After 21 moves RR has ceded his advantage as white, having gained space at the expense of an isolated pawn, and being behind in terms of centralising his rooks.

RR v Neville Layhe after 21 ... Rfe8

3rr1k1/1pq2pbp/p1b2np1/2B1p3/1PP1P3/P4B2/3NQ1PP/R4RK1 w
Time for a not very subtle threat:
22 Qe3 Qd7 Neville goes straight for penetration down the d-file. Nd7, with the threat of b6 would have been ghastly for white - the best he would seem to be able to hope for would be disruption of his pawn structure with the weak foot soldiers being picked off one by one.

23 Ra2 Qd3, 24 g4 Qxe3+, 25 Bxe3 Rd3, 26 Kf2 Red8
27 g5 Nh5 diagram right

RR v Neville Layhe after 21 ... Nh5

Kicking the rook with Be2 looks natural, but if he plays Rc3 do I have anything better than dancing the knight between b1 and d2 for a draw.? Perhaps I should try Be2 first in case he chooses to retreat along the file, but RR elected to take the knight immediately, hoping for chances in the endgame.

28 Bxh5 gxh5, 29 Rc1 Ba4, 30 Ke2 Bb3
Rb2 seems to set up another draw by repetition so
31 Raa1 Ba4, 32 Nf3 Rb3, 33 Nd2 Much though I'd like to play Ne1 I need to stop the check picking up the h-pawn. Did at least spot that Rd1 loses to Rxe3+.

33 ... Rb2, 34 Rc3 Offering draw by repetition chances with Rcb1 may be objectively better, but RR is naturally greedy.

34 ... Bc2, 35 Kf3 Ba4, 36 Nf1 With his rooks on a1 and c3 RR is petrified of an f-pawn shove by black, but rather than offering to repeat moves he again looks elsewhere.

36 ... Bd1+, 37 Kg3 Re2, 38 a4 intended as a dual purpose move, removing a4 as a retreat square for the bishop and preparing a later b5.

RR v Neville Layhe after 38 ... Rc2

38 ... Rc2 diagram left

Time for some exchanges, but possibly not the ones Neville was hoping for.
39 Rxd1 Rxd1, 40 Rxc2 Rxf1, 41 Rd2 Grabbing the d-file seems natural, though my silicon friend says shove the queenside pawns.

41 ... Bf8, 42 Rd8 Kg7, 43 b5 c5 and Bd2 must be better

43 ... axb5, 44 axb5 Re1, 45 Rd3 Rb1, 46 b6 Rb4
47 Rc3 Be7, 48 Kf3 f6, 49 gxf6+ Kxf6, 50 c5 Ke6
51 c6 bxc6, 52 Rxc6+ Kd7 reaching the diagram right which shows a simple tactics test for white which RR failed.

RR v Neville Layhe after 52 ... Kd7

53 Rh6 (53 Rc7+ Ke8, 54 Rxe7+ Kxe7, 55 Bc5+ is the easy win)

Steve Lloyd v RR after 16 Rc1

2r1r1k1/p2qbppp/1pb1pn2/8/2PP4/P1NB4/1B3PPP/2RQR1K1 b
16 ... Red8, 17 Qe2

Big think. Can I take the d-pawn now given that a knight move will be the obvious response? Yes, I think I can.

17 ... Qxd4, 18 Nd5 Qc5, 19 Bxf6 Bxf6, 20 Qh5

I'm guessing we'd both seen this far (right) ...

Steve Lloyd v RR after 20 Qh5

... but that Steve hadn't predicted

20 Bxd5 which conveniently frees up e7 as an escape square for my king as well as setting up an attack of my own.

21 Qxh7+ Kf8, 22 cxd5 Qxc1, 23 Rxc1 Rxc1+, 24 Bf1 Rxd5, 25 Qe4 Now g6 and Rdd1 adds a bishop to the material deficit, but for reasons which are not entirely clear RR goes for the a-pawn relying on finding a mating attack or queening an early alphabet pawn instead.

