RR v Bill Armstrong after 13 ... c5

r1bq1rk1/1p1n1pbp/4n1p1/pNp1p3/2P1P3/4BN2/PPQ1BPPP/3R1RK1 w
RR had already decided he didn't want a game with knights perched on d4 and d5 (hence his on b5) for no better reason than he fancied something different. What now? Let's plonk a bit on d6. The knight would inevitably have to exchange itself for the c8 bishop, but how about putting the rook there instead? Although it's retreat will be cut off by the inevitable Nd4, getting at it will be harder for black.

14 Rd6 Nd4 No surprise there. Should whip the knight off immediately, with Bf4 the continuation after the likely e-pawn recapture. This would give more protection to the rook and build latent pressure against c7, but RR tries to be more sophisticated by delaying action.

15 Qd2 Qe7, 16 Bg5 f6, 17 Bh4 Rf7, 18 Qd3 Bf8

RR v Bill Armstrong after 18 ... Bf8

With the d6 rook attacked first by the queen and protected by the knight RR foolishly believes that he can afford at least one more move before taking on d4, for where can black move his queen? Qe8 and RR has Nc7, or Qd8 gives RR Rd5 and the knight is pinned. Bill soon shows this to be wrong:

We join the chess game after 10 moves.

RR v John Booth after 10 ... a6

r2q1rk1/1p1bppbp/p1np1np1/8/2PNP3/P1N1B3/1P2BPPP/R2Q1RK1 w
Decisions, decisions. Exchanging on c6 must have virtues, as does moving the rook off a1 so as not to be the firing line of blacks fianchettoed bishop. f4 to commence kingside action and to keep the enemy knight out of e5. More preparation. RR rejects all these in favour of queenside expansion.

11 b4 Ne5, 12 f3 Protecting e4 and keeping enemy pieces out of g4. However with a couple of unprotected pieces on the long diagonal RR may need to watch out for Nxf3+ with a discovered attack at some later stage.

12 ... Rc8, 13 Qb3 Qc7, 14 Nd5 Rac1 superior as it gets the rook off the diagonal and black would regret Nxc4 when white replied Nd5

14 ... Nxd5 15 cxd5 Qc3 16 Rac1

RR v John Booth after 16 Rac1

16 ... Qxb3 Computer says Nxf3+ is good for black leading either to a simplified position in which black is a pawn up, or a complex one in which black effectively gives up his queen for rook, bishop, pawn, control of the c-file and the initiative. Thus white should have played
16 Qxc3 Rxc3, 17 Bd2 kicking the rook out of his position.

Back to the game.
17 Nxb3 Rxc1, 18 Rxc1 Rc8, 19 h3 Should RR have initiated the exchange of the final pair of rooks himself? Quite possibly, but he decided that allowing black to do this would mean RR's dark squared bishop ended up on c1 preventing the black opposite number arriving on b2, whilst by covering g4 h3 would also prepare f4.

19 ... Rxc1+, 20 Bxc1 Bc8= Time to look around, but not surprised to discover that the progress in the other games is insufficient to make clear the direction the match is heading. Play on, but a plan is needed, which with minor piece endings is often far from easy. Na5, or f4 and then Na5? Queenside expansion with a4? RR alighted upon grabbing kingside space and centralising the king with intent to cramp black.

21 f4 Nd7, 22 Kf2 Kf8, 23 g4 h6, 24 Ke3 Ke8

RR v John Booth after 24 ... Ke8

Now for a little disruption.
25 b5 Nc5 Presumably not liking the pin after an exchange on b5, but this gives RR the chance to acquire the superior pawn structure.

26 Nxc5 dxc5, 27 bxa6 bxa6 28 Kd3 Silly. e5 must be better as the g7 bishop is then restricted.

28 ... Bd4, 29 Be3 Bxe3, 30 Kxe3 Kd7

RR v John Booth after 30 ... Kd7

It is well known that in a same coloured bishop ending you want your pawns on the opposite colour squares to the bishops, so g5, h4 and e5 are the candidate moves, but RR comes up with an inferior plan:

31 a4 believing that black won't want to respond a5 and allow Bb5+. Shoddy thinking, but John co-operated by trying to sit on the position.

