Down to the wire, part II

 After the bright start to my Oxfordshire League campaign that I have already recounted, my next few games in the competition followed a depressing pattern. Wantage would be at home against one of the top teams, giving me black against a significantly stronger opponent. I would play well for a time, and often have some chances to fight for the win, only to lose my way towards the end of the night. This stopped only at the end of November when I received a gift in a drawn pawn ending, so between that slice of good fortune and the improved showing at Coventry, things finally seemed to be looking up.

And they needed to be. The next match was a proverbial six-pointer against our relegation rivals Didcot, with both teams looking a little depleted. I had the black pieces on board 1 against Nicholas Gough, rated 1956 though a lot higher until an unhappy recent run of form. In a new county almost everyone is unknown to me, and often to my database too, so I couldn't prepare at all for this key game.

Even so, the opening went my way. I gradually built an advantage with some strategic finesse, then lost it in a noteworthy fashion. Yet again in 2025 one of my games had to go the distance, and this time the fate of the match was riding on it:

[Event "Oxfordshire League Div. 1"] [Site "?"] [Date "2025.12.02"] [Round "7.1"] [White "Gough, Nicholas L"] [Black "Young, Daniel J"] [Result "0-1"] [ECO "C88"] [WhiteElo "1956"] [BlackElo "2027"] [Annotator "Young,Daniel J"] [PlyCount "116"] [EventDate "2025.??.??"] [SourceVersionDate "2025.12.02"] [WhiteTeam "Didcot"] [BlackTeam "Wantage"] {This was an important match against our fellow relegation candidates. I couldn't find any games of my opponent so I was on my own for this one.} 1. e4 {(1:20)} e5 {(1:20)} 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O {Up to this point, no-one had challenged me in a main line Ruy Lopez since I took up 1... e5. Finally it was time to hint at what had been prepared.} Be7 6. Re1 b5 7. Bb3 O-O {Now if 8. c3 to reach the main lines, Black may play the sharp Marshall Attack (8... d5!?)... or he may not. But I think there's no harm in asking White if he wants to take the chance, for if he doesn't then he has to fall back on one of the anti-Marshall systems, which aren't quite so threatening.} 8. a4 {The bluff (or was it?) works. This is the most popular anti-Marshall (8. h3 also has its supporters) but the decline of 1. e4 at elite level owes much to its lack of bite.} Bb7 9. c3 $6 {And this is inconsistent. The reason that 8. a4 and 8. h3 can be considered anti-Marshall variations is that White forgoes c2-c3: the pawn move obstructs White's queenside development, and a local advantage on the kingside is what gives the Marshall its power. White tries to revert to the normal lines a move later, but Black is now very happy to switch to Marshall-like play in which ...Bb7 is clearly a more useful move than a2-a4.} (9. d3 {is usual and more modest, just preparing to develop. From here Black usually responds in kind with 9... d6 and conventional Ruy Lopez play, though the ambitious 9... d5!? might be workable too.}) 9... d5 $1 10. exd5 {(1:09)} Nxd5 {(1:19)} 11. d4 {An admission that White is only trying to equalise - but after all, Black is ahead in development and has his share of the centre, so why not? It is simply too dangerous to accept the pawn sacrifice now.} (11. Nxe5 $143 $2 {gives Black a decisive positional advantage after} Nxe5 12. Rxe5 Nf4 13. d4 (13. g3 $2 Nd3 14. Re2 Bc5 $19 {causes f2 to collapse}) 13... Nxg2 14. d5 Bd6 15. Rg5 Nf4 16. Bxf4 Bxf4 17. Rg2 Re8 $19 {: White's kingside is exposed and the d-pawn is vulnerable, whilst Black has two bishops and the e-file.}) (11. axb5 axb5 12. Rxa8 Bxa8 $11 {is fine for White provided he plays d2-d4 soon, but as another example of how greed can backfire on him,} 13. Qe2 $2 {is a fatal loss of time:} Nf4 14. Qxb5 Nd3 15. Rf1 e4 {White's options are limited by the unhappy bishop at c1, and} 16. Nd4 Nxd4 17. cxd4 Qxd4 $19 {wins back the pawn with an overwhelmingly active position.