RR v Rich Wiltshir after 9 Qc2

r1bq1rk1/ppp2pbp/3p1np1/2nPp3/2P1P3/2N2N2/PPQ1BPPP/R1B2RK1 b
It seems to me that pawn moves are always the most difficult to judge since they cannot be reversed later should they prove poor. Moreover they are so often double-edged, serving the purpose for which you are attracted to the move in the first place, yet often leaving a weakness. Which is more important? Rich was faced with this dilemma in the diagrammed position. His knight on c5 is in danger of being dislodged by a white b4, yet playing a5 himself concedes the b5 square for a white knight. Rich passed on the pawn push, electing instead to prepare the thematic f-pawn push:

9 ... Ng4 with intent to retreat to h6 further supporting f5 should white play h3. However I suspect Nh5 is more normal here. The knight cannot immediately be dislodged with g4 because of black's light squared bishop, and in some lines may later arrive on f4.

10 b4 Na6, 11 a3 f5, 12 Bg5 Qe8, 13 c5 fxe4
14 Nxe4 dxc5 15 bxc5 15 b5 Nb8 superior, but I confess was not even considered

15 ... b5

RR v Rich Wiltshir after 15 ... b5

With d6, c6, Rb1 and Stockfish's favoured a4 all available, RR is spoilt for choice, but unerringly he alights on one of the poorer options, though is still in control.

16 cxb6 axb6, 17 Rac1 Nc5, 18 Nxc5 bxc5, 19 Qxc5 RR has conceded much of his positional advantage but now has a material one, if only a single pawn. However at our level such advantages tend to be more enduring, so from a practical point of view it's not all bad. In the simplification that follows RR gadually loses the edge, but not the pawn.

19 ... e4, 20 Nd2 Qe5, 21 Bxg4 Bxg4, 22 Be3 Rfd8
23 Qxc7 Qxd5, 24 Qc6 Be2, 25 Rfe1 Bd3 ,26 Qxd5+ Rxd5, 27 Nc4

RR v Rich Wiltshir after 27 Nc4

Black's bishop pair constrain the movement of white's rooks so that white needs the extra pawn for equality. White does have a pesky knight though, protecting the a-pawn, preventing black's dark-squared bishop arriving on b2, and threatening to fork the black rooks from b6. Rich decides that this is enough to order the execution of said knight, but his choice frees up white's game.

27 ... Bxc4, 28 Rxc4 Rxa3, 29 Rxe4 Ra1, 30 Rxa1 Bxa1
31 Kf1 Rb5, 32 Ke2 Kf7, 33 Kf3 Rf5+, 34 Rf4 Ke6
35 g4 Rb5, 36 h4 Bf6, 37 Re4+ Kd5, 38 g5 Be5
39 Ra4 Rb7, 40 Kg4 Rc7, 41 Ra5+ Ke6, 42 Rxe5+ 1-0

Ifan Rathbone-Jones v RR after 10 d5

rn1q1rk1/p1pp1ppp/1p2p3/3P4/2P1b3/2B1PN2/PP3PPP/R2QK2R b
The game started life as a Nimzo-Indian, but this is I suspect not a mainline position. White is threatening Qd4 forking mate on g7 and the bishop on e4. What do I do? Think straight would help. After exd5 the bishop is temporarily protected. f6 would pre-emptively block the mate threat. Bg6 perhaps, though that asks for the h-pawn to march. So RR comes up with c5, not a bad move, but played with the wrong motivation leading his though processes into the wrong channels. Is he really going to allow the d-file to be opened and perhaps permit the exchange of queens? Unnecessary defensive moves simply help the attacker. Indeed the threat of the bishop will need to be countered by f6 sooner or later, so h6 is setting up holes around the king. exd5 seems much more natural, creating a queenside majority with obvious attractions for later play.

