Busted French line?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: Harvey Williamson, bob, hgm

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
alchemist888
Posts: 8
Joined: Fri Mar 29, 2019 2:02 pm
Full name: Denis S.

Busted French line?

Post by alchemist888 » Fri Oct 02, 2020 3:52 pm


There was a discussion in TCEC chat about this French line. Intuitively, this should be lost, since Black's can almost never play Nf5. And the typical counterplay with ...f6 is complicated by lack of development.
Of course, it takes a lot of analysis to prove that Black's is losing. But I was inspired by how the Grob opening was parsed on this forum and decide to create this topic.

Also both games from TCEC were losing for black:
https://tcec-chess.com/#div=sf&game=5&season=19
https://tcec-chess.com/#div=sf&game=6&season=19

That would be great if someone share their analyzes and thoughts about 1.e4 e6 2.d4 d5 3. Nc3 Nf6 4.e5 Ng8?!

Tony P.
Posts: 203
Joined: Sun Jan 22, 2017 7:30 pm
Location: Russia

Re: Busted French line?

Post by Tony P. » Fri Oct 02, 2020 5:19 pm

The line hasn't been explored deeply on chessdb.cn yet :( You can submit moves for analysis there and see what SF says at depth 22 (and also make a better use of transpositions than Homefish would). So far, the score is estimated to be between +50 and +70 (centipawns?).

mmt
Posts: 296
Joined: Sun Aug 25, 2019 6:33 am
Full name: .

Re: Busted French line?

Post by mmt » Fri Oct 02, 2020 6:18 pm

I'll do some deep analysis.

User avatar
Ajedrecista
Posts: 1484
Joined: Wed Jul 13, 2011 7:04 pm
Location: Madrid, Spain.
Contact:

Re: Busted French line?

Post by Ajedrecista » Fri Oct 02, 2020 6:53 pm

Hello:

This weird move can lead to even weirdest positions, like in Árnason - Petrosian of 1983:

Árnason - Petrosian (Tallinn, 1983)



Please take a look at the game after the ninth move: white knights on c3 and f3 and black knights on a6 and h6. However, Volume C of Encyclopedia of Chess Openings (fifth edition, 2006) does not give errors or dubious moves to this game. It gives a move deserving attention (10. Bf2!?). This ECO Volume C also explores the following line deviating from that game played in 1983:

11. Qe2 {11. a3!? deserves attention} c4 12. h3 b5 {12. ... Bb4!? deserves attention} 13. g4 b4 14. Nd1 {white stands slightly better}

The link for chessdb.cn is:

Chess Cloud Database Query Interface

Regards from Spain.

Ajedrecista.

alchemist888
Posts: 8
Joined: Fri Mar 29, 2019 2:02 pm
Full name: Denis S.

Re: Busted French line?

Post by alchemist888 » Fri Oct 02, 2020 9:28 pm

Ajedrecista wrote:
Fri Oct 02, 2020 6:53 pm

This weird move can lead to even weirdest positions, like in Árnason - Petrosian of 1983:

Árnason - Petrosian (Tallinn, 1983)
Yeah, interesting game. But most likely 5. f4 is not the strongest reaction: it reduces DSB mobility, and doesn't support development; white center is not threatened at the moment, and they can calmly maneuver; f4 will be played one way or another, but at the right moment.
It looks like 5. a3! is the best move which also was played in both TCEC games.
With 5. a3 we prevent Bb4, we also no hurry to put the knight on f3 or e2, because in order to be able to Qg4 (attacking g7). As example 5...c5 is not great for black because 6.dxc5 Bxc5 7.Qg4! g6 8.Nf3 a6 9.Bd3 Nc6 10.0-0 h5 11.Qg3 Nh6 12.Bg5 Be7 13.Bxe7 Nxe7 14.Qf4 and white got a huge positional advantage.
Johansen - Mogstad, ICCF 2016 1-0 https://www.iccf.com/game?id=833258
So 5...b6 can be the best reaction on a3, but as we can see from the TCEC games, Black was not even close to the equation.

Madeleine Birchfield
Posts: 76
Joined: Tue Sep 29, 2020 2:29 pm
Location: Dublin, Ireland
Full name: Madeleine Birchfield

Re: Busted French line?

