Chess.com 2024, 3rd round, E12 - Narrow road to a loss

30^th of June

Yet another game where I did not explore my opponent’s options enough. 

He has tied up my main forces and is trying to pick up the kingside pawns. If I played h5 rather than g4 on the last move, it would’ve been an option – but I was playing for a win rather than a draw, a clear evaluation mistake. The rook also should have stayed on the first rank rather than go to b3.

I was overly self-critical here. The engine shows several ways of holding the draw without too much creativity on white’s part. It’s the fact that I didn’t see last black’s move that took me off the psychological balance. Of course, the loss earlier in the week still rankled.

Not sure if it’s the loss to furiously fast or just general poor play on my part, but I really screwed up this part of the tournament. I see some drawing chances with trapping the black king, but I’m sure he’ll find his way around them.

His plan is to install the knight on the blockading d2 spot. The question is whether I’ll manage to grab the pawn on b4 to open another line of attack. Looks doable, I just need to tie him up first with Rc2:

I missed an idea here. Black could go after the b4 pawn with Rc4 and Rdd4. White still had defensive chances, but that was the only way to play for a win.

England had an incredible comeback in the 1/8 final. A scissor kick goal in the last minute of added time and then an extra goal in extra time. I don’t see them getting far with that kind of play though; they were minutes away from leaving the tournament.

Thankfully, the VFK game went down one of the lines I’ve calculated, which makes things much easier with fewer pieces on the board. Incredibly, I managed to hold on to my extra pawn. Not sure yet if the endgame can be won, but the more exchanges happen the better.

I didn’t notice it at the time, but white did a major mistake with 27. Rb1. It’s understandable, since the position is complex and while he got most of the calculations right, a single mistake is enough to tip the balance. The right options were 27. Re1 or Ra1. The eval went from equal to -1.66

Hotrik has finally finished his tournament.

The smoke has cleared from the VFK game and now I need to figure out how to convert. Exchanging rooks is in my favour. Not so sure about exchanging bishops though, but if my rook will stay on the right side of the pawn, that might be fine too. I’m quite sure he’ll use a4 as a blockading square with Ba4. The extra kingside pawns are in my favour, since I can go after them after distracting white with the a-pawn passer. By the way, no pressure, but this might be the only game that gives me a chance to get 2nd place.

So, the plan is:

• Force the white bishop to a4.
• Get the king active and install my bishop on d5.
• Use threats of rooks’ exchange to push my a-pawn.

I know – very fuzzy, but at least there are some ideas.

He is obviously going after my disjointed pawns. The only defense I see is Rf4, which strangely works against a brute force approach.

Rf4 is not too bad, but I should have considered giving up a pawn as the price for reducing material or activating the rook. 34. a4 was an interesting option, with the idea of Rd4-Rd7. The eval is -0.68.

On to the chess_lover chap, where I’m continuing to suffer.

I’m quite sure he’ll respond to g5 with h5. The question is whether I can sacrifice the pawn on g6 and not lose the game. Otherwise, he cements the kingside with g6 and the game moves to the queenside, probably after he installs a knight on f4.

I can’t see lines that work and if I play N3h4, he’ll just respond with Ng5. Looks like he again found the right sequence of moves to block my threats and promote his.

Yes, indeedy. The eval is already -1.57 despite material being equal. All my observations were correct, though unhelpful.

I can’t quite understand the idea of this move. Does he want to go c5? On the other hand, he’s not fighting for the d-file which is still firmly in my control. I’ll leave this for tomorrow – feel a bit unwell.

Hindsight being 20/20, the idea of the move is simple – hold the fort. The eval is +0.4.

1^st of July

Going back to the game above. After thinking it through a bit more, I see the following points:

• The black rook has to guard the 7^th rank to prevent Qd7. So, he’s unlikely to play Rc8.
• I shouldn’t let him play c5; it frees up black too much and creates counterplay, especially if he gets to keep the rook on c7.

I don’t see a good breakthrough, however; considering how previous games went, I’d rather play it safe and let him make the committal moves. So, decided to play f3 to strengthen the e4 pawn and open the king a path should we go to the endgame.

