Weekend 3

[games, pgn]

Tim Dickinson writes:

Thanks to all who played, especially those who joined us for the first time this season (Aidan, Robin, Matt L, Robert) and also to Phil N who played through the pain of a toothache.

It wasn't the best weekend in Division 3, but the headline is that Oxford 1 enter the championship pool being clear top of pool A. That ch-pool effectively starts in round 7 late March, as our opponents Wessex are 99% qualified as well [it would take big Ox1 + Poisoned Pawns r7 wins to demote them – game points is the tiebreak].

In the Ox3 carnage, Sean managed 2/2 to score 66% of the team's game points over the weekend – well done to him!

I was particularly pleased to see Aidan return to Ox1 in good form, he was our other 100% man and I thought his Sunday win was well-controlled. In the game-player below, do please admire the way Aidan's king calmly walks from g1 to a1 on moves 26-33, and was almost mated there at move 43 necessitating a temporary queen sac with 44. Rh1.

Hats off also to K for converting the notoriously difficult KQ v KR ending near-flawlessly on Sunday. Do play through that game, try to ignore the dodginess of the opening, and look out for the key position Kh7 Rg7 v Kf6 Qe8 after which the win is a doddle. Well, it is if you're refreshed with a drink in your armchair at home. It is considerably harder if you've been playing for 6'30" as K had.

Oxford 1


Oxford 1 2204 Rhyfelwyr Essyllwg 1 2122
w Zakarian, David fm 2374 ½ - ½ Jones, Richard S im 2394
b Harvey, Marcus R 2434 ½ - ½ Kett, Timothy J 2220
w Rose, Matthew 2303 0 - 1 Spice, Alan P 2099
b Smallbone, Kieran D 2211 ½ - ½ Trevelyan, John 2129
w Rawlinson, Aidan M 2171 1 - 0 Brown, Tom 2083
b Dickinson, Tim R 2134 1 - 0 Smith, Olivia wf 2011
w Healey, Michael W 2088 1 - 0 Freeland, Alexander 2066
b Ehr, Jennifer 1922 ½ - ½ Fathallah, Joe 1976
5 - 3

The Saturday match featured a lot of opening ‘theory’, huge plus advantages early and some tricky endgame play on the top boards. Our top 3 were vying with each other for the prize of top disaster opening outcome – and certainly the judges had trouble splitting them. First of all, on board 1, David Z finds himself in this position after 16 moves of a King’s Gambit:

Zakarian (2374) - Jones,R S (2394) (B)
4NCL Division 2a (5.11), 23.02.2013

Here, something clearly has gone wrong, as even the weaker, sub-2000 player can see that unless there is some tactical coup here, that those 2 black bishops are worth about one white king. Fritz tends to agree, giving Black an advantage of -3.05 around about here – astonishing in a position which for the moment is materially equal.

A clear contender for Top Disaster all right, although David keeps his cool and manages to scramble to a drawn position via a series of other minus figures (½–½, 31). It might seem like a harsh call on White’s conversion ability but in all fairness there’s so much empty space on the board. It is not easy to sniff out the key significant weaknesses in this position — let’s face it, there are a lot of suspects — and exploit them. Try it yourself before turning on the analysis engine.

Kett (2220) - Harvey (2181) (B)
4NCL Division 2a (5.12), 23.02.2013

Next door a Nimzo has turned out looking poor for Marcus, who (B, to play) seems lamely to give away an exchange: 20. ... f6 21.Bh6+ Kh8 22.Bxf8 Rxf8. It does need saying though that the black square weaknesses around his king are cause for concern amongst the Twitterati – (Fritz: +1.62) – as White already threatens the uncommonly strong Bb5 after which Bxd7 will allow a march to mate by the white pieces.

This loss of the exchange, with the c-pawn thrown in soon enough as partial compensation, moves the game towards the conversion stage in the ending, as Marcus wriggles with a Knight + pawns against the Rook + pawns.