}) 11... exd4 12. cxd4 {I was suspicious of this choice since I assumed (correctly) that I would get comfortable play against the IQP, but nonetheless White judged it well - he needs a foothold in the centre.} (12. Nxd4 $143 $6 {gives up the centre and keeps his other knight out of c3. If White insists on avoiding the structural weakness then Black can quickly take over, for example} Nxd4 13. Qxd4 $2 Re8 14. Nd2 Bf6 15. Rxe8+ Qxe8 {White has problems with his back rank, and Black can take advantage with} 16. Qg4 h5 $1 17. Qg3 (17. Qxh5 $4 {loses to} Nf4 18. Qd1 Rd8 $19 {. Black can take on g2 at his leisure and White's queenside pieces will never get into the game.}) 17... Qe2 {...h4 is now a major threat, and after the forced} 18. Nf1 $8 h4 19. Qh3 Rd8 $17 {Black is in control: his pieces are obviously way better than their counterparts, and he hasn't even had to invest any material.}) 12... Bb4 {12... Bf6 also came into consideration, but I was still in a Karpovian frame of mind and decided to play it prophylactically. The text stops White developing smoothly with Nc3 for the moment, and he can only get that move through by allowing trades - which dampen any initiative that might make up for his IQP. The tempting 12... b4?! would be less good, however: it does fix the queenside and control c3, but it allows White to jump onto a new circuit with Nbd2-c4.} 13. Bd2 Bxd2 14. Nbxd2 (14. Qxd2 {was what I expected, reserving the knight for c3 and also not letting me harass the b3 bishop with ...Na5. The computer comes up with a very sharp continuation in} b4 {(I wouldn't have considered this, but the blockage on d2 might justify giving away c4)} 15. Bc2 g6 16. Qd3 {Preparing to develop the queenside, but allowing} Nf4 17. Qe3 Nd5 18. Qd3 Nf4 19. Qe3 {White appears to have nothing better than to repeat, but Black could hazard playing on with the fascinating} Nxg2 $5 20. Kxg2 Nxd4 21. Be4 Nxf3 22. Bxf3 Bxf3+ 23. Qxf3 Qg5+ 24. Kh1 Rad8 $44 {The engine thinks Black has full compensation for the piece, and I could believe it - besides two pawns and White's leaky king position, the knight on b1 is essentially stalemated.}) 14... Nf6 {Just asking politely what White is doing about his d-pawn. 14... Nf4 is more natural, but 15. Ne4 felt active enough for White so I chose to control that square.} 15. Nf1 {(58) Without doing anything special, Black has diverted this knight from its stable on c3 to the backwater of f1. Now I look to squeeze more concessions out of White.} (15. Ne4 {would have been met by} Nxe4 16. Rxe4 Na5 17. Re3 Nxb3 18. Qxb3 $11 {, since the d-pawn becomes more and more of a static weakness as pieces disappear.}) 15... Na5 {(56)} 16. Bc2 $5 {This bishop is usually White's best piece in the Ruy Lopez, and it's no surprise that he wants to keep it. The engine thinks this is a mistake; I disagree, but it is a luxury that puts some obligation on White, since he needs to get creative to make it work.} (16. axb5 axb5 17. Ne5 {is a safer continuation, evading the trade on f3. The weakness of b5 means that Black doesn't win a pawn for long after} Nxb3 18. Qxb3 Qxd4 19. Rxa8 Bxa8 20. Qxb5 $11 {, after which White has dissolved his structural weakness and has no problems.}) 16... Bxf3 17. gxf3 $6 {Simplistically defending the pawn, but White needed to do more than this - there is no future in a passive position so rife with weaknesses.} (17. Qxf3 $142 $1 {is the way to go, abandoning the d-pawn in order to bring his pieces back to life.} Qxd4 18. Ne3 {(at the board I expected 18. Rad1, but doubted White would have enough for two pawns after ...Qxb2)} Qxb2 19. Qe2 $44 {A surprisingly tricky idea, threatening to win Black's queen with 20. Bxh7+. White will gain further time against the queen, and the loose queenside combined with the sidelined knight on a5 offer him a surprising degree of compensation. I have to admit that I'm less cheery than the engine about White's chances, but still this is better than the game turns out since White is condemned merely to defend.