A recent trip to the Cambridge International Open (https://lichess.org/broadcast/3rd-cambridge-international-2025/round-6/Wa4LxDIM) to do some commentary together with Natasha Regan (recordings are here for the next 2 weeks: https://www.twitch.tv/wim_natasharegan) reminded me how much fun commentating is and how nice it is to walk around the tournament hall and watch some games. One game that couldn’t fail to catch my eye was the spectacular round 6 game between GM Daniel Fernandez and the young player Avyukt Dasgupta. After the game, I joined in the analysis session and we discovered some really lovely lines together which I’d like to share with you here! The analysis session was even immortalised by the Dutch artist Pia Sprong 😉 on https://piasprong.nl/detail/1971bbfe-f61f-4871-b776-69961b32b9a0/id/CIO-2025-13.jpg
Fernandez,Daniel Howard (2475) – Dasgupta,Avyukt (2142) [B40]
1.e4 c5 2.Nf3 e6 3.c3 Nf6 4.Bd3 b6 5.0–0 Bb7 6.Re1 Qc7 7.Na3 a6 8.e5 Nd5 9.Be4 Be7 10.Nc2 d6
Daniel has the uncanny knack of being able to generate sharp and irrational play from any type of opening and here, instead of the normal 11.d4, he sets the board on fire with a very sharp idea.
11.c4 dxe5
Black doesn’t back down and takes on the challenge!
12.cxd5 exd5 13.Bf5 e4
This is the point: the knight on f3 has nowhere safe to run which means Black regains the piece by force. However, …exf3 will open the e-file with Black’s king still uncastled so winning the piece back will not come without danger!
14.d4
Threatens Ne5 and thus forces Black to capture the piece immediately. It was the only move I considered but Daniel pointed out the creative idea of 14.b4!? opening the long diagonal for the dark-squared bishop. After 14…0-0 (possible now as White is still not threatening to move the knight away from f3) Daniel suggested 15.bxc5 bxc5 16.Rb1!? His idea is simply that after a future …exf3, Qxf3 followed by Bb2, White will have 2 bishops pointing towards the opponent’s king, both rooks on open files and a knight that can support the white queen’s assault on the kingside via e3-f5/g4. The engines suggest recapturing the piece and then returning Black’s extra pawn for development with 16…exf3 17.Qxf3 Nc6 18.Qxd5 Nb4 with an approximately equal position.
14…exf3
15.dxc5
Watching the game, I thought that White should delay this capture and play the immediate 15.Bg5. I felt that Black’s best defence against Bg5 was always to play …f6 and I thought that delaying the pawn exchange left Black more constricted – for example, Black doesn’t have access to the e5 square for the knight to block the e-file. The engine prefers Daniel’s choice however! After 15.Bg5, Black has 2 choices
- 15…Nc6 is met by 16.dxc5 when 16…bxc5 17.Ne3 leads back to the game with what I thought was the most accurate move order. I’d thought 16…0–0 could be met by “at least” 17.cxb6 Qd6 18.Bxe7 Nxe7 19.Qxf3 with an edge and I expected to be able to find more than this minimal option. Only, as often turns out, that doesn’t exist! So this would be Black’s best reply.
- 15…f6 I wanted to meet with 16.Bxf6 although the engine thinks this is way over the top preferring both 16.Bh4and especially 16.Bd2 with an overwhelming White advantage (nearly +4!). My idea was 16…gxf6 17.Qxf3 to meet 17…0–0
with the winning 18.Qg4+ Kh8 19.Bxh7 Kxh7 20.Re3. However, 17…Nc6 is defensible for Black, though obviously quite fraught in a human game!
So some interesting ideas, but truth was not on my side!
15…bxc5 16.Bg5 Nc6
Once again I was expecting 16…f6 and this is indeed the best move. Once again the engine likes Daniel’s 17.Bh4 best though this is less promising than after 15.Bg5 f6 16.Bh4 (around 0.6-0.7). I got carried away trying to make 17.Ne3 work but I got stuck on 17…d4 which the engine confirms is very strong for Black. I was trying to come up with ways of withdrawing my bishop from f5 to allow my knight to jump there but I couldn’t find anything. 18.Qa4+ is met by 18…Qc6 (18…Bc6 19.Nd5) 19.Bc2 dxe3 and White has nothing.