Next door on Board 3, Matt has the pleasing position seen below, the result of a Tarrasch French. He’s on move and after Black has just retreated his Nb6-c8 Fritz is calling this one to the tune of +2.55:

Rose (2303) - Spice (2099) (W)
4NCL Division 2a (5.13), 23.02.2013

A sample line here would be 17.c4 dxc4 18.Nxc4 Qf7 19.Bg5 Be7 20.Rad1 – not necessarily the best computer moves from Black, but sensible enough – and Black finds it hard to do anything constructive, like castle. As it was, Matt’s choice 17.Ba3 wasn’t bad, but lacked the directness that we have come to suspect from him in such positions. Subsequent plans adopted allowed Black to castle and by the time time trouble loomed on move 35 Black was firmly in control and his 40th move forced resignation.

Boards 1 and 3 resulted in one solitary ½ point, a net loss of the other ½ these positions merited - and Kieran’s solid draw on move 44 meant we were 1-2 down on the top board. As Marcus was entering hand-to-hand combat, it is time to look at the bottom boards which provided the match muscle. Starting with the returning Aidan, who showed spectators just where his chess thinking had gotten to, after a few years’ break: The Catalan, with the white bits. Oh dear – although this rapidly churned into a King’s Indian position where, when the crowd reached the Indian in Daventry some 4 hours later, Black seemed to be doing just that little better:

Rawlinson,A (2171) - Brown,T (2083) (W)
4NCL Division 2a (5.15), 23.02.2013

Black’s temporary pawn sacrifice has reaped some dividend, with the Na4 looking a little fragile, but needed there to defend against the incursion of pieces on b2. But, after 20.Rc6 f5 21.Bxe4 Rxe4? [Played to defend the c-pawn but that can’t be taken: 21...fxe4 22.Rxc4 Ba6 23.Rc2 Bd3 24.Rd2 Ne5 would be crushing]. Now White seizes the initiative – on such small margins in the difference between success and failure: 22.Rd1 Bb7 23.Rc7 Rxe3 24.Rdxd7 1–0

On the next board Tim won prettily after White allowed an incursion by the bishops:

Smith,O (2011) - Dickinson (2134)
4NCL Division 2a (5.16), 23.02.2013

21. ... Nb4 21.Nxb4 Bxb4 22.Re2 a3 23.Ba1 Qc6 24.e4 Qc1 0–1

And Mike’s game allowed for a ranging discussion while waiting on the main course

Healey(2088) - Freeland (1819)
4NCL Division 2a (5.17), 23.02.2013

But ... Ke7 was a blunder here and Mike exacted accurate tactical penalties – a test for the crowd.

On bottom board Jennifer was no worse for all her game and agreed a draw just in time to secure some win points, just after adjournment:

Fathallah (1976) - Ehr (1595)
4NCL Division 2a (5.18), 23.02.2013

25.c4 Ba6 26.Nxa6 Rxa6 27.c5 R6a7 28.Rb3 Rd7 29.Qe2 Qh4 30.Qf3 Qe7 31.Rfb1 Rb8 32.a3 Rdb7 33.axb4 Rxb4 34.Rxb4 Rxb4 35.Rxb4 axb4 36.Qd1 Qa7 37.Qb3 Qa5 38.h4 g6 39.Kc2 Qb5 40.Qd3 Qa4+ ½–½

So that gave us 4 ½ points and only Marcus remained for the attention of pseudorandom tweeterama – and it was as exciting OTB as it proves in PGN format.

Kett – Harvey (B)

Just before time control it’s time to see what will be left after the multiple queen promotions: 57... b2 58.Kxf7 a1Q 59.h8Q Qxa8 60.Qxa8 b1Q 61.Qa5+ ah – the start of some geometry bringing about a Q+2 v Q ending. ... Kd4 62.Qd2+ Kc4 63.Qd5+ Kc3 64.Qxe5+ Kd3 and all seems lost. A technical win with the Black king completely out of the game... ... still, Marcus carries on. 65.f4 Qb7+ 66.Kf6 Qh7 67.Qd5+ Ke3 68.f5 Kf4 69.Ke6 Qg8+ 70.Ke7 Qg7+ 71.Qf7 Qe5+ 72.Qe6 Qg7+ 73.Ke8 Qh8+ 74.Kd7 Qa8 75.f6 Qa7+ 76.Ke8

... Kg5 77.Qf5+ Kh6 78.e5??? so natural, but wrong: ... Qf7+ 79.Kd8 Qc7+ ½–½

That’s pretty neat, really.