White’s last move, Qf4, was another miss in my calculations. His threat is simple: g4-g5 and win one of the bishops. This doesn’t give me time for e4-e3, which I was planning as a response to Qg3. The only reasonable response I see is Qa5. The problem is that after Qa5, things look quite bleak as and when he plays Qc7+:

Unfortunately, I don’t see any alternative here that doesn’t lose immediately. Another miscalculation in a long series of miscalculations.

There were two strong ideas here that I didn’t notice.

1. 1. Be7, and if white plays g4, Bd6.
   2. Qb5, and if white plays g4, Be5.

Both are to do with fighting the pin by attacking the queen. While neither fully equalize, they are way better than the option I reconciled myself to.

I was planning Rd4, however now I see that white can play Ke3 which has a path to a draw. As soon as the rook moves away from the d-file, white plays Nd6+, exchanges the bishop, gets the d-pawn and the game’s over:

Hence, I’d rather exchange on d2 and try keeping the knight there.

At this point, white has fully equalized – the engine displays zeros all over for the eval.

I was planning Nd4, followed by Kd3 here. He is obviously going after the various weaknesses spread around the board; the question is whether I can hold on to them. One option for black is Be5 which almost certainly wins a pawn at the price of a minor piece swap.

As always, interesting lines here, and again, as always, they must be full of holes. One thing I do know – if I manage to get to a rook ending with f and h pawns, that’s a draw. This is one consolation.

A pawn sacrifice with c4 was still the way to go here. Unfortunately, I didn’t see black’s very narrow path to a win following Nd4.

Furiously fast scores another point against VFK – one more game played at 2250 level from him. This is probably his true rating.

White made another inaccuracy here; 32. Rb8+ should have been played before Ba4.

No surprises here. There is a trap with Re7, but I’m sure he’ll spot it. My plan yesterday was more prosaic: get the king active, try to exchange rooks and remove the obstacle on a4.

He played h5, as expected. I need to move to defense, Ng3-e2, maybe Qc3, Nd2.

There was an idea with 24. b3 or b4, winning the pawn back. White is still at a -1.6 eval, but I should’ve at least analyzed this.

2^nd of July

4 weeks into the tournament, and I still have half the games active. As the flight to the UK is coming up, I’m sure I’ll have the tournament proceeding there; not something I was planning on. At least I know now that I won’t be dealing with round 4.

Black’s last move was a bit of a surprise. Not because I haven’t seen it, but because I was sure Be5 is far more dangerous. The first order of the day is to check if the rook endgame isn’t losing.

This would be the f and h pawn endgame that was emerging through the fog a couple of moves ago and chess lore says it’s holdable. Whether I’ll manage to do that is a wholly different question, but I don’t intend to lose two rook endgames to the same person.

He can try to evade exchanges, but then the king just goes to e4 and I should be fine. So, conclusion – our final battle will be a rook ending with f and h pawns for black and a pawn for white.

Incorrect, unfortunately, as again – I did not see black’s devious plan in the pawn endgame.

Now to the problematic Draqlek game:

He nicely evaded my mini-trap with Rc4, h6+. I can delay taking on h6 by playing Kf4, but then he’ll just force me to do the same with Rc4 which creates an annoying zugzwang. I think my only chance lies with the gxh6, Kf3 line. Let’s check…

I missed an idea here: 36.g6+. While obvious when the comp shows it to you, but I should be able to spot such simple options. I also didn’t see that after 36. gxh6 Kxh6 white can just hold the position, as black cannot take the pawn on g5; the white h-pawn would be unstoppable.

Needless to say, all the options I calculated below are inferior to those two.

On closer inspection, it looks like all the roads lead to Rome – namely we end up in a rook endgame with black’s d-e-f pawns against white’s g-h pawns. I still don’t know if there’s a draw possible, but it doesn’t matter whether I play gxh6 or Kf4 – the result is the same.

There is an empty queue again. At this point, I’ve sped up to the point where in all games my opponents move slower than me.