31 ... Kd6, 32 Bc4 f6, 33 a5 Back on a black square where it belongs. The next few moves are a bit daft as both sides fail to recognise the advantage in playing g5. White's plan is to get his king to c4, push e5 to gain c5 after which Kb6 hopefully wins quicker than blacks now free movement in the centre and on the kingside. Black meanwhile seems to be adopting an I don't believe you can get in approach.

33 ... Bb7, 34 Kd3 Bc8, 35 Kc3 Bb7, 36 h4 RR wakes up. Now it is too late for g5 from black.

36 ... Bc8, 37 g5 hxg5, 38 hxg5 fxg5, 39 e5+ Kc7, 40 fxg5

RR v John Booth after 40 fxg5

As is so often the case it is a very short path from a tenable position to a hopeless one, and black now has too many weaknesses to hope to survive.

RR v Karl Lockett after 10 NxB

10 ... Qh4, 11 Qf3 Nc6, 12 O-O-O Sometimes deciding which side to castle is as difficult as choosing the files for ones rooks. I've a suspicion that in this case it doesn't make much difference objectively whether RR castles short or long, even though the resulting positions may well lead to different games. Additionally in this case black gets to choose whether castling is on the same or opposite wings.

12 ... O-O, 13 g3 Qh6+ RR had expected Qe7 here, followed by a launch of queenside and/or central pawns. After all play on a wing is usually best countered by play in the centre or by setting up a race with an attack on the opposite wing. The chosen move encourages Kb1, which white wants to play anyway, and leaves black needing to take care that his queen doesn't get trapped in the next few moves. Stockfish green lights this check, so we are in personal choice rather than weak move territory.

14 Kb1 Ne7, 15 h4 Qg6 Perhaps an attempt to offer an exchange on f5, but black is better to pose the en passant question with f5 immediately.

16 g4 f5

RR v Karl Lockett after 16 ... f5

17 exf6 gxf6, 18 h5 Qf7, 19 Qg3 d5, 20 Nc3 c5 With his own king position now slightly compromised, Karl decides its time to open his own lines of attack. RR has the choice of whether the half open file against his king will be the b-file (dxc5 bxc5) or the c-file. RR chooses the latter, Stockfish would have picked the former.

21 f4 Spent time considering Nb5, but RR couldn't convince himself that this led anywhere useful after Nc6.

21 ... Nc6, 22 Ne2 cxd4, 23 Nxd4 Nxd4, 24 Rxd4 Rac8

RR v Karl Lockett after 24 ... Rac8

25 Qg2 A multi-purpose move holding back the e-pawn because of the threat against the d, and providing additional protection against back row mates delivered on f1 and for the c-pawn. How many, if any, of these are necessary is a moot point. Probably better simply to centralise the h-rook.

25 ... Rfe8 The plan was for f5 at this point, but couldn't identify the edge after some responses including Rc4, so switched back to forward march against his king. Not a decision of which my silicon commentator approves. With the board opening up we're now at the stage of the game where the computer often thinks there is only one sensible move. Humans find more, some of which they subsequently rue.

26 g5 fxg5, 27 fxg5 Qg7 Rf8 may seem like wasting the previous Re8, but with the f-file now open doubling major pieces on it makes sense with threats of back row mates and ganging up on c2. If nothing else RR would have to give consideration to how serious the threats are. Instead he has to decide on where to place his rook - d1, d2 or g4, all of which have their attractions

28 Rd2 Rf8, 29 Qe2 The loose g-pawn is a two for one offer.

RR v Karl Lockett after 29 Qe2

29 ... Rce8, 30 Rg1 Kh8, 31 Rdd1 Rf5, 32 g6 Qh6 Had expected h6. Admittedly the final endgame with white having a supported passed pawn on g6 would favour RR, but with all the heavy artillery still on, black's first task is to get that far. Besides does black really want to take on h5?

33 Qb5 Rd8 Re7 anticipated, though the white queen still moves in.

34 Qc6 Rxh5 He does want to take the h-pawn!

35 Qxe6 Qc7 more powerful, whilst I'm told that g7+ leads to mate in 18 - won't lose any sleep missing that.

RR v Karl Lockett after 35 Qxe6

Black is struggling for a move.

35 ... Rg5 RR couldn't resist an audible "You'll regret it". Fortunately as one of the good guys Karl didn't object about this breach af etiquette. RR played

36 Qf6+ Karl regretted it, as he's losing both rooks at a minimum.