Post by Madeleine Birchfield » Fri Oct 02, 2020 9:33 pm

Tony P. wrote:
Fri Oct 02, 2020 5:19 pm
The line hasn't been explored deeply on chessdb.cn yet :( You can submit moves for analysis there and see what SF says at depth 22 (and also make a better use of transpositions than Homefish would). So far, the score is estimated to be between +50 and +70 (centipawns?).
chessdb.cn is the Chinese Chess/Xiangqi database, not the western/international chess database.

Alayan
Posts: 429
Joined: Tue Nov 19, 2019 7:48 pm
Full name: Alayan Feh

Re: Busted French line?

Post by Alayan » Fri Oct 02, 2020 9:49 pm

The API access to both is on the same website.

Tony P.
Posts: 203
Joined: Sun Jan 22, 2017 7:30 pm
Location: Russia

Re: Busted French line?

Post by Tony P. » Fri Oct 02, 2020 9:55 pm

Oops, I thought almost everyone on these forums was aware that it's for both Xiangqi and Western chess. It has a dedicated thread.

The switches between Xiangqi and chess, as well as between Chinese and English, are at the top right corner of the webpage. Also, the link given by Ajedrecista leads directly to the chess version in English.

supersharp77
Posts: 913
Joined: Sat Jul 05, 2014 5:54 am
Location: Southwest USA

Re: Busted French line?

Post by supersharp77 » Sat Oct 03, 2020 1:01 am

alchemist888 wrote:
Fri Oct 02, 2020 3:52 pm

There was a discussion in TCEC chat about this French line. Intuitively, this should be lost, since Black's can almost never play Nf5. And the typical counterplay with ...f6 is complicated by lack of development.
Of course, it takes a lot of analysis to prove that Black's is losing. But I was inspired by how the Grob opening was parsed on this forum and decide to create this topic.
Also both games from TCEC were losing for black
That would be great if someone share their analyzes and thoughts about 1.e4 e6 2.d4 d5 3. Nc3 Nf6 4.e5 Ng8?!
Well my take on this particular line is although it does lose some time by moving the knight back to the original square Ng8?! and loses valuable time..black can justify the loss of time due to the fact that with 4.e5 the position is "closed" at least for a while...Featured game has The Great GM Tigran Petrosian with the black Pieces.. :D :wink:



Pgn games file http://www.mediafire.com/file/hke41wpg1 ... gn.7z/file

My datebase seach did not pull up many played games with this line (less than 100) of the 90-100 games (some were duplicates) many white wins and alot of drawn games with a limited number of black wins..overall I think this line is wholly unexplored..Most well known players using Ng8 were Tigran Petrosian and David Bronstein

mmt
Posts: 296
Joined: Sun Aug 25, 2019 6:33 am
Full name: .

Re: Busted French line?

Post by mmt » Sat Oct 17, 2020 10:23 pm

I have the results of a long analysis with Oct 5 SF dev, nn-baeb9ef2d183.nnue, 450 GB of cache, most important 7-piece EGTBs. The TCEC game that SF lost was:


Before 5... b6 with multi PV 3:

75/96 96:59:49 8,561,685,310k 28,887k -1.09 1. ... Ne7 2.Nf3 b6 3.Ne2 c5 4.h4 h6 5.h5 Qd7 6.Rh3 Nbc6 7.c3 a5 8.Nf4 Ba6 9.Bxa6 Rxa6 10.a4 cxd4 11.cxd4 Nf5 12.g4 Nfe7 13.Bd2 Ra8 14.Kf1 Rb8 15.Kg2 Nc8 16.Rg3 Be7 17.Qe2 Bd8 18.Bc3 N8e7 19.Nd3 Na7 20.b3 Rc8 21.Qd2 Qc7 22.Rc1 Qd7 23.Qe3 Nac6 24.Bd2 Na7 25.Rh1 Qc6 26.Rb1 Qb7 27.Kh3 Nec6 28.Ra1 Be7 29.Rc1 Qd7 30.Kg2 Bd8 31.Nf4 Ne7 32.Ra1 Nac6 33.Nd3 Qb7 34.Rc1 Na7 35.Rb1 Nac6 36.Kh3 Na7 37.Qe2 Qd7 38.Kg2 Nac6 39.Rc1 Na7

