...and over to me

About a week ago, I set you this position:

First of all, the significance: this position is from the game Pietrzycki v Rudd, from the Devon v Somerset match. The score in the other fifteen games was 7½-7½, so the match was to hang on the result of this game.

The crucial factors of this position, as I see them, are the following:

1) The black bishop is both a better piece, and a currently better-placed piece, than the white knight.

2) The black pawns are far enough advanced that they might have a decent chance of queening; the white ones are not.

3) The white rook is somewhat awkwardly placed, although it does have the good point that while it's on a4, black always has to watch for the possibility of white's playing a3.

So my take on the position is that black has a definite advantage. Is it enough for a win? My gut feeling would be probably not, but in practice it's difficult for white to defend. Among my thoughts while playing this position was that while R+B v R is a theoretical draw, it is not an easy draw, and at 40/100 + G/20, I'd fancy my chances of winning it.

In the actual game, white played 49.Ne5, to which I responded with the natural 49...Bg7. Now white missed a trick here: after 50.f4!, it is hard to see how I make progress. (50...Rxf4 51.Nd3+ Kb5, for example, can be met by 52.Rxb4+ and that's an immediate draw.) In practice, I'd have probably gone for 50...gxf3+ 51.Nxf3 Rc4, and tried to round up the a-pawn. We would have probably ended up in R+B v R at some point.

But white's actual move was 50.Ng6?!, after which white has made things rather more difficult for himself. 50...Rd2 meant white had to deal with threats to f2 and a2: the way to deal with these was probably 51.Kg3, reaching a similar position to the last line. Instead, 51.Nf4? probably pushed the position over the edge.

After 51...Kb5 52.Ra8 (52.Ra7 comes to much the same thing after 52...Be5 53.Ng6 Bd4), I then played 52...Be5 to stop the king's coming up to snaffle the g-pawn. 53.Kg3 would have been met with 53...Rd4 54.Rf8 Rxf4! 55.Rxf4 Ka4, and the black pawn easily wins the race.

(The full analysis runs 56.Kxg4 Bxf4 57.Kxf4 Ka3 58.Kg5 Kxa2 59.f4 b3 60.f5 b2 61.f6 b1=Q and now 62. f7 Qb4/b8 and 63...Qf8, or 62.Kh6 Qf5 63.Kg7 Qg5+ 64.Kf7 Kb3 and white is stuck: Ke7/f7 do not threaten to advance the pawn, and Ke6 followed by f7 allows ...Qd8.)

So white played 53.Ng6, and the game continued Bc7 54.Rc8 (what else? ...Ba5 was a serious threat whatever) Bb6 55.Ne5 Rxf2+ 56. Kg3 Rxa2 57.Kxg4. I'd managed to emerge a pawn up, and now set about shielding the b-pawn from any white pieces thinking of coming back to defend. 57...Rd2 prevented the knight's passage back to d3 (note that 58.Nc4? fails to 58...Rd4+). Thus white had to spend a tempo on 58.Kf3, after which 58...b3 59.Nc4 Rf2+ gave him an unwanted decision:

If white now plays 60.Kg3 (60.Kg4 is not all that different, except that it doesn't attack the rook.), then I play 60...Bc5, and the knight has nowhere good to go to escape the attack. Wherever it goes, I can play 61...b2. An amusing line is 60.Kg3 Bc5 61.Ne5 b2 62.Rb8+ Bb6 63.Nd7 (or 63.Nc4 Kxc4 64.Rxb6 Rd2 and my king snakes in to shepherd the pawn home) Rf6!, and the pawn queens.

So instead white tried 60.Ke4 Bc5 61.Kd5. Unfortunately for him, I'd already seen the answer to this: 61...Rf5+ 62.Ke6 Kxc4! 63.Kxf5 b2, with no way to stop the pawn queening. The game ended a few inconsequential moves later.


So in the end, I won the game and Somerset thereby won the match. (Yes, I play for Somerset, not Devon. It's because I lived in Somerset for twenty years before moving across the border.) Should I have won the game? Well, I suspect not. I get the feeling some of the contributors to the previous post would have put up a more tenacious defence than my actual opponent did.

