Sorry, but the whole Aronin-Taimanov variation 7...Nc6 seems dubious! 
The modern trendy KID is bad for black
Sort:
#2
7...Nd7 gives white the same queen's side advantage, but without the black attack on the white king. 7...exd4 cedes the center. Both are even more dubious than 7...Nc6. Look at the corresponding engine evaluations and statistics.
#1
Yes, the Bayonet Attack probably refutes the King's Indian Defence on a theoretical level. Kasparov gave up on the King's Indian Defence after Kramnik pummeled him. Radjabov gave up too. Osipov tried to play it in the ICCF correspondence World Championship, but suffered some losses.
On a practical level it is still playable at GM level especially as a surprise.
In short time controls even more.
It is still good to try and win as black even at greater risk of losing.
'King's Indian Defence is riskier for black than King's Gambit for white' - Bronstein
Maybe one day someone will find an improvement and the KID will be revived again at the top level...
Aug 11, 2022
0
#5
I recently saw some compelling analysis that argued in favor of black being okay after 9…Nh5 10. Re1 a5 11. bxa5 Rxa5 12. Nd2 f5, which is a sharp and fairly unexplored pawn sac. And 12. a4 f5 as well.
Aug 11, 2022
0
#6
The masters database does not agree with you, on chess.com nor on chessbase.
(speaking of chessbase, it's about time i started updating my megabase)
Sure, it scores slightly worse than it should for black, but it's not that dubious.
Aug 11, 2022
0
#7
tygxc wrote:
#2
7...Nd7 gives white the same queen's side advantage, but without the black attack on the white king. 7...exd4 cedes the center. Both are even more dubious than 7...Nc6. Look at the corresponding engine evaluations and statistics.
7...Na6 is an interesting idea that seems to be getting more popular recently.
The same illiterate comments about KID being "bad for Black".
He has a choice between 9...a5 and 9...Nh5 10.Re1 f5 11.a4 a5. He is fine in both lines.
Aug 12, 2022
0
#9
dpnorman wrote:
I recently saw some compelling analysis that argued in favor of black being okay after 9…Nh5 10. Re1 a5 11. bxa5 Rxa5 12. Nd2 f5, which is a sharp and fairly unexplored pawn sac. And 12. a4 f5 as well.
Thanks for pointing it out! 😎 However, SF14+ suggests 12. a4 (ev=+1.1) instead of 12. Nd2.
#9
Here is how Radjabov fared with 9...a5.
https://www.chessgames.com/perl/chessgame?gid=1994975
And here is how Radjabov fared with 9...Nh5 10 Re1 f5 11 Ng5
https://www.chessgames.com/perl/chessgame?gid=1688946
Aug 12, 2022
0
#11
Yigor wrote:
dpnorman wrote:
I recently saw some compelling analysis that argued in favor of black being okay after 9…Nh5 10. Re1 a5 11. bxa5 Rxa5 12. Nd2 f5, which is a sharp and fairly unexplored pawn sac. And 12. a4 f5 as well.
Thanks for pointing it out! 😎 However, SF14+ suggests 12. a4 (ev=+1.1) instead of 12. Nd2.
This is a joke right
Yigor wrote:
dpnorman wrote:
I recently saw some compelling analysis that argued in favor of black being okay after 9…Nh5 10. Re1 a5 11. bxa5 Rxa5 12. Nd2 f5, which is a sharp and fairly unexplored pawn sac. And 12. a4 f5 as well.
Thanks for pointing it out! 😎 However, SF14+ suggests 12. a4 (ev=+1.1) instead of 12. Nd2.
12.a4 just transposes to the 9...Nh5 10.Re1 f5 11.a4 a5 12.bxa5 Rxa5 line which is fine for Black, but with Black having avoided 12.b5! which is a bit annoying.
Aug 12, 2022
0
#13
tygxc wrote:
#2
7...Nd7 gives white the same queen's side advantage, but without the black attack on the white king. 7...exd4 cedes the center. Both are even more dubious than 7...Nc6. Look at the corresponding engine evaluations and statistics.
