Inspired by mr Berger, anyway, starting another -more simple- thread as my other postings clearly
were beyond the ability to understand by most chess or computer kiddies here.
Almost in quora style: what's the Best opening ?
Answer, well it's not 1.g4, and also not 1.f3
Continuing, i've done research on this for many years, and while originally it was often
said that only GM's could make accurate evaluations (eg. of end positions after an opening line)
in recent years with the Nnue eval it's now possibly to make judgments about opening lines
with computer analysis as well. Note that in the past i also looked at GM evals like
an opening database with GM evals in the program Chess Assistant, yes i have a lot
of experience in this field (that's why i'm not a nine year old chess kiddie anymore).
As result of this computer chess revolution (nnue's) opening theory was slightly, but
not heavily modified (hey, maybe the Gm's were not so wrong after all with their
earlier human evals, contrary to some earlier comps with more inaccurate evals).
While earlier on opening theory had been described (already pretty accurately) in
the GM repertoire series by Quality chess,
https://qualitychess.co.uk/sections/1/opening/
some resulting modifications of the Nnue revolution were published in series of books by Thinkers Publishing
in Belgium, (with varying quality btw), with names as the 'Modernized' (Ruy Lopez, Slav, etc etc.
https://thinkerspublishing.com/product- ... ng-theory/
In total almost 35 books or so (sometimes called 'Revisited, instead of Modernized. I've looked at
many of them and compared it with my own analysis which often was running parallel with such findings.
After 1.e4 the move 1...e5 is a solid defense. While after 2.Nf3! the Petroff (2..Nf6) is
a solid defense, usually 2...Nc6! is played, after which White can best play 3.Bc4 or Bb5
After 3.Bc4 the Giuoco Piano or Pianissimo, acording to the book the Modernized Italian
some innovations and more modern White strategies were found (eg. Nbd2 and then
Ndf1 in Ruy Lopez style, but if Black is aware of such methods Black can keep the draw
ofcourse. After 3.Bb5, the Ruy Lopez (RL), we get an opening already analyzed for centuries.
It was known that the Berlin and RL Marshall countergambit (give Black sufficient drawing
resources, nevertheless there are books as Modernized Berlin, and Modernized Marshall.
https://thinkerspublishing.com/product- ... ng-theory/
Also, after 1.e4 the Sicilian 1..c5 is quite solid, as already shown by Euwe, but more sharp (*).
While we know that Najdorf (with ..d6) is most solid for Black, chess kiddies as Magnus
C when he was young got good results with the (sharp) Dragon (...g6). Yes there also
is a book about the Modernized Dragon.
Continuing with the Najdorf, while for years either Bg5 (classical) or Be3 (the English attack)
were considered to be strong, in more recent years attempts were made with an early h3
move, preparing an early g4 (as eg also in the Keres attack against)the Sicilian Scheveningen.
This move (usually 6) .h3 is called the Adams attack btw, after English GM Mike Adams.
While theory changed a bit over the years, at high levels GM's look closely to the
developments of such theory, usually with Chessbase, and found that again the move ...e5
for Black (as in the conventional Najdorf against Be3) is the most solid and Black ofcourse
can always keep a draw (although now i'm repeating myself, i guess).
So instead of 1.e4 you might like to look at 1.d4, at least if someone think of getting
a slight advantage for White (*) after the opening (something which in the end will turn
out to be impossible, in line with the 0.0 NNue eval for the start position btw).
In line with the question by P.Berger about Kings Indian and Gruenfeld.
More about this in my next posting.
(*) as posted already a few times earlier on my blog and in this crazy forum,
for a proper study of opening theory you don't only need to look at the eval,
but also at the sharpness of a position. That's why gambits exist, and they
sometimes (**) can give better results in practical -human- play.
(**) in fact often, at lower levels of (otb) play
to be continued
(with first 1.d4, and then 1.Nf3)
PS all written down by yours truly in detail btw (for various levels of play)
www.amazon.com/Learning-Chess-Openings- ... B00WKE1HJC
so what's the best chess opening ?
Moderator: Ras
-
jefk
- Posts: 1025
- Joined: Sun Jul 25, 2010 10:07 pm
- Location: the Netherlands
- Full name: Jef Kaan
-
Jouni
- Posts: 3655
- Joined: Wed Mar 08, 2006 8:15 pm
- Full name: Jouni Uski
Re: so what's the best chess opening ?
According to chessdb all moves draw easily, except 1. g4 1. f3 and possible 1.Na3 and 1.Nh3.