Anglian Avengers 1 2182 Oxford 1 2199
w Merry, Alan B 2251 0 - 1 Zakarian, David fm 2374
b Player, Edmund C 2248 1 - 0 Schmidt, Jan-Peter fm 2269
w Lindner, Daniel cm 2202 ½ - ½ Rose, Matthew 2303
b Talsma, Paul A 2185 ½ - ½ Dickinson, Tim R 2134
w Munson, Shaun D 2177 0 - 1 Smallbone, Kieran D 2211
b Tozer, Philip AA 2174 0 - 1 Rawlinson, Aidan M 2171
w Taylor, Adam C 2146 1 - 0 Healey, Michael W 2088
b Gray, Mark S 2074 1 - 0 Foster, Chantelle L 2042
4 - 4

The top two boards this time around gave two gorgeous ending displays: First, David Z’s advantage seems to be dwindling down to a draw that seems emphasised by the overall symmetry of the position:

Merry(2251) - Zakarian (2374) (B)
4NCL Division 2a (6.211), 24.02.2013

... h3! Persuading White to round up that invader, at the expense of a few tempi: 44.Nf2 a5 45.Nxh3 Nc5 46.Nf2 Nxa4 47.Nd3 Nc5 48.Nxc5 dxc5 49.Bc3 Kf7 and White spies another pawn offer 50.Bxe5

50 ... Bf6! Unfortunately it is poisoned and after this tactical coup White can’t exchange pieces, but nor can his bishop now cope with the passed pawns and the control the Bf6 has on proceedings 51.Bf4 Bb2 52.e5 a4 53.e6+ Ke7 54.Be3 Bd4 55.Bc1 Bf6 56.Ba3 Kd6 57.Kg2 Kxd5 58.g4 Be7 59.Kf3 Kc4 60.Ke4 Kb3 61.h4 Kxa3 62.h5 gxh5 63.gxh5 c4 64.h6 c3 65.h7 c2 66.h8Q c1Q 67.Qe8 Qc5 68.Qd7 Qd6 69.Qc8 Kb3 0–1

On the adjoining board, Jan was always struggling with the White pieces, but seemed to be setting up a blockade of sorts, using the opposite bishops and the pawn placements. Unfortunately for him another h-pawn move set up an elegant conversion:

Schmidt (2269) - Player (2248)
4NCL Division 2a (6.212), 24.02.2013

56 ... Qc1+ 57.Kg2 h4 58.Bb3 Qb2+ 59.Kf1 Qb1+ 60.Kg2 Qb2+ 61.Kf1 Qxa3 62.Qd5 Qc1+ 63.Kg2 Qc7 64.Kh1 Be5 65.Kg2 Qe7 66.Kf1 Bc3 67.Kf2 Qe1+ 68.Kf3 Qf1+ 69.Ke4 Qe2+ 70.Kf4 Bd2+ 0–1

The top four boards became 2-2 after Matt’s rather fortunate draw in 15, and Tim’s solid draw in 18.

On board 6, Aidan was once again treating the crowd to the delights of the Catalan (a very Messi opening) and those who were still standing at move 25 were treated this this 10-move passage of play -

Move 25 (W)

Move 35 (W)

During which the White King wanders nonchalantly in to a1 – and we all awaited developments – which happened on move 36 with h3, the prelude to a king side expansion that fully stretched Black’s capacity to resist. The final moves were introduced by a witty ignoring of the threat to his queen in the diagram position:

Rawlinson,A (2171) - Tozer,P (2174)
4NCL Division 2a (6.216), 24.02.2013

44.Rh1! Rxa6 45.Rh7+ Kf6 46.Rxc7 Bb5 47.Bb3 Bc4 48.Rb2 Re6 49.Rh2 Rxe3 50.Rhh7 Rf8 51.Bxc4 dxc4 52.d5 Re1+ 53.Kb2 c3+ 54.Kb3 1–0 It’s mate or loss of Englishman’s home (followed by death)

Mike had an off day against an opponent who refused to let his youth get in the way of some sober positional play:

Taylor (1966) - Healey (2088)
4NCL Division 2a (6.217), 24.02.2013

... Ra8 and ... Rf7 may be the computer recommendations for a continuing bad rash over the black position, but (or so) Mike decides to punt in the centre: 23. ... d5 24.c5 Qg3+ 25.Bg2 Bc7 26.exd5 Be5 27.d6 b6 28.Ne2 Bxf5 29.gxf5 1–0

