Influence of engines on theory

[Roger de Coverly]

Those who punt the Kings Indian Attack from time to time will no doubt recognise this position.

SetUp "1"] [FEN "r1bqk2r/pp2bppp/2n1pn2/2pp4/4P3/3P1NP1/PPP1QPBP/RNB1K2R w KQkq - 1 7"]

The most popular move is 0-0 with e5 to follow. e5 could also be played immediately. That's what theory has said since the time of Ray Keene's Flank Openings, if not before. Also it was considered incorrect for White to try anything on the Queenside.

In his tenth round game in the 4NCL against Lupulescu, Adams played 7. a4 which is either a new move or a newish one. Giri played it against Bacrot in last year's FIDE Grand Swiss. Checking on what stockfish thought, it also suggested a4 and recommended Adams' next few moves of 0-0, Re1 and h4.

That's a problem for the veteran player, you can find that you have to unlearn theory. Adams didn't ever play e5 either. Black took on e4 eventually.

I've found in my own games that for all the attractive attacks after e5 in Flank Openings, you cannot actually force any of them through. The attack then runs out and you lose to a queenside invasion. One of my recent losses was against Paul Littlewood in last year's English Seniors. After the game in which his queenside invasion was well ahead of any mating theats, he mentioned that his father had told him that such positions were best avoided for Black.

[John Upham]

I get the impression that even modern engines have a poor "understanding" of closed positions and openings such as the KIA that can remain closed for some time. To my mind they appear to be somewhat pessimistic and provide assessments that are difficult to make anything of.

For so-called open games or positions with at least one open file or files then the assessment perhaps become more realistic and reliable.

I'd like to ask Matthew Sadler about this!

[Tim Spanton]

Presumably the point of 7.a4 is to slow Black's queenside counterplay, rather than playing on the queenside per se (I realise we are getting into the world of semantics here), before switching to White's usual kingside play.

[Roger de Coverly]

Presumably the point of 7.a4 is to slow Black's queenside counterplay, rather than playing on the queenside per se (I realise we are getting into the world of semantics here), before switching to White's usual kingside play.

Adams had several opportunities to play e5 and start the usual trundle of the White pieces towards the Black King. So h2-h4 would be played and possibly it would go to h6 if not blocked, Also the c1 Bishop would go to f4 with over protection of e5. Then there would be a Knight's tour, Nb1-d2-f1-h2-g4 to add more over protection and hopefully join a mating attack. You can play Bg2-h3 with possible e6 sacrifices, also Kg1-g2 and Re1-h1 to add the Rook to the attack.

I've concluded that this approach doesn't always work and that the plan of Giri and Adams of just playing in the centre may be preferred, less risky anyway as control can more easily be maintained. In the event, Black took on e4 and exchanged some pieces.

[Joey Stewart]

I find it's a really difficult opening to use as white, so easy for black to close down the position, equalise and then steadily outplay you on the queenside - the attack takes absolutely ages to come against blacks king.

If you want to "ruin" a Sicilian game there are much safer lines for white like all that Bb5 stuff that almost everybody on the internet is playing these days and has presumably been given the green light by the silicon overlords of opening theory.

[Roger de Coverly]

If you want to "ruin" a Sicilian game there are much safer lines for white like all that Bb5 stuff that almost everybody on the internet is playing these days and has presumably been given the green light by the silicon overlords of opening theory.

I think authors who advocate playing 3 Bb5 against 2. .. Nc6 and 2. ..d6 suggest 3. g3 or 3. d3 against 2. .. e6. which was how the Adams game started.

The sequence 1. e4 c5 2. Nf3 e6 3. g3 d5 4. exd5 exd5 5. d4 Bg4 6. Bg2 Nc6 7. 0-0 Nxd4 looks very good for Black until you realise that Qe1 can be played.
[SetUp "1"] [FEN "r2qkbnr/pp3ppp/8/2pp4/3n2b1/5NP1/PPP2PBP/RNBQ1RK1 w kq - 0 8"]

[MartinCarpenter]

I get the impression that even modern engines have a poor "understanding" of closed positions and openings such as the KIA that can remain closed for some time. To my mind they appear to be somewhat pessimistic and provide assessments that are difficult to make anything of.

For so-called open games or positions with at least one open file or files then the assessment perhaps become more realistic and reliable.

I'd like to ask Matthew Sadler about this!

I'd not be sure. Certainly from poking engine vs engine games etc on some very obscure French defence positions the modern NNE versions of Stockfish seem to at least have a vastly improved understanding of what is going on.

Watching them take apart earlier engines can be quite amusing