Jouni
-
jefk
- Posts: 1025
- Joined: Sun Jul 25, 2010 10:07 pm
- Location: the Netherlands
- Full name: Jef Kaan
Re: so what's the best chess opening ?
yes all first moves theoretically lead to a draw (but not 'easily' for a human otb) except 1.g4? (after which
is for a non-Gm still also not always easy to win with Black ) but nevertheless, in the past chess playing have
been searching for a possibility of finding a tiny opening advantage for White after the first move.
Continuing my discourse about searching for a possible opening advantage for White,
it's clear that 1.e4 cannot lead to a forced win (despite Fischer's comment 'best by test').
Note that the following summary about opening theory is well known even by many amateurs,
but i post it nevertheless, both for those not so aquainted with opening theory, and for those
still thinking that it might be possible for White to get a fundamental advantage in all situations.
Thus let's look first now at the move 1.d4, a move usually thought to lead to more 'positional' play;
but there are many exceptions (and sharp variations, sometimes even difficult to
avoid by Black (with 1..c5 Black can make a choice to go for sharp variations as eg.
the Benko gambit or Benoni, but this would go at the cost of counteracting a possible
slight or tiny opening advantage for White). After 1...e6?! White usually plays 2.e4, which leads
to the French opening; although Black then also can hold the position, it's often not easy to play.
So fundamentally 'best' for Black now after 1.d4 are 1...d5 or 1...Nf6!
Note that i have made evaluations of (the end positions) of such variations with engines
as Obsidian (based on Lc0 evals) which often give more nuanced evals than SF and the
Chinese database with almost always simply give a 0.0 result, making understanding (and
well thought out opening choice) more difficult.
After 1...d4 White can play 2.Nf3, or 2.c4, leading to positions which often transpose.
After 2.c4 we get the socalled queens gambit, which is a pseudo gambit because after
the move 2...dxc4?! Black can never keep possession of the extra pawn. While the Queens
gambit accepted QGD can lead to interesting and sharp positions, it's known for quite some time that
the moves 2...c6! and 2...e6 are more solid. The first leads to the Slav defense, and the latter one
to the Queens Gambit Declined, Nimzo-Indian or QGD or Catalan, or SemiSlav (after a later ...c6).
Have a look at eg. ref 1. Note that an early Nc3?! by White, enabling an early Bb4 by Black, the Nimzo-Indian
usually leads to very solid defense lines for Black, and thus it's usually better (as known by many GM's) to first play Nf3.
have a look at eg ref 2. Problem with the move 2...e6?! is that White can go for the Catalan, preferably by
immediately playing the move 3.g3! (instead of 3.Nf3). Reason that this is preferred is because of the Bogo Indian,
if Black would play 3...Bb4+, after which White can play 4.Nbd2! followed by first playing Bg2
and only later the move Nf3. An important distinction, whereby White in theory can maintain
a tiny opening advantage with the Catalan. Not for nothing the Catalan is the preferred opening
in the GM opening repertoie series, described in book nr one by GM Boris Avrukh (3).
Black however, can simply avoid the Catalan by playing 2...c6. Then after 3.Nf3 e6 (or Nf6! or 3...dxc4 the
pure Slav) we get a lot of variations/positions as Semi Slav (with 3...Nf6 and 4...e6). While there are sharp
positions possible after Bg5 and then the Botvinnik or anti-Moscow gambits, it has been shown White
can never gain a fundamental advantage against such defensive variations. Also with e3 instead
fof Bg5, the anti-Meran, Black can hold on it's own and there are a number of good books
by Thinkers Publishing describing in detail such d4 variations (eg. ref 4)
After the (fundamentally) -slightly-better move for Black 1...Nf6! the openings can transpose
to the QGD or SemiSlav positions, but in addition there now also are options with ...g6, socalled
black fianchetto openings as Gruenfeld (or the slightly worse Kings Indian with d6). While Kings
indian used to be an imbalanced defensive with strategic resources for Black in particular a
king's attack, it nowadays has been analyzed in enormous detail (yets also in a book by TP) and
thus at GM level isn't played so often anymore. The Gruenfeld defense (with ...d5), however, remains
extremely solid, and 'fundamentally'' the best for Black , at least at higher levels because Black needs
to know quite precisely how to play. Have a look at the Modernized Gruenfeld, for example (5).
Concluding, also with 1.d4 there's not fundamental opening advantage for White.
So maybe White should try 1.c4, the English instead of 1.d4 to gain an opening advantage ?
Well not really, after 1...e5 the reversed Sicilian, Black can maintain equality. Not always
easy, because some of the reversed Dragon positions with a later g3 for White are
not so easy to handle for Black, but nevertheless. Note that Sicilian in itself already
is a strong defense for Black, thus a reversed Sicilian even with an extra half tempo White
in itself is not sufficient for a fundamental advantage for White. see books such as in ref 6:
Then, maybe after the amazing opening choice 1.Nf3!! the Reti, avoiding the Gruenfeld
(because after 1.Nf3 g6?! is not so good for Black anymore) White may still be able
to gain a slight opening advantage ? ok, I will discuss this in the next posting.