Hastings update

I'm currently on 1½/2 at Hastings; my first-round victory against Chkhaidze got a mention in the daily commentary.

My second-round game featured two recent winners of the Brilliancy Prize; I won it in last year's event, while Simon Williams won it the year before. Expectations were high going into the game; we did not disappoint.

Unfortunately, chesspublisher fouled up when trying to process this game. It's available on the commentary link, though.

Opening Concepts: Queen's Gambit, Chigorin

The Queen's Gambit, 1.d4 d5 2.c4, has at its heart a simple strategic idea, that of exchanging the white c-pawn for the black d-pawn and then establishing a centre with a later e2-e4. There are a number of ways black can meet this idea, but they tend to fall into a small number of distinct categories:

1) Preventing white from carrying out this plan in the first place by defending the d5 pawn with a pawn: this is mostly the province of the Queen's Gambit Declined, 2...e6 3.Nc3 Nf6/Be7, although there are some lines of the Slav, 2...c6, which could also fall into this category.

2) Allowing white to carry out his plan, but only at the cost of a pawn: this idea arises in lines of the Slav, 2...c6, and semi-Slav, 2...e6 3.Nc3 c6; this usually requires white to consciously play one of the gambit lines, otherwise the play will mutate into cases 1) or 3).
3) Allowing white to carry out his plan, but exploiting the weakness that arises as a consequence of it, namely the semi-backward and somewhat weak pawn on d4. This is the strategy of the Queen's Gambit Accepted, 2...dxc4, and the Chigorin, 2...Nc6 (see the diagram on the left).

So what are the ideas behind the Chigorin? Well, two stand out immediately: the idea of playing ...dxc4 followed by capturing on d4, and that of playing ...e5 thanks to the support from the knight on c6. For example, if white plays 3.Nc3, black can reply with 3...dxc4, and then 4.d5 Na5 renders it rather tricky for white to get the pawn back, whereas 4.e3 allows black to immediately break in the centre with 4...e5.


The obvious counter to both those ideas is the development of the other knight with 3.Nf3, which is what I played against Tim Seymour in our game at Coulsdon. His response was to renew the threat against the d4 pawn with 3...Bg4, after which I played 4.cxd5. Now 4...Qxd5?! would allow easy development of the white pieces with gain of tempo after 5.Nc3 (this is generally applicable to the Chigorin, incidentally: if white can play cxd5 Qxd5 Nc3 without the knight's being pinnable with ...Bb4, he's probably got a good game), so Seymour played 4...Bxf3, after which I had a choice.

I could either play 5.gxf3 Qxd5 6.e3, after which 6...e5 leads to double-edged play, or I could play 5.dxc6 Bxc6 6.Nc3 e6 7.e4. This is what happened in the game, giving me the position shown above here.


The pawns on d4 and e4 form an impressive classical centre, but they are easily attackable, and this informs black's choices over the next few moves: 7...Bb4 8.f3 (the only reasonable way to defend the e4 pawn; 8.Bd3? would drop the d-pawn) Qh4+! 9.g3 Qh5 (now the f3 pawn is another pawn I have to worry about) 10.Be3 0-0-0 (note that the rook is developed straight into an attacking position against the d4 pawn) 11.Bg2 f5, and we have the position above. Notice that my impressive centre is still there, but black has developed nearly all his pieces to squares where they attack it directly or indirectly. (The knight will be coming to f6 to attack the e-pawn.)

The position here is certainly full of tension, and it's a shame that the actual game was then a bit of a let-down; just a few moves later, black made a blunder which converted his temporary sacrifice of the e-pawn (he has no good way to defend it after I play 12.Qb3) into a permanent one, and I won in fairly short order in the endgame.

A game from the 4NCL

This game, a win by me against a Scottish IM, proved to be critical in Bristol 1's 4NCL match against Barbican 2 on Sunday; we ended up winning 4½-3½.

Coulsdon Christmas Internationals

Next week, I will be out of the area, at an international chess tournament. This is not unusual for me. What is unusual, however, is that I am not attending as a player, but as an arbiter.

The event in question is the Coulsdon Christmas International; the CCF run three international events a year, and a number of players have got title norms or FIDE ratings from them, as you can see here.