12 Rh3 exd5, 13 cxd5 d6, 14 Nh2 Qd7 Again RR is seeing a big attack where none exists, though it would help had he not played h6 earlier. Now he comes up with a slow plan to remove the queens, but all it does if give time for white to mass his forces effectively. Qe7 would have been much more flexible, protecting the bishop, ready to give sideways protection to g6, maintaining protection for f6 when played, and leaving d7 for the knight. 15 Ng4 f6, 16 Rg3 Kh7, 17 h5 Qf5 Continuing with the unnecessary queen exchange plan

18 f3 Bc2, 19 Qd2 Qd3, 20 e4 Qxd2+, 21 Bxd2 An unexpected recapture. King takes keeps the bishop on the long diagonal and allows the a1 rook easy access to the h-file.

Ifan Rathbone-Jones v RR after 21 Bxd2

What are white's threats? Ne3, hit bishop, then Nf5 forking g7 and d6. Or Nxh6 gxh6 followed by Rg6.

21 ... Rg8 Played as a counter to the latter threat, though now white can play Bxh6 with impunity as gxh6 would lead rapidly to mate. Instead
21 ... f5 would have been a whole new game, with black having attacking as well as defensive options. The gameline led to the loss of the d-pawn:

22 Ne3 Ba4, 23 Nf5 Be8, 24 O-O-O Nd7, 25 Rh1 Ne5, 26 Nxd6 after which black never got close to equality. For those who like watching late resigners resign late, the remaining moves were

From which introduction you will gather that RR's final game at Crewe involved an early exchange of queens, with David Hall co-operating in this state of affairs as black.

1 Nf3 g6, 2 c4 Bg7, 3 d4 d6, 4 e4 e5
5 dxe5 dxe5, 6 Qxd8+ Kxd8

RR v David Hall after 6 ... Kxd8

White must be happy here. Black can't castle, though quite possibly wouldn't want to, and the blck bishop on g7 will take a while to get into the game or find a way of exchanging itself. Yet black may target establishing a knight on d4 which will be immovable without allowing black a passed pawn and a freed bishop. Plenty of options, plenty of life. There are too many alternative lines to be considered for a concise write-up, so instead I've decided to go for sparse commentary.

7 Be3 Nd7 A surprise to RR giving the obvious attractions of the d4 square.

8 Nc3 c6, 9 Rd1 f6, 10 c5 Bh6, 11 Bc4 Bxe3 12 fxe3 We are warned of the weaknesses of doubled pawns, but RR is hoping that his will be compensated for by having one and a half open files for his rooks to operate on.

12 ... Ke7 13 Na4 I'm sure RR found a good reason at the time for choosing this rather than b4, but it now eludes me.

13 ... Nf8 Black too can find strange knight moves.

14 Ke2 Be6, 15 Bxe6 Nxe6, 16 b4 Nh6, 17 Nd2 Nf7

RR v David Hall after 17 ... Nf7

RR wanted to use the f-file, so rather than allowing a big swap on the d-file with Nc4 he continues with his outline plan.

18 Rhf1 Rad8, 19 Rf2 Nfg5, 20 Nc3 h5, 21 Rdf1 Rhf8
22 h4 Nh7, 23 a3 Rd7, 24 Nc4 Rfd8 Now Nd6 would be undermined immediately by b6, whilst the knight on h7 prevents effective use of the f-file.

25 Rc1

RR v David Hall after 25 Rc1

With both players looking to win, black now sacs a pawn to penetrate white's position, an echo of the Neville Layhe game in round 1 except that Neville needed no sac to get in. Will RR again be successful in pushing the enemy out?

25 ... b6, 26 cxb6 axb6, 27 Nxb6 Rd2+, 28 Ke1 Rxf2
29 Kxf2 Rd2+, 30 Ne2 Ra2, 31 Rc3 both a4 and Rxc6 are better. White has the passed pawn but blacks pieces will start circling the white king, so white needs his threat to materialise more quickly.