#1
Yes, the Bayonet Attack probably refutes the King's Indian Defence on a theoretical level. Kasparov gave up on the King's Indian Defence after Kramnik pummeled him. Radjabov gave up too. Osipov tried to play it in the ICCF correspondence World Championship, but suffered some losses.
On a practical level it is still playable at GM level especially as a surprise.
In short time controls even more.
It is still good to try and win as black even at greater risk of losing.
'King's Indian Defence is riskier for black than King's Gambit for white' - Bronstein
Engines are TERRIBLE at evaluation positions during the opening phase of closed openings.
The 7...exd4 line has been the trend of late and Black is ok, but he needs to find a number of only moves. Ntrilis did an extensive article on this in NIC Yearbook maybe a year ago.
The problem with 7...Nc6 these days is not in the Bayonet. It is in the line 9.Ne1 Nd7 10.Be3 f5 11.f3 f4 12.Bf2 g5 13.Rc1! Where 13...a6, 13...Ng6, and 13...Rf6 are all unsatisfactory for Black!
Black has satisfactory answers to both the Bayonet and Korchnoi's 13.a4, which has often been dubbed the "main line", if there exists such a thing in the King's Indian
#14
"It is in the line 9.Ne1 Nd7 10.Be3 f5 11.f3 f4 12.Bf2 g5 13.Rc1"
++ Here is an ICCF game:
https://www.iccf.com/game?id=948246
"The problem with 7...Nc6 these days is not in the Bayonet."
++ See the 2 Radjabov games above #11.
Aug 13, 2022
0
#15
Here are two less biased examples of 9...a5 are two games of Jones from same tournament a win and a loss
https://www.chessgames.com/perl/chessgame?gid=2135231
https://www.chessgames.com/perl/chessgame?gid=2137494
It's why very strong GM's still play KingsIndian, you get winning chances. Jones initially dealt very well with a new idea by Gelfand. Jones is a strong GM, but Gelfand probably just has a bit more quality. It's main reason tend to lose games, rather than an opening being refuted. Used to be interested in 9...Nh5 10Re1 a5 but it seens 9...a5 10pxp Rxp 11a4 leads to same sort of position and can tweak where put knight. Black can delay/ prevent c5 break with b6 or c5 themselves before looking for counter-play with f5.
Aug 13, 2022
0
#16
tygxc wrote:
#14
"It is in the line 9.Ne1 Nd7 10.Be3 f5 11.f3 f4 12.Bf2 g5 13.Rc1"
++ Here is an ICCF game:
https://www.iccf.com/game?id=948246
"The problem with 7...Nc6 these days is not in the Bayonet."
++ See the 2 Radjabov games above #11.
So you think 1 Correspondence Game suddenly makes 13.Rc1 good for Black?
You think 2 games constitute a Refutation?
SMH!
@17
"So you think 1 Correspondence Game suddenly makes 13.Rc1 good for Black?"
++ A correspondence game represents a few years of human grandmaster analysis and engine calculation. If you assert 13 Rc1 is good for white, then you have to find an improvement over the ICCF WC game.
"You think 2 games constitute a Refutation?" ++ Radjabov gave up on the King's Indian after that, like Kaparov gave up on it after his 2 losses to Kramnik.
Aug 13, 2022
0
#18
tygxc wrote:
@17
"So you think 1 Correspondence Game suddenly makes 13.Rc1 good for Black?"
++ A correspondence game represents a few years of human grandmaster analysis and engine calculation. If you assert 13 Rc1 is good for white, then you have to find an improvement over the ICCF WC game.
"You think 2 games constitute a Refutation?" ++ Radjabov gave up on the King's Indian after that, like Kaparov gave up on it after his 2 losses to Kramnik.
Our wonderful omniscient @ThrillerFan appreciates only his own opinion. 🤣
Aug 13, 2022
0
#19
You think that Radjabov's rapid game against Anand is anything approaching best play for black? Just allowing Bxh5, and Qxh5.
The most popular and trendy Aronin-Taimanov variation: Bayonet attack:
is bad for black both by engine evaluations (ev=+0.8/+1.0) and statistically. For example, the Lichess master database, restricted to the year 2021, shows 36% of white wins and only 8% of black wins.