There are two points of North Devon interest here. The first is my attempt to get myself a FIDE Arbiter norm. The second is the attempt of Roger Hutchings, who has emerged from retirement this season, to get himself a FIDE rating. To get a full rating from the event, he has to get at least one point with John Torrance also getting at least one point. If John Torrance fails to score one point, Roger can still get a partial rating, but he then needs to score at least one point against the rated players.

Title Decider?

The game between me and Peter Marriott is usually the one that decides a Barnstaple internal tournament these days. The last time he beat me, he went on to win the summer tournament. Ever since then, I've had the upper hand, but the games haven't been easy.

Tonight's offering had all the usual levels of excitement and insanity. Neither of us is known for orthodox handling of the openings, and my 6.g4!? was about par for the course.

Opening Concepts: Hippopotamus

There are a number of chess openings that fall into the category of "universal systems"; sets of moves that can be played against most reasonable setups by the opponent. The Stonewall Attack, featured in the Matoewi-Phillips game here, is one such, and another is the Hippopotamus.

The position on the left here, taken from my game against David Grant at the Wellington College International, is a typical Hippo position. It started off 1.e4 d6 2.d4 g6 3.Nc3 Bg7 4.Be3 a6 5.a4 b6 6.Qd2 Bb7 7.f3 Nd7 8.g4 h6 9.h4 h6 10.Nge2, reaching the diagram, after which I replied 10...Ne7. Those ten black moves, in some order, constitute the Hippopotamus. Note that there are some subtleties in my move-order:

a) I started off with ...d6 and ...g6 because those don't immediately commit me to a Hippo. Had white played 3.c4, I would have changed tack with 3...Nf6 4.Nc3 Bg7, transposing to a King's Indian. Only once the knight had gone to c3 did I firm up my choice of opening.

b) I played 4...a6 for similar reasons: if white does not play 5.a4, I will not immediately rush to commit my b-pawn; it may be better left on b7 or advanced straight to b5 in some lines. Once white had played 5.a4, though, I needed to play 5...b6 in order to prevent his crippling my structure with 6.a5. (Not that this is necessarily the best move, but it cuts across my plans.)

c) I made sure to play 9...h6 before 10...Ne7 because, with the white queen and bishop lined up on the diagonal, the bishop can come to h6 if I play them in the other order.

Now, you may ask, once black has those ten moves in, what does he do? Well, that depends on what white does. If white chooses to make few pawn moves himself, and develops followed by castling quickly, there is often good play to be had with ...g5 and ...Ng6 or ...b5 and ...Nb6, aimed against the king on the relevant side. Alternatively, if white has committed himself rather more, black can play for a central pawn break.


The latter was what happened in this game. It continued 11.Ng3 d5 12.Bg2 (note that white cannot close the centre with 12.e5 because it all falls apart after 12...c5 13.f4 g5!) c5 13.f4 Nf6 14.g5 Ng4 15.Bg1 cxd4 16.Bxd4, reaching the position on the right here. I then completed my set of central breaks with 16...e5, blasting open the centre and giving me a position where I had the two bishops in exchange for white's having a central passed pawn.

Post-game analysis showed that this was not objectively justified, but the positions that resulted were extremely complex and hard to find the right moves in over the board, which is about par for the course with the Hippo. And this is the critical feature of the opening that determines whether you should play it: it's not really suitable for chasing an objective advantage that you can easily exploit, but it's highly appropriate for chasing an unclear position, one likely to be won by the player who is happier in sharp tactical melees.

Blind Spots and Beamed Headlights

I have a love/hate relationship with county matches. The "hate" part is all about the travelling I'm required to do for them; any county match further from home than Taunton requires leaving Bideford at 08:30 and getting back at 00:10 the following day. Of course, part of this is because I play for a county in which I no longer live, but, given how much North Devon is represented in the county team, playing for Devon probably wouldn't help much.