31 ... f5, 32 a4 Nf6, 33 Rxc6 Ng4+, 34 Ke1 Nxe3
35 Nd5+ Nxd5 The natural way out of the check fork, but
35 ... Kd7, 36 b5 Nxg2+ allows black to bring his king towards the queenside and pick up a couple of pawns, creating equality though a dynamic position in which either side can go wrong. The chosen line however creates a position which is whites to lose, which he doesn't. 36 exd5 Rxe2+ A "clever" trick which white is more than happy to be on the receiving end of as it leaves him with three connected passed pawns and a king able to protect against black pawn expansion.

37 Kxe2 Nd4+, 38 Kd3 Nxc6, 39 dxc6 Kd6, 40 b5 e4+
41 Kd4 Kc7, 42 a5 Kc8, 43 b6 Kb8, 44 a6 1-0

1 e4 d5, 2 exd5 Nf6, 3 Nc3 Nxd5, 4 Nxd5 Qxd5 If you don't know that RR plays this stuff, where have you been the last 20 years?

5 Qf3 c6, 6 Bd3 e5, 7 Qxd5 cxd5 black has the centre.

Derek Price v RR after 7 ... cxd5

8 c3 f5 Nc6 is the normal developing move, but RR goes for a bigger centre, nervous that he may be overextending himself.

9 Ne2 Bd6, 10 Bb5+ Nc6 Bd7 better. If white exchanges blacks king is already on the march towards supporting the centre, if not what was the check about?

11 d4 0-0, 12 0-0 e4, 13 Bf4 Bxf4, 14 Ng4 Ne7
15 Ba4 g5, 16 Ne2 f4

Derek Price v RR after 16 ...f4

Those black pawns may look impressive, but a well-timed c4 will burst the ballon. It definitely feels more like a position from a Wednesday night kick-about than a top division clash.

17 Rac1 Be6 lacking the courage for Bg4

18 Bb3 b5 Putting paid to white's attempts to play c4. Sometimes if you prepare too carefully you never get the move in.

19 a4 a6, 20 axb5 axb5, 21 Ra1 Nc6 aiming to arrive on c4, though more forceful would have been a king advance ready to help his pawns push the enemy off the back of the board.

22 f3 Na5 At this point RR is a little concerned about the exchange sac Rxa5 followed by fxe4, but this proved not to be on Derek's agenda.

Derek Price v RR after 22... Na5

23 Bc2 Nc4, 24 Rfb1 exf3, 25 gxf3 Nd2, 26 Rbd1 Nxf3+ winning a pawn with check is always nice. Now to see whether RR can complete a win without any significant blunders.

27 Kg2 Nh4+ Stockfish recommends taking a pair of rooks off on a1, allowing the knight use of the superior d2 square. RR is nervous of allowing an enemy rook into his position though.

28 Kf2 Kg7, 29 Re1 Kf6 Inaccurate as it allows black to grab the h-pawn. Not the most important pawn on the board, but Bf5 offering the exchange of bishops is clearly better.

30 Bxh7 Nf5, 31 Bxf5 Bxf5, 32 Ng1 g4, 33 Ne2 Kg5
34 Ng1 Be4, 35 Kf1 Rh8, 36 Ne2

Derek Price v RR after 36 Ne2

In recent moves white has done little more than sit and say you can't find a way in, allowing black time to mass his pieces prior to an "Oh yes I can attack". Time to abandon control of the a-file.

36 ... Rxa1, 37 Rxa1 Rxh2, 38 Rd1 g3 rejecting the simplification
38 ... Rxe2, 39 Kxe2 Bf3+ leaving a trivial pure pawn ending for black to win.

39 b3 Kg4, 40 Ng1 Bc2, 41 Ra1 f3, 42 Nxf3 Kxf3
43 Kg1 Be4, 44 Re1 Kf4, 45 Kf1 Rh1+, 46 Ke2 Rxe1+
47 Kxe1 Bd3 0-1