}) 17... Nc6 {Black is still playing simple chess. I return a knight from the edge of the board to the centre, training my sights on the d-pawn once more. White therefore needs to divert my queen from the d-file, and this leads him to finally release the tension on the a-file.} 18. axb5 axb5 19. Rxa8 Qxa8 20. Be4 {(32) Defending for now, but I was ready with an unexpected answer.} Nxe4 $5 {(42) Outwardly a very strange decision, and yet there is a method in the madness. Although White is gifted the centre and can undouble his pawns into the bargain, he also loses the pin that was keeping the d-pawn safe. So once it is attacked again it will be forced to move, and then Black's pieces can start to slip between the cracks.} 21. fxe4 $6 (21. Rxe4 $142 {It's hard to reconcile oneself to an ugly move like this, but it would have kept the rook active. Here White could have gained some more meaningful play with} Qd8 22. Ne3 Qf6 23. Kf1 Rd8 24. d5 $13 {, which is a genuine threat in spite of the pin: not} Qxb2 $4 25. dxc6 $1 Rxd1+ 26. Nxd1 $18 {when Black is threatened with mate as well as the loss of his queen.}) 21... Qd8 22. d5 Qg5+ $6 {I failed to resist the siren call of this check. It seemed like a useful tempo before moving my knight, however it is White who should profit by it since my queen will soon need to spend a move going to h4 or f6 anyway.} (22... Ne5 $142 23. Ng3 Qf6 {homes in on the f3 square, and if} 24. Kg2 g6 25. Qe2 h5 $15 {retains some initiative. White has more space but also the commitments that come with defending it.}) 23. Kh1 $6 (23. Ng3 $142 {fights for the initiative. It's not so much about this knight as freeing the f1 square - after} Ne5 24. Rf1 $11 {Black will have to lose time dealing with the threat of 25. f4.}) 23... Ne5 {I felt I was doing well here. My knight will be hard to dislodge, White's kingside is compromised, and a later ...f5 could further chew up his pawn centre at the same time as getting my rook into play. The evaluation is correct, but it turned out not to be as easy as I'd hoped...} 24. Qd2 $6 {By now it is White who is keen for simplification to ease the pressure on his kingside, but this is an unfortunate way to go about it since Black now gains a precious tempo.} Qh4 {Naturally refusing the trade, and now threatening ...Nf3.} 25. Qe3 {(14) Now comes one of the critical moments of this game. When I saw the following x-ray tactic it felt like the logical conclusion of the play, which has been tending in my favour up to now. But all that glisters is not gold.} Ng4 $2 {(33) This turns out to be a very instructive mistake, and one on the theme of cashing in a positional advantage prematurely. Black wins a pawn, but is forced to concede a trade of queens and thereby loses the initiative.} (25... f5 $6 {This also came to mind for the reasons outlined in the note to my 23rd move, but doesn't really go anywhere. After} 26. Qg3 {my intention was} Qxg3 (26... Qe7 $142 {keeps more chances, forcing White to release the central tension with} 27. exf5 Rxf5 {, but} 28. Qb3 $1 $132 {gives White counterplay on the queenside} (28. f4 $2 {loses a pawn to} Qf7 $17)) 27. Nxg3 fxe4 28. Nxe4 $11 {The f-file is open and White's pawns are all isolated, but I preferred to win a pawn. In truth White isn't in much danger here, since his king can deal with f2, and d5-d6 will liquidate another weakness.}) (25... h5 $142 {is a better (i.e., more patient) path. The arrival of the knight on g4 will carry more sting if it's defended when it gets there.} 26. Qg3 Qf6 {Black needn't feel queasy about inviting} 27. f4 (27. Nd2 {there's no time for White to calmly improve his pieces, since} h4 28. Qe3 Ng4 29. Qe2 Qxb2 $17 {wins a pawn for no compensation.