Black’s last move is very sneaky. I originally thought that Kc4 equalizes, but the pawn ending looks very shaky. There are very interesting calculations there, where I see a path to a draw. I might be again missing something; same as with the other endgame:

I did miss something, same as in the other endgame. Black is playing top line again, also same as in the other endgame. In this case, I’m not going to blame myself, since black’s path to a win is study-like.

However, let’s assume for a moment he won’t go to the pawn ending; are there any other black moves I need to be worried about? Kd6 leads to an immediate draw, while the other options just allow me to untangle and have an active king. Alea jacta est!

I get my main question answered a few minutes later – he is going for the pawn ending and we blitz out a bunch of moves:

A few quick moves later, I see a draw on the horizon. I just can’t figure out what else he can cook up here. If so, it would be one of the best defenses I’ve ever done. Too bad I stumbled at the finish line in the other game.

At the risk of sounding like a broken record, I’ll say it again – most black’s moves are the only path to a win. He must have seen it about 15 moves ahead, back when the exchange on d4 has happened. Very impressive. It’s understandable that I assumed a draw with both h- and f-pawns queening, but he saw further.

So, I lost that game as well. After writing the note above, I’ve noticed that he can force a queen exchange and then grab the pawn on a4 with a tempo to spare. This explains why he played a5 on the 42^nd move.

This is especially painful, since apparently, I played that game well. Only my opponent played at IM level. 10 games left.

This is a very tough loss. 97.8 precision from black in this game is the highest I’ve seen in a full game. Perhaps I should have focused on making a draw at the very start, when my opponent went into a symmetric mode. The 21…Be7 move is one of the most surprising I’ve seen in this tournament and it’s rightfully shown with two exclamation marks. The pawn endgame is certainly a learning experience, though I doubt I’d be able to detect such deep lines and ideas even after going through this analysis with a magnifying glass.

In terms of sheer quality, this is the game I’m most proud of in the tournament. I found many tricky defenses and held for a long time despite black playing top line almost the entire game.

It’s very difficult to focus on other games after such a letdown.

Looks like c5 wasn’t his plan; black just wants to park the bus. The position is almost entirely symmetrical, with the only difference being white’s control of the d-file and entry points at b6 and d6. I can try routing the knight to c5, though not sure it will change the eval much. Still, that’s my only path of getting something out of this game.

There was also a gradual space acquisition with Nf5, h4, g3, Kf2. A Karpov-style approach – a mental note to look for micro improvements in such situations.

This looks lost – I honestly screwed up here. I’ll try some activity with Ke6, but will likely resign in a few moves.

White had the option of 27. b4 which would’ve finished the game sooner. I was a bit too pessimistic here. The eval is +1.78, so it’s not resignable just yet.

He’s putting up decent resistance. I’m thinking Bh6 and Nh4-f5 should give him something to think about.

Black’s plan is clear: go Rd7 and break through the d-file. I think Qd2 should hold the fort. On closer inspection, I see that black may get all kind of interesting ideas with breaking through on d3, so it’s not all as peaceful and calm as I thought.

Re3 was a big mistake. I missed out on 29. f5 which would have equalized. On the other hand, Re3 leaves the first rank unprotected and chases after phantoms.

Another game where I’m lost. This is all tumbling down as a house of cards… White’s treatment of the French Advance was unlike anything I’ve seen before. My only idea is to untangle the kingside with g5.

All moves continue to be top line for white, and indeed I’m lost, as the eval is +1.89.

3^rd of July

My plan is :

• Get the white rook into a passive defense with Rc2. This also ties up the white knight due to Rxg2 and Re2+ threats.
• Get the king to d5.
• Exchange the rooks by coming back to the e-file.
• Chop off the pawn on b4.

The plan is easily negated. White has multiple ways to secure the draw, as I wasn’t creative enough in this endgame. The engine flatlines on several top lines.

I’ll publish the table, as painful as it is.

e3 is probably my only chance here, though that’s not saying a lot. I’ll give it a shot.

The engine quotes that as another major mistake and points out that Kd5 is mandatory before playing e3. I was in “Hail Mary” mode and just looking for practical chances while mentally resigning. The eval is already +3.

Continued here: https://www.chess.com/blog/RomanKleiner/chess-com-2024-3rd-round-e13-under-the-heatwave