76/6 96:59:49 8,561,685,310k 28,887k -1.02 1. ... Nd7 2.Nce2

76/88 96:59:49 8,561,685,310k 28,887k -1.00 1. ... b6 2.Nf3 c5 3.h4 Nc6 4.Ne2 h5 5.c3 Ba6 6.g3 g6 7.Bh3 Qd7 8.0-0 Nh6 9.Re1 Bxe2 10.Qxe2 c4 11.Bg5 Be7 12.Bf4 Nf5 13.a4 Na5 14.Nd2 Kf8 15.Qf3 Kg7 16.Reb1 Qc7 17.Rc1 Qd7 18.Qe2 Rag8 19.Kg2 Ra8 20.Re1 Rac8 21.Rh1 Kf8 22.Qf3 Qd8 23.Rhb1 Kg7 24.Kg1 Qc7 25.Qe2 Nh6 26.Re1 Nf5 27.Red1 Nc6 28.Rdb1 Na5 29.Kg2 Nh6 30.Re1 Nf5 31.Rh1 Qd7 32.Nf3 Nb3 33.Rae1 Rcg8 34.Bg5 Bxg5 35.hxg5 Ne7 36.Nh4 Qxa4 37.Qf3 Qd7 38.Qf6+ Kf8

After showing SF the line that was played in the game and letting it analyze them, it would play 7... Nc6 instead of 7... h6 (no multi PV):

84/82 22:21:27 2,440,217,800k 30,318k -0.88 1. ... Nc6 2.Ne2 h5 3.g3 Ba6 4.c3 g6 5.Bh3 Bh6 6.0-0 Bxc1 7.Rxc1 Nh6 8.Ng5 Rc8 9.Re1 Bxe2 10.Qxe2 c4 11.Rb1 Qe7 12.a4 Rc7 13.Qf3 Nf5 14.Bg2 Qf8 15.Qf4 Qe7 16.Rbd1 Na5 17.Ra1 Nc6 18.Rab1 Na5 19.Bf3 Kd8 20.Bd1 Rc8 21.Kg2 Kc7 22.Qd2 Qf8 23.Bc2 Kb8 24.Qf4 Qe7 25.Nh3 Nc6 26.Ra1 Na5 27.Reb1 Qd7 28.Ng5 Qe7 29.Qd2 Rc7 30.Re1 Rd7 31.Nf3 Ng7 32.Rab1 Ka8 33.Qg5 Qf8 34.Qf6 Rc7 35.Ra1 Kb8 36.Reb1 Ka8 37.Ng5 Kb8 38.Rf1 Ka8 39.Rfe1 Nb3 40.Rad1 Na5 41.Nh3 Kb8

and 9... Qd7 instead of Nec6

68/75 18:23:46 2,000,997,225k 30,214k -1.05 1. ... Qd7 2.Rh3 Bb7 3.c3 a5 4.Rh4 Nec6 5.Ng3 Ba6 6.Bxa6 Rxa6 7.Kf1 a4 8.Rg4 Na5 9.Qe2 Ra8 10.Be3 Nbc6 11.Kg1 Nc4 12.Rd1 b5 13.Kh1 cxd4 14.Bxd4 Qc7 15.Re1 Qb7 16.Kg1 Rc8 17.Nf1 Qd7 18.Ne3 Qb7 19.Rb1 Qc7 20.Rc1 Qb7 21.Re1 Qd7 22.Rd1 Qb7 23.Rg3 Qc7 24.Kh2 Qb7 25.Re1 Qc7 26.Rf1 Qd7 27.Kg1 Rd8 28.Rd1 Qc7 29.Re1 Rc8 30.Qc2 Rd8 31.Nxc4 dxc4 32.Rg4 Ne7 33.Be3 Rd5 34.Qe2 Nc6 35.Rd1 Nxe5 36.Nxe5 Qxe5 37.Rxd5

69/82+ 19:26:23 2,116,385,984k 30,241k -0.97 1. ... Qd7

If 9... Nec6 is played like in the game, the evals go bad, down to -1.22 and white probably can win it.

So the conclusion is that black has several different viable moves and this French line is not busted. However, it will be a tough game for black to hold.

Post Reply