}) 27... Ng6 $17 {, as White can't maintain this idyllic pawn front.}) 26. Qg3 Nxf2+ {I made this move with the impression that Gough had overlooked it, but perhaps I was being unkind since White gains the most from this transaction. Although Black wins a pawn, for now at least, without queens it's much harder to challenge his pawn centre. The extra space, combined with my weak queenside pawns, allow White to finally generate some chances of his own on that side of the board.} 27. Kg2 Qxg3+ 28. Nxg3 Nd3 29. Re3 $1 {Over the next few moves, my opponent fights back excellently. I'd seen that Re2 fails to ...Nf4+, and the submissive 29. Rb1 can be dismissed on principle, so I thought I must be doing very well. White's active defence is, of course, much better than either of those moves.} Ne5 {29... Rxb2 30. Rb3 would probably turn out similar to the game only with both b-pawns gone. Since pawn exchanges generally favour the defender in endgames, I decided to steer clear.} 30. Rc3 {(12)} Rc8 {(30)} 31. Rc5 {This was the third move in a row I'd failed to foresee. Suddenly the d-pawn casts a long shadow, and by now I understood that I had to readjust and start thinking about how to save the game. The turnaround did not come at a good time for the team - although we'd won easily on board 2, elsewhere the situation looked bleak. And we really needed a result from this match.} c6 32. Nf5 Kf8 $8 33. Nd6 {This is fine, but equally it is the point White starts to go astray again, and for the same reason as me a few moves earlier - being seduced by a pretty tactic. As Gough pointed out afterwards, 33. Nd4 would regain the pawn without any adventures, and should be an easy draw. Instead he plays for more.} Rc7 $8 {Losing a pawn, but Black has no choice and should be glad to get some air for his rook. Not 33... Rd8? 34. dxc6 and wins.} 34. Nxb5 Rb7 $8 (34... Rc8 $143 $2 {I took a long time on my 33rd move, for at first I gloomily considered this the best available, but finally I spotted my follow-up to the game continuation. If I hadn't then I would probably even have lost this game, since} 35. Na7 Rb8 36. dxc6 {and yet again the passed pawn has its say.} (36. Nxc6 $143 $2 {actually throws away the win for White, since} Rxb2+ 37. Kh3 Rb3+ 38. Kh4 Re3 $1 {rounds up the e-pawn, whilst White's passed pawn is too close to the black king in this line. If} 39. Nxe5 Rxe4+ 40. Ng4 f5 $11) 36... Rxb2+ 37. Kg3 Rb3+ 38. Kh4 $18 {The c-pawn costs Black a piece.}) 35. dxc6 {(5)} Nxc6 $8 {(10) Just clinging on! With the passed pawn gone, Black has at least eliminated any danger of losing. But would that be enough?} 36. Rxc6 $2 {(White draw offer) Alas for White, this peace offering would be valid after almost any move besides the one played. For once it is precisely in a pure rook ending that Black finds winning chances, because the text trades White's active piece for my passive one.} (36. Kf2 $142 {White need not force events and can calmly improve his king position.} Ke7 37. Ke3 $11 {White has no problems once his king is in on the action, but compare this to what follows.}) 36... Rxb5 $17 {I was almost ready to accept the draw - I'd been walking a tightrope for the last few moves, and all of my pawns are on the same side. Indeed it is White who can boast the outside passed pawn. But from looking at the other boards I realised that the match situation required me to play for a win. After I'd played my move, it dawned on me that in fact Black has very good winning chances. It is all about the activity of the rooks.} 37. Rc2 {I hated this move on White's behalf, but active defence doesn't work so it seems that there's nothing else but this meek retreat. We now reach another important moment in the game, and a good position for less experienced readers to stop and think about Black's next move.