All of which, added to Chantelle getting lost in some tactics in a sharp Lopez and dropping two pawns while unable to prevent a trivial ending, left the Anglian Avengers 4-3 to the good and looking to whitewash the Oxford teams over the weekend. Kieran’s game chances were being hyped by Tim on Twitter, as we sped back to Oxford from Daventry, and while we were being fed a lot of technicalities about the ending, this was at the expense of a right interesting bhaji bunfight in the opening: 1.d4 d5 2.Bg5 f6 3.Bh4 Nh6 4.e3 Nf5 5.Bd3 h5

Munson (2177) - Smallbone, (2211)
4NCL Division 2a (6.215), 24.02.2013

Five moves of liquid post-modernity give us galloping bishops, dim-rim Knights and weakening pawn structures – all this for Black to be 0.88 to the good.

50 moves later we were still in the technical conversion phase, with

White to move

56.h6 Rd8 57.h7 b3 58.Rh2 a4 59.h8Q Rxh8 60.Rxh8 Kc7 (whew) 61.Ra8 b2 62.Rxa4 b1Q and now it’s Q v R with a pair of kings thrown in 63.Rf4 Kd6 64.Kf6 Qg1 65.Rf5 Qg4 66.Rg5 Qe6+ 67.Kg7 Ke7 68.Kh7 Kf8 69.Ra5 Kf7 70.Ra7+ Kf6 71.Rg7 Qe8 72.Rg1 Qh5+ 73.Kg8 Qf7+ 74.Kh8 Qa7 75.Rg4 Qc7 76.Rg2 Qc8+ 77.Kh7 Qb7+ 0–1

Oxford 2


Oxford 2 2011 Anglian Avengers 2 2048
w Cornford, Matthew 2026 0 - 1 Scott, Gordon W 2038
b Duggan, Christopher 2025 0 - 1 Myall, Ivan J 2078
w Nandi, Robin J 2066 0 - 1 LeMoir, David A 2064
b Burrows, Nick 1938 0 - 1 Wallis, Ian J 2021
w Ludbrook, Matthew 1996 ½ - ½ Gregory, Stephen J 2052
b Scott, David A 2017 1 - 0 Cook, Michael P 2039
1½ - 4½

Nothing much went right for Oxford 2 on the day, against an Anglian team bent on Avenging.

(Avenging what? You might ask. What did we ever do to the Anglians? Indeed.) Let’s have some highlights.

On top board, Matthew took a wrong view in the Grunfeld, and his Kf1 in the opening, when 0-0 was possible, was taken with a view to pushing that h-pawn to High Wycombe and back. It eventually did its job, but by move 25 the chips were on the cards:

Cornford - Scott (2038)
4NCL Division 3s Daventry Court Hotel ENG (5.55), 23.02.2013
25 ... Rxd4 26.cxd4 Rc1 27.Qa8+ Bf8 0–1

This was one of the later results, as next door Dugg had already left the hotel to jog over to see how the firsts were shaping out against the Welsh lads after his attempt to repeat the opening from the previous weekend was blown off course by a simple error in what was already a poor position:

Myall (2078) - Duggan (2025) (B)
4NCL Division 3s Daventry Court Hotel ENG (5.56), 23.02.2013

The threat is Qd2, so ... Kh8 is called for. Instead 10. ...f6 11.exf6 Qxf6 12.g5 spells the end (1-0, 21)

Nandi - LeMoir (2064) (W)
4NCL Division 3s Daventry Court Hotel ENG (5.57), 23.02.2013

Passing this board here I wondered if hacking on c6 followed by Be3 and an extended grovel was an option. That this seems a plausible suggestion – those black bishops would be fierce – suggests Black is comfortable already – and after 15.Nd2 Bc5+ material is about to exchange hands on unequal terms 16.Kh1 Bf2 17.Re2 e3 18.Nf3 Bd3 19.Rxe3+ Bxe3 20.Bxe3 b5 21.Bb3 0–0–0 22.Ng5 Rde8 23.Bf4 Re2 24.Ne6 Re8 0–1

What light there was to shine on the Oxford team came from the lower boards where Matt drew what looked like a very technical discussion in the closed Sicilian, and Dave Scott managed this technical but elegant conversion of the extra pawn. Do not rush, saith the Gods of the Endgame – and that’s what Dave didn’t do:

Cook (2039) - Scott (2017) [B]
4NCL Division 3s Daventry Court Hotel ENG (5.60), 23.02.2013

29.... a5 30.Rc6 Kf7 31.Kg4 Ke7 32.Kf3 Kd7 33.Rc5 c6 34.a4 Kd6 35.Rg5 b4 36.Rxa5 b3 37.Rh5 Re8 38.Rh1 c5 39.Rd1+ Kc6 40.Rc1 Kd5 41.a5 c4 42.a6 b2 43.Rd1+ Kc6 44.a7 Kb7 0–1

Final word goes to the Avenger’s and ex-City boy Ian’s win against newcomer to Oxford Nick. A pawn down after the middle game transition to the ending, it’s time for Black to take a punt:

Wallis (2021) – Burrows
4NCL Division 3s Daventry Court Hotel ENG (5.58), 23.02.2013

28.c3 Ndf5 29.Nxf5 Nxf5 30.Bc2 Nd4 31.Bd1 Nf5 32.Bb3 Rd6 33.d4! which sets the seal on the result as Be6 will follow any capture by the e-pawn (1-0, 47)


Sussex Smart Survivors 1 2030 Oxford 2 2030
w Therrien, Alex L fm 2258 1 - 0 Ludbrook, Matthew 1996
b Grant, David J 2084 ½ - ½ Scott, David A 2017
w Bucher, Grant W 2039 1 - 0 Cornford, Matthew 2026
b Cumming, Rhys 2001 1 - 0 Stembridge, Ed 2050
w Payne, Matthew J 1952 0 - 1 Nandi, Robin J 2066
b Brewer, Callum D 1850 0 - 1 Duggan, Christopher 2025
3½ - 2½

From early on the bottom boards were looking good, and we got two points in the bag just after time control: Robin was sweeping all before him – those bishops again [is it me, or are bishops getting stronger now that the Pope is resigning?!] – and Fritz even suggests that the ‘shot’ that White uncorks here isn’t the best and that a quieter move (Qc2) would leave him better placed:

Payne,Matthew J (1952) - Nandi,Robin J
4NCL Division 3s Daventry Court Hotel ENG (6.69), 24.02.2013

16.Qxa7 0–0 17.Qd4 Ra4 18.b4 Re8 19.Qc5 Qxc7 and soon the pressure on the white position yielded a piece which was duly converted (0-1, 49)

The score became 2.5 – 0.5 after Chris D converted a solid positional crush on bottom board and Dave Scott’s earlier draw on board 2 came to fruition after a repeat on move 23 in a complex, hardly drawn position, but one which was perhaps favouring Black, albeit that he was a pawn adrift.

A long wait ensued while the crowd – real, imagined, or virtual – wondered if Ed S would make it to time control unscathed – having left himself 6 or so minutes for 24 moves.

(Ans: No – 0-1, 30)

2.5 – 1.5

And could Matthew C hold this position?

Bucher,Grant W (2039) - Cornford,Matthew (W)
4NCL Division 3s Daventry Court Hotel ENG (6.67), 24.02.2013

27.Be2 Rbd8 28.Bg4 Re7 29.c4 Na7 30.Bf2 h5 31.Bf3 Red7 32.d5 exd5 33.exd5 ... “no” (1-0, 54)

2.5 – 2.5

So, then – could Matt L hold this one on top board?

Therrien (2258) - Ludbrook (1996) (W)
4NCL Division 3s Daventry Court Hotel ENG (6.65), 24.02.2013


This game had seemed dynamically equal for long portions but with the disappearance of pieces to leave the array in the diagram the only soldiers standing, the logic of countervailing pawn phalanxes began to come under (first) visual and (then) analytic scrutiny. In plainer English, white’s king side majority was more mobile than the lot on the other side, while the Bishop was so dominant it might even now be taking its place in a Vatican conclave. In fact, that b1-h7 diagonal is so slippy it could easily be acting as a portable runway runway to airlift the powerful prelate to Rome...

45.Ke3 Ne8 46.h4 Nd6 47.g5+ Kg7 48.f5 and that b-pawn can be snaffled but only at the expense of the Black king hxg5 49.hxg5 b4 50.Kf4 Nc4 51.f6+ Kf7 52.Kf5 Nd6+ 53.Ke5 Ne4 54.Bxe4 dxe4 55.Kxe4 c4 56.Kd4 c3 57.bxc3 1–0

A brave fight by the lads, just failing.