The "love" part is all about the matches themselves. Meeting all my old friends from the Somerset League, getting practice playing serious games against tough opposition without its having an immediate effect on my FIDE rating, and on this occasion, winning 11-5 against Hampshire. Nowadays that's not such an amazing result, but when I first started playing for Somerset in 1987, Hampshire were by far the strongest team in the region, and beating them still gives a sense of achievement. (It's a little as if one's national football team beats Uruguay.)

In this particular county match, I also got the opportunity to notice some weaknesses in my own play; this is not something into which I usually have great insight, so I'm pleased when it happens. Take the following two positions:


This is the position after I played 20...Bd5, after which my opponent played 21.c4, attacking my bishop. A natural enough move, but I'd completely missed it. And the reason I'd missed it was that I have a blind spot in my play - I habitually overlook moves that allow en passant captures, even if the capture is not any good. (I had considered 21.c3 and was not going to capture on c3 had Yeo played it, so it certainly wasn't a case of rejecting the move based on how I would have responded to 21.c3.) As it happens, the move I'd overlooked probably wasn't the strongest anyway, but on other occasions it might have been.

From the diagram, play then continued 21.c4 Bf7 22.c5 Nxc5 23.Nxc5 Qxc5 24. Rac1 Qd5, and my opponent's next move was the natural and good 25.Bxf5, winning his pawn back.

I had also considered another move for him here, and this was 25.Bc4, attacking my queen. Some thought reveals why this is not a good move: I can play 25...Rxe1+ 26.Qxe1 Qd7, and white isn't getting his pawn back any time soon.

However, my mind had been distracted by another option for black there, and that was the positional queen sacrifice 25...Qxc4. This is probably not as good as the simple line I've just given, but it attracted itself to my consciousness a little like an oncoming car's headlight does when it's set to "beam", and it would take great effort to restrain myself from playing it in the game. This is, I suppose, an inverted blind spot: when an opportunity to create an unusual material balance occurs, I have to make a special effort to notice anything else.

Trying to find blind spots and beamed headlights is not easy, but it's probably something that will pay off for most players looking to improve their game. I hope it's something that will improve mine.

Opening Concepts: Sicilian Dragon

Today's Opening Concepts article is on the Sicilian Dragon, beloved choice of juniors everywhere. It is characterized by the moves 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6, reaching the position shown on the left.

There are many moves white can play here - 6.f4, 6.g3 and 6.Be2 all have their advocates - but the most aggressive and theoretically critical line is the Yugoslav Attack, 6.Be3 Bg7 7.f3 Nc6 8.Qd2 0-0, after which there is an important theoretical fork. One important line is 9.0-0-0, after which black can make the thematic Sicilian central advance 9...d5!?; sacrificing a pawn for open lines - the sacrifice line runs 10.exd5 Nxd5 11.Nxc6 bxc6 12.Nxd5 cxd5 13.Qxd5 Qc7 (this does not drop a rook because 14.Qxa8 Bf5 forces white to give the queen up with 15.Qxf8+.)


The other main option white has is to prevent the ...d5 advance with 9.Bc4, to which black usually responds with 9...Bd7. White's usual plan from that position is to build up a kingside attack, starting with something like 10.h4, after which many games have been won with some combination of h4-h5 (which, if met with ...Nf6xh5, is followed up with g2-g4 to kick the knight back), Be3-h6 and an exchange of bishops, and 0-0-0 to bring an extra rook into the attack. Black in turn will try attacking on the open c-file with moves like ...Nc6-e5, ...Qd8-a5, and some rook to c8 with the possibility of sacrificing the exchange on c3.



However, there is another way for black to play this position, and that is to forestall white's plan with a move like 10...h5, the Soltis Variation, giving us the diagram on the right. The idea is quite simple: without an h4-h5 advance, it's much harder for white to open a file for his rooks. He now has to work for g2-g4, but even then, it's not so easy: after ...h5xg4, h4-h5 can often be met by ...Nf6xh5, and there's no easy way to kick the knight away.

And so we come to a game I played recently: I had the white side of this position and tried to build up an attack by redirecting my c3 knight over to f4 to support an h4-h5 advance. Unfortunately for me, this allowed black an attacking idea of his own: ...a7-a5-a4 to take advantage of the absence of a defending piece. My king's position ended up smashed to pieces, while my own attack never really got off the ground.