} (37. Kg3 $143 $6 {Feeling intuitively that the game continuation is hopeless for White in the long run, I wondered if White could just give up his b-pawn to activate the king. However, not every pawn-down rook ending confined to one flank is a draw, and this would merely have accelerated the end.} Rxb2 38. h4 h5 {Fixing the weakness.} 39. Rc8+ Ke7 40. Kf4 g6 41. Rc7+ Kf6 42. Rc6+ Kg7 43. Kg3 Rb3+ 44. Kg2 Re3 $19 {White's fundamental problem is that the e- and h-pawns are too far apart, so his king gets overstretched by the enemy rook.}) 37... Rb3 $1 {Confining White's king to the first two ranks. Any chance to restrict the enemy monarch is of paramount importance in the endgame, as I remembered from a video series on Chess24 by the late Mark Dvoretsky. White is now forced to switch purely to passive defence - never a good omen.} 38. Kf2 Ke7 {Black meanwhile has no such difficulty and hastens to bring his king towards the e-pawn.} 39. Re2 Ke6 40. Kg2 {(4)} Ke5 {(8)} 41. Kf2 $2 {White is defending both of his weak pawns for now, and decides to wait on events. This strategy is doomed to fail.} (41. Rc2 $142 $1 {Now that Black's king is further away from its own pawns, White can hope for some counterplay and should certainly roll the dice. Black retains considerable winning chances, but has to show some care.} g6 (41... Kxe4 $143 $2 {is a hasty choice that would spoil everything after} 42. Rc7 $1 Rxb2+ 43. Kg3 Rb3+ 44. Kg2 $11 {, and White's rook gets amongst the black pawns since ...Rf3 is unsafe.}) 42. Rc5+ Kd6 $1 43. Rc2 h6 44. Rf2 Ke6 {Carefully shepherding the black pawns. White is starting to run out of moves here, so the engine decides to throw caution to the wind with} 45. e5 $5 (45. Re2 Ke5 46. Rf2 f6 47. Re2 h5 $19 {and White faces the spectre of zugzwang.}) 45... Rb5 46. Rf6+ Ke7 47. Rf2 Rxe5 48. b4 $17 {White has given up the e-pawn to at least lift the lid on his b-pawn. Since Black also has a passed pawn on f7 my feeling is that this isn't enough, but White has to try something.}) 41... g5 $6 {This is not an objectively bad move, but it sets the scene for a blight on this otherwise very satisfying endgame. When the opponent has no counterplay, time is not of the essence - in fact, prolonging the torture is often the quickest way to win since it may provoke the opponent. In this case, it's sounder for Black to move the pawns up gradually and avoid giving away so many squares on the f-file.} (41... f6 $142 42. Kg2 h5 {Now that the f-pawn is defended, the recipe in the game does not work, and if White marks time with} 43. Kf2 {then he deprives the rook of this f-file, enabling} g5 44. Kg2 Rb4 {, harvesting the pawn in a different way.} 45. Rd2 (45. Kf3 g4+ $19) 45... Kxe4 $19) 42. Kg2 h5 $4 {It's hard to accept, but this weakening move should cost Black the win. The correct way to play is 42... f6, pivoting back towards the previous variation.} 43. Rf2 f6 44. Rf5+ Ke6 (44... Kxe4 $144 {Even with the g- and h-pawns out of the firing line, this isn't enough to win either.} 45. Rxf6 Rxb2+ 46. Kg3 Ke5 47. Rh6 h4+ 48. Kh3 Kf5 49. Ra6 $11 {White's king is in front of the enemy pawns, and whilst its position causes some alarm Black has no way to play ...g4+ safely.}) 45. Rf2 $4 {(2) Luckily for me, Gough was by now in serious time trouble, and couldn't find anything better than returning to cover the b-pawn. I'd noticed the correct move whilst he was thinking, although I didn't realise that it could have been White's salvation.} (45. h4 $142 $1 {The point is to break up Black's pawn front. Every trade of pawns brings White closer to a draw, and for this purpose the b-pawn has no significance.} Rxb2+ 46. Kg3 Rb3+ 47. Kf2 g4 {There would have been plenty of play left even here, but the computer appears to be able to hold it after} (47... gxh4 $144 {will reduce to a drawn RP vs R position:} 48. Rxh5 h3 49. Kg1 Re3 50. Kh1 Rxe4 51. Rxh3 Rg4 {Even though White's king gets cut off on the h-file, he finds enough activity with his rook on the long side.