Oxford 3


Oxford 3 1792 Anglian Avengers 3 1917
w Hayward, Philip T 1942 0 - 1 Hopkins, Phil 2032
b Neatherway, A Philip 1859 0 - 1 Lamont, Richard 2018
w Terry, Sean 1908 1 - 0 Peck, Silas 1882
b Henbest, Kevin B 1729 0 - 1 Fogg, Martin 1942
w Irving, Alastair 1698 0 - 1 Hall, Antony C 1970
b Asatryan, Robert 1618 0 - 1 Woodcock, Keith D 1658
1 - 5

A dismal day for the Oxford 3 team, which was largely outgraded on the bottom boards and struggled to avoid blundering on the top three. A selection appears below!

Phil’s error was both instructive, and immediately deadly:

Lamont, - Neatherway
4NCL Division 3s (5.166), 23.02.2013

14 ... c5?? 15.Bxf7+ Kh8 16.Bxe8 Rxe8 17.Nxf6 Nxf6 18.Qe5 b4 1–0

Poor Phil worked his way tirelessly to invade the Black position, only to be hit by the dual calamity of having an immediate win, and a passing spectator in charge of a Twit gun:

(sorry, Phil)

Naturally this ended up 0-1, 56

We arrive late for the match - our car arrived at 1.55pm and me / Alistair / Kev check in ("Patzer see a check-in, ...") before arriving at the board. It is only when Alastair arrives that his opponent is revealed to be slightly shy, and has a preference against talking - thus ruling him out of playing Alastair! Fortunately, I haven't arrived either - as if ... - so a last minute switch is agreed.

Terry(1908) - Peck(1713)
4NCL Division 3s (5.167), 23.02.2013

I managed to mismanage the opening against my opponent and I was in some trouble until, in this position, he took the wrong option. Exchanging is fine, as is ... Qg5, but 13 ... Qh2 simply allows 14. e5! With a reversed position assessment and an easy attack. (1-0, 25 - time)

The final 3 boards saw Kev drop a piece, Alastair refuse many a perpetual before getting mated around time control, and Robert A allowing too easy an attack against his king.


Bristol 2 1899 Oxford 3 1791
w Ponter, Ian P 2005 1 - 0 Burrows, Nick 1938
b Hassall, Mark I 1943 0 - 1 Neatherway, A Philip 1859
w Hendy, Jerry 1834 0 - 1 Terry, Sean 1908
b White, Martin J 1828 1 - 0 Henbest, Kevin B 1729
w Meadows, Michael 1898 1 - 0 Irving, Alastair 1698
b Taylor, John Paul J 1890 1 - 0 Asatryan, Robert 1618
4 - 2

Sunday saw the same battling spirit, a point more, but no match points. Highlights:

Alastair found his king under sustained siege and here is the denouement:

Meadows,Michael - Irving,Alastair
4NCL Division 3s (6.159), 24.02.2013

19.... Nxf4 20.Nxe6 Nxe6 21.Rxe6 Ra7 22.Rde1 Qd8 23.Rf6 Rf8 24.Bc4 1–0

Phil found the direct route handy when Black decided to sac an exchange to win a few pawns on the queen side:

Neatherway (1859) - Hassall (1943)
4NCL Division 3s (6.156), 24.02.2013

40.Qf5+ Kh8 41.Qe6 Nh5 42.Rg6 Bf8 43.Bxh6 Qxd3 44.Bxg7+ 1–0

While I found the direct route to the White king was via the g-file:

Hendy - Terry (1908) (B)
4NCL Division 3s (6.157), 24.02.2013

24. ... h5 25.Qxg7 0–0–0 26.f3 Rdg8 27.Qf7 Rxg2+ 28.Kh1 Here, I stopped to think; my original plan was ... Rg3 xf3 which seemed good, but I was attracted by Qxc2, expecially when I saw the mate which happened on the board (but not before sweating and nearly missing the Rook pin!) 29.Rac1 Rh2+ 30.Kg1 Nxh3# 0–1

I felt good about this, until as I signed the sheet, I suddenly thought with a clutch – “what if 29. Rfc1?” The heart raced a little before I realised that this gap in the analysis was, quite fortunately, not relevant...

28 Feb 2013