Computer Chess Openings: Modern defense

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

Moderator: Ras

User avatar
Graham Banks
Posts: 46076
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Computer Chess Openings: Modern defense

Post by Graham Banks »

lkaufman wrote:There is a reason that the Elite GMs regularly play 1...e5 against each other and almost never 1...g6. It is a better move. But in my opinion not nearly as much better as the above evals imply.
Both Komodo and Stockfish are tuned for best results, and these results are achieved with evals that wildly overstate transitory positional factors. I don't fully uonderstand why this is so or what if anything we can do about it, but I'm open to good ideas.
Hi Larry,

in the online Chessbase database, 1...g6 almost has the best success rate against e4, narrowly edged by only the Sicilian. The result percentage in the table below refers to White's success rate against each move.
Of course, other online databases may differ.

Perhaps you're right about it not being preferred so much by elite GMs - not sure.

Graham.

Image
gbanksnz at gmail.com
Sven
Posts: 4052
Joined: Thu May 15, 2008 9:57 pm
Location: Berlin, Germany
Full name: Sven Schüle

Re: Computer Chess Openings: Modern defense

Post by Sven »

Graham Banks wrote:in the online Chessbase database, 1...g6 almost has the best success rate against e4
For training I use a database containing only games where both players have a rating >= 2400. There 1...g6 has a much higher success rate for white (about 58%) compared to 1...e5, 1...c5, 1...e6 or other standard openings which are usually in the range of 54% .. 55%. A higher success rate (for white) does not necessarily mean that it is a "losing" move for black, or that black has already a disadvantage. For me it just says that the move might increase the probability of playing positions that are difficult for black and can already be lost by playing one "suboptimal" move.

Also, success rates suffer from including an unknown portion of "bad moves" from the past that were preferred for some time until someone found a refutation. And finally, in some cases success rates suffer from losing won (or drawn) positions later in the game due to other reasons (blunder, time, ...).

Therefore I never really trust success rates.
jdart
Posts: 4435
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: Computer Chess Openings: Modern defense

Post by jdart »

I think of the Modern as being A41-A42, while you are including Pirc lines too.

IMO among the more difficult lines for Black are those in which White gets in an early h3, so that Black does not have .. Bg4.

Some GMs even go h3 as early as move 4, for example:

1. d4 g6 2. Nf3 Bg7 3. e4 d6 4. h3

--Jon
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Computer Chess Openings: Modern defense

Post by carldaman »

jdart wrote:I think of the Modern as being A41-A42, while you are including Pirc lines too.

IMO among the more difficult lines for Black are those in which White gets in an early h3, so that Black does not have .. Bg4.

Some GMs even go h3 as early as move 4, for example:

1. d4 g6 2. Nf3 Bg7 3. e4 d6 4. h3

--Jon
A41-42 and also B06 are the usual ECO codes assigned to the Modern Defense.
B07-B09 is the Pirc Defense.
Henk
Posts: 7261
Joined: Mon May 27, 2013 10:31 am

Re: Computer Chess Openings: Modern defense

Post by Henk »

End conclusion: modern defense is perfectly playable.

[d] r1bqk1nr/pppnppbp/3p2p1/8/2BPP3/5N2/PPP2PPP/RNBQK2R w KQkq - 2 5
jdart
Posts: 4435
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: Computer Chess Openings: Modern defense

Post by jdart »

Ok, but 1. e4 g6 2. d4 d6 3. f3 Nf6 4. c4 Bg7 5. Ne2 (which a bunch of Graham's line begin with) is a Saemisch King's Indian (E83).

--Jon
User avatar
Nordlandia
Posts: 2845
Joined: Fri Sep 25, 2015 9:38 pm
Location: Sortland, Norway

Re: Computer Chess Openings: Modern defense

Post by Nordlandia »

Why is that Carlsen never plays or face najdorf variation?
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Computer Chess Openings: Modern defense

Post by carldaman »

jdart wrote:Ok, but 1. e4 g6 2. d4 d6 3. f3 Nf6 4. c4 Bg7 5. Ne2 (which a bunch of Graham's line begin with) is a Saemisch King's Indian (E83).

--Jon
Yes, playing an early Nf6 usually transposes to other openings -- and thus no longer called the Modern.

CL
User avatar
Graham Banks
Posts: 46076
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Computer Chess Openings: Modern defense

Post by Graham Banks »

carldaman wrote:
jdart wrote:Ok, but 1. e4 g6 2. d4 d6 3. f3 Nf6 4. c4 Bg7 5. Ne2 (which a bunch of Graham's line begin with) is a Saemisch King's Indian (E83).

--Jon
Yes, playing an early Nf6 usually transposes to other openings -- and thus no longer called the Modern.

CL
Having some Pirc lines and KID lines thrown in amongst the Modern Defense lines is a bonus. :)
gbanksnz at gmail.com
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Computer Chess Openings: Modern defense

Post by carldaman »

Graham Banks wrote:
carldaman wrote:
jdart wrote:Ok, but 1. e4 g6 2. d4 d6 3. f3 Nf6 4. c4 Bg7 5. Ne2 (which a bunch of Graham's line begin with) is a Saemisch King's Indian (E83).

--Jon
Yes, playing an early Nf6 usually transposes to other openings -- and thus no longer called the Modern.

CL
Having some Pirc lines and KID lines thrown in amongst the Modern Defense lines is a bonus. :)
The more the merrier! :D