} 52. Re3+ Kf5 53. Re8 Kf4 54. Ra8 $11) 48. Rxh5 Rf3+ 49. Kg2 Re3 50. Ra5 Rxe4 51. Kg3 f5 {Maddeningly, even connected passed pawns aren't enough for Black, since the h-pawn is able to make a nuisance of itself.} 52. Ra6+ Ke5 53. h5 Re3+ 54. Kh4 Kf4 55. Ra4+ Kf3 56. h6 {There's nothing more than} g3 57. h7 g2 58. h8=Q g1=Q $11 {, when perpetual check is inevitable.}) 45... Re3 {(6) Black finally wins this pawn outright, and now it's just a matter of time until the black pawns march down the board.} 46. Rd2 {46. b4 Rb3 would not have changed anything. By now this was the last game going, and Wantage were 3-2 down, so there was still some pressure, but I was able to put it away with no further surprises.} Rxe4 $19 47. Kf2 Rb4 48. Kg2 f5 49. Kf2 {(I stopped recording)} Rb3 50. Kg2 g4 51. Rc2 Ke5 52. Kf2 h4 53. Kg2 Kf4 54. Rc4+ Kg5 55. Rc2 g3 $6 {Good enough, though we've already seen that pawn trades can make things go wrong.} (55... f4 $142 {is simpler, not fearing} 56. Rc5+ (56. Rd2 $144 f3+ 57. Kf2 Kf4 58. Rd4+ Kf5 59. Rd2 Ke4 $19 {wins quickly thanks to zugzwang}) 56... Kf6 57. Rc6+ Ke5 58. Rc5+ Kd4 $19 {when the king invades.}) 56. hxg3 hxg3 57. Rc3 $2 {A final oversight, but the position was already lost.} Rxc3 58. bxc3 f4 {Black need only run after the b-pawn, so White resigned. At the bottom of a league every point is precious, so this was an invaluable point. I feel it was deserved on the whole, and a game packed with insightful moments.} 0-1

This was a hard-fought practical game, though consequently not without its share of mistakes. Still, if there were no errors in chess games I wouldn't have a blog to write! The troubles started in the opening, with White giving me a supercharged Marshall Attack by 9. c3?! - it's completely understandable to want to avoid that particular minefield, but it is justified on positional grounds rather than by an exact tactical sequence. 8. a4 should be seen as the gateway to an anti-Marshall system, not an anti-Marshall move in itself.

My opponent thus had to accept a stale form of IQP to avoid anything worse, and in the early middlegame I gained the upper hand just by putting pieces on sensible squares and consistently targeting it. It helped that White played straightforwardly to defend it, rather than risking more active counterplay (17. Qxf3! or 21. Rxe4) to try changing the narrative. But 25... Ng4? could have ruined everything, and just as I was closing in on a truly serious advantage. Winning a pawn is all well and good, but a queen exchange is usually one of the biggest strategic tremors in the game, and I didn't consider its ramifications. The loss of the initiative was so serious that within a few moves I had to find a string of only moves just to survive.

I'm often slow to react to a change in evaluation, and I didn't immediately realise that White's fateful decision to accept the pure rook ending was the decisive mistake. 'All rook endings are drawn' is one of those unhelpful chess principles that is a half-truth at best (just like 'always simplify with a material advantage', funnily enough) and in this case Black's positional advantage was enough to win the game even with equal pawns. Even then, there could have been a sting in the tail if White had found 45. h4!, which would have disastrously broken up my structure.

In the end it was a precious win, and enough to pull a draw out of the hat for Wantage. Crucially, I was able to keep a cool head at the climax of both this game and the last, which was certainly not the case in my younger years. I suppose that comes with having a bit more confidence in such positions, so maybe the whole process of improving my technique has been a virtuous circle.

One side-effect of my learning journey in the endgame has been that a lot of the games I've chosen to present recently have been fairly dry. That is a true reflection of my recent practice, but I do hanker after the excitement of old, and the next game was much more fun.