I thought White was just winning after 16…Nc6, but as always I was veering a touch too wildly in my assessments!
17.Ne3
A very powerful idea: the bishop on g5 is not hanging because of Nxd5+ winning the black queen and Nxd5 is a huge threat anyway. Black has only 1 hope!
17…Nd4
Covers the d5-pawn with the bishop on b7 and introduces some ideas of blocking the e-file with …Ne2+. I thought however that it was all in vain
18.Bxe7
This caught me by surprise as I was expecting 18.Nxd5 Bxd5 19.Rxe7+ Qxe7 20.Bxe7 Kxe7 21.Qd3 which I assumed would be winning for White. It’s much better of course, but Daniel had a few doubts about it and the engine just gives it +1 which is close to winning, but maybe not yet over the edge! 21…fxg2 22.Be4 is the idea. Daniel’s choice however gives Black some unexpected chances.
18…Qxe7
Again, I thought this was losing and again I was rather over the top! The emotion this time because I’d spotted quite a nifty idea:
18…Ne2+ first of all
Now White has 2 possibilities:
- 19.Kh1 is met by the very nice 19…d4
when it’s a little tricky for White to find anything reasonable. White can’t waste time because …dxe3 followed by …fxg2+ mate is serious business! The engines do find equality for White but nothing more. For example: 20.Bg4 Qxe7 21.Bxf3 Bxf3 22.gxf3 0-0 23.Qxe2 dxe3.
2. 19.Kf1 was what Daniel wanted but I had prepared 19….Qxh2 (quite a shock to see that on the board!)
20.gxf3 Nf4. Unfortunately, as Daniel pointed out in a flash, 21.Qa4+is not good news for Black after 21…Kxe7 22.Ng4+! Luckily, we found the lovely 20…Nd4 which is rather stronger (Qa4+ can now be met by …Bc6!) and Daniel’s 21.Qxd4 cxd4 22.Bc5 though White’s best is not sufficient after our and the engine’s line 22…Qh1+ 23.Ke2 Qxe1+ 24.Rxe1 dxe3 25.Kxe3 Bc8! preparing to block the e-file after 26.Kd4+ Be6! The engines point out the better move order 20.Qa4+! Kxe7 21. gxf3 when 21…Rhe8 22.Qa3 Kf8 23.Qxe2 Qf4 24. Qxc5+ Kg8 25.Kd3 d4 is a mess, but a 0.00 mess according to the engines!
However, even after 18…Qxe7 Black has decent chances!
19.Nxd5 Ne2+ 20.Rxe2 Qxe2 21.Nc7+
21…Kf8
Avyukt had got a little pessimistic around this stage – and I assumed Black was in big trouble too – whereas Daniel was convinced that the game was still in balance, and Daniel was absolutely right!
21…Ke7 was Black’s best, and Daniel demonstrated White’s best line as well: 22.Qd7+ Kf6 23.Bd3 A very subtle move order as Daniel feared 23.Nxa8 Bxa8 24.Bd3 Qe6. 23…Qe5 and now Daniel was looking either at 24.Bc4 immediately or 24.Nxa8 Bxa8 25.Bc4 (now or after a preliminary 25.Rd1). The engines think the latter line is just equal! During analysis I got very keen on 22.Qc1, aiming for c5 or g5 but the engines hate it! 22…Rad8 23.Qxc5 Kf6 and it’s all over!
22.Nxa8
22…Bxa8
This is Black’s big mistake! Daniel suggested 22…g6 which isn’t quite right as White has the engine 23.Qb3! (we were looking at 23.Qc1 with a similar idea but the engine idea hitting the bishop on b7 is much more forcing) Bxa8 24.Qc3 (This was Daniel’s spot in another variation and it links everything up perfectly!) 24…Rg8 25.Re1! with a big advantage for White.
Black did however have the possibility of 22…Qxd1 23.Rxd1 g6! Covering the threat of Rd8+ and hitting another piece leaving 2 white pieces en prise and too far apart to defend each other! After the inevitable exchanges, the engines only see complete equality!
Now however Black’s resistance is broken!
23.Qd8+ Qe8 24.Rd1 Bc6 25.Qd6+ Kg8 26.Qxc6 1–0
A really exciting game from both players! Thanks to both as well for the really interesting analysis session afterwards: I’d forgotten how much I’d missed that part of playing chess!