same conclusion in 2014. However, a complete reanalysis (2023) shows that Black can draw with precise play.
In particular, after 37.d5, Black draws with either 37…Rg1! (A) or 37…Ra5! (B), instead of the game move 37…Rc1+?

D.2-1



A) 5…Rg1! 6.Rb3 Ke7 7.Kc5 Rc1+! 8.Kd4 Rd1+ 9.Rd3 Ra1 10.Rc3 Ra4+ 11.Rc4 Ra3 12.Rc7+ Kd6 13.Rxf7 Ra4+ =.

B) 5…Ra5!

This position is a reciprocal zugzwang: White to move draws, Black to move loses. 

D.2-2



White to play (See the attached file Smit-Pytel, the line 3...Kd6 4.Rb6+ .Ke7 5.d5 Ra3? 6.Rb3? Ra5 7.Rb7+ Kf6!; and file Smit-Pytel 8.Rc7) 

A) 8.Rb5 Ra7! 9.Kc5 Ke7! transposes to D.1-3.

If the rook stays on the 7th rank, Black exploits the weakness of g3, sacrifices the rook, and reaches a drawn R vs P ending.

B) 8.Rc7 Ra3! 9.Kc5

D.2-2a



With the white rook on b7, Black draws both with 9...Rxg3 and 9...Rc3+ (see D. 2-5a). In this position, however, Black has only one saving move: 9…Rc3+!

9…Rc3+! 10.Kd6 Rxg3! 11.Kc6 Rc3+! 12.Kd7 Rd3 13.d6

D.2-2b



Here, unlike in D.2-5b, Black draws because the rook is ideally placed on the d-file.

13…Kf5! 14.Rc4 f6! 15.Kc7 g5 16.fxg5 fxg5! 17.Kxg5 18.d7 h4 19.d8Q+ Rxd8! 20.Kxd8 h3 =.

Black to play loses (See the attached file Smit-Pytel, the line 3...Kd6 4.Rb6+ .Ke7 5.Rb7+ Kf6? 6.d5+–)

The black rook is forced to leave the 5th rank: 6...Ra1 6.Kc5+–.

Attacking the g-pawn is pointless, since after 6...Ra3 7.Rb3, the rook protects the pawn and cuts the black king off along the e-file. 

7...Ra5 8.Re3 Ra1 9.d6 Rd1 10.Kc5+–.

D.2-3

This position arose in the game after 33.Rb3? Ke7 34.Kd3 h5 35.Kc4.



ECE II again evaluates this as winning for White, but the ending is drawn. Petronijević correctly observes that Smit’s 41.d6? missed a win and that 41.Rb5 would have won instead, but his analysis overlooks several earlier drawing opportunities for Black.

1) Line (3...) 1...Kd6 2.Rb6 Ke7 3.Rb5 Ra3 4.Rb7

D.2-4



ECE II gives only 4…Ke6?, which loses. However, 4…Kd6 (A) and 4…Ke8 (B) both draw.

A) 4…Kd6 5.Rxf7 Rxg3! 6.f5

6.d5 Rg4 7.Kd4 Rxh4 8.Rf6+ Kd7 9.Ke5 Rg4 10.Rf7+ =.

6…g5 9.Rf6+ Ke7 10.Re6+ Kf7 11.Rh6 gxh4 12.Rxh5 h3 =.

B) 4…Ke8 5.Rb3 Ra5! 6.d5 Ke7! 7.Rb7+ Kf6 =. A reciprocal zugzwang.

2) Line (3...)1...Ra4 2.Kc5 Ra5 3.Rb5 Ra3 4.Rb7

D.2-5



Smit evaluated this as lost; ECE II continues with 4…Ke8 5.d5 Ra6? But Black draws with either 4…Kf6 (A) or 4…Ke8 (B).

A) 4…Kf6 5.d5

D.2-5a



This position was analyzed by Emms in his The Survival Guide to Rook Endings (Everyman Chess, 1999). According to his analysis, (5…) 12...Rxg3 (a) draws, while (5…) 12...Rc3+ (b) loses. The assessment of 12...Rxg3 is entirely correct; however, Black can also draw with 12...Rc3+.

b) 12…Rxg3 13.d6 Rc3+!

Emms analysed 13...Rf3? 14.Rb4 Rc3+ 15.Kd5 Rc1 16.Re4 Rd1+ 17.Kc6 Kf5 18.Re5+ Kxf4 19.Rd5 Rc1+ 20.Rc5 Rd1 21.d7+–; and 13...Rg4? 14.Rb8 Rg1 15.d7 Rc1+ 16.Kd6 Rd1+ 17.Kc7 Rc1+ 18.Kd8 Re1 19.Rb6++–.

14.Kd5 Rd3+! 15.Kc6 Rc3+! 16.Kd7 

D.2-5b



We have transposed to the game after 42.Kd7. Black has only one move to draw - 16…Rc4! This move was suggested by Beliavsky and Mikhalchishin (Winning Endgame Technique, Batsford, 1995, p. 73) as a better try. In Practical Endgame Play (Everyman Chess, 2008, p. 128), Grivas demonstrated that Black draws if, after 12.Kd8 Rxf4 13.d7 Rxh4 14.Rb6+, he plays 14...Ke5! instead of 14...Kf5? (Beliavsky/Mikhalchishin). 

16…Rc4!

16...Kf5? was played in the game Smit–Pytel, and after 17.Rb5+, according to Smit (ECE I, No. 1181, p. 284), Black resigned.

17.Rc7 Rxf4! 18.Ke8 Rxh4 19.d7 Re4+ 20.Kf8 Rd4! 21.Rc6+ Kg5 22.Ke7 Rxd7+! 23.Kxd7=.   


B) 4…Ke8 5.d5

D.2-5c



Petronijević incorrectly gives 5…Ra6? Instead, 5…Rxg3 draws. 

5…Rxg3 6.Kd6

D.2-5d



6...f5! 7.Re7+ Kf8 8.Rc7 Ke8 9.Rc8+ Kf7 10.Kc6 Rc3+! 11.Kd7 Rh3! 12.d6 Rxh4 13.Kc7 Rxf4 14.d7 Rc4+  or 14...Rd4=. 

3) Line (3...)1…Ra7 2.d5

D.2-6



Petronijević’s 2…Rc7? loses. Black can draw with 2...Kd6, 2...Ra4 or 2...Ra5.   

All three transpose into known drawing structures.

Back to D.2

Both ECE editions present the position from 35.Kc4, while Marić begins at 32…Kf8. His 33.Rb3!, awarded an exclamation mark, actually throws the win away. 

White should play 33.Kb3!, seizing the a-file and marching the king to the queenside.

1.Kb3! Ke7 2.Ra2! Rb5+ 3.Kc4 Rb1 4.Kd5 h5 5.Ra7+ Kf6 6.Kc6 Rc1+ 7.Kd7 Rd1 8.Ra4 Rd3 9.Rc4! Rxg3 10.d5 Rg4 11.Rc6+! Kg7 12.d6 Rxf4 13.Kc7! Rxh4 14.d7! Rd4 15.Rd6 Rc4+ 16.Kb6 Rb4+ 17.Kc5 Rb8 18.d8Q Rxd8 19.Rxd8+–.

A comparison of Ghitescu-Rajković with Smit-Pytel is extremely instructive. The key difference lies in the position of the f4-pawn, which in the latter case weakens the g3-pawn and leaves White’s kingside more vulnerable, directly increasing Black’s chances of a draw. As a result, Black possesses significantly more defensive resources in Smit-Pytel than in Ghitescu-Rajković.