(hint: while the Reti can avoid the -solid- Gruenfeld, it also forsakes any advantage
in the Catalan because now Nf3 has been played before Bg2, thus making the little
trick against the Bogo-Indian with Nbd2! (as with 1.d4) impossible....
Thus usually the Reti will transpose to QGd, unless Black plays 1...c5! after which
White can play 2.e4, transposing to Sicilian, or 2.c4, the Symmetrical English.
In both situations Black can hold the position(s), believe it or not (**).
references
1) https://thinkerspublishing.com/product/ ... zo-indian/
2) https://www.amazon.nl/Complete-Opening- ... 9464201932
3) https://qualitychess.co.uk/products/1/2 ... is-avrukh/
4) https://thinkerspublishing.com/product/ ... -pavlovic/
5) https://thinkerspublishing.com/product/ ... d-defense/
6) https://qualitychess.co.uk/products/1/7 ... hardcover/
(**) but isn't it possible that suddenly at move 114 in the Najdorf Sicilian with Bg5 after a few new moves
a new position is arising which suddenly turns out to be a win eg. checkmate for White in 187 moves ?
Well no, not really. When investigating whether the Earth isn't flat it might be also of course 'possible' that
suddenly at Antartica people find a whole (not found with the 3d computer map) after which you can
gain access to hollow earth and meet some aliens who have lived there for a million years and subsequently
telling you your 3D model is a hallucination and you actually live within 6d space with four spatial dimensions
and two time dimensions. Yes that is not impossible. But just as implausible as finding a White win
in all lines from the start position. In fact even more plausible, but now i'm digressing into Ufology (lol)
is for a non-Gm still also not always easy to win with Black ) but nevertheless, in the past chess playing have
been searching for a possibility of finding a tiny opening advantage for White after the first move.
Continuing my discourse about searching for a possible opening advantage for White,
it's clear that 1.e4 cannot lead to a forced win (despite Fischer's comment 'best by test').
Note that the following summary about opening theory is well known even by many amateurs,
but i post it nevertheless, both for those not so aquainted with opening theory, and for those
still thinking that it might be possible for White to get a fundamental advantage in all situations.
Thus let's look first now at the move 1.d4, a move usually thought to lead to more 'positional' play;
but there are many exceptions (and sharp variations, sometimes even difficult to
avoid by Black (with 1..c5 Black can make a choice to go for sharp variations as eg.
the Benko gambit or Benoni, but this would go at the cost of counteracting a possible
slight or tiny opening advantage for White). After 1...e6?! White usually plays 2.e4, which leads
to the French opening; although Black then also can hold the position, it's often not easy to play.
So fundamentally 'best' for Black now after 1.d4 are 1...d5 or 1...Nf6!
Note that i have made evaluations of (the end positions) of such variations with engines
as Obsidian (based on Lc0 evals) which often give more nuanced evals than SF and the
Chinese database with almost always simply give a 0.0 result, making understanding (and
well thought out opening choice) more difficult.
After 1...d4 White can play 2.Nf3, or 2.c4, leading to positions which often transpose.
After 2.c4 we get the socalled queens gambit, which is a pseudo gambit because after
the move 2...dxc4?! Black can never keep possession of the extra pawn. While the Queens
gambit accepted QGD can lead to interesting and sharp positions, it's known for quite some time that
the moves 2...c6! and 2...e6 are more solid. The first leads to the Slav defense, and the latter one
to the Queens Gambit Declined, Nimzo-Indian or QGD or Catalan, or SemiSlav (after a later ...c6).
Have a look at eg. ref 1. Note that an early Nc3?! by White, enabling an early Bb4 by Black, the Nimzo-Indian
usually leads to very solid defense lines for Black, and thus it's usually better (as known by many GM's) to first play Nf3.
have a look at eg ref 2. Problem with the move 2...e6?! is that White can go for the Catalan, preferably by
immediately playing the move 3.g3! (instead of 3.Nf3). Reason that this is preferred is because of the Bogo Indian,
if Black would play 3...Bb4+, after which White can play 4.Nbd2! followed by first playing Bg2
and only later the move Nf3. An important distinction, whereby White in theory can maintain
a tiny opening advantage with the Catalan. Not for nothing the Catalan is the preferred opening
in the GM opening repertoie series, described in book nr one by GM Boris Avrukh (3).
Black however, can simply avoid the Catalan by playing 2...c6. Then after 3.Nf3 e6 (or Nf6! or 3...dxc4 the
pure Slav) we get a lot of variations/positions as Semi Slav (with 3...Nf6 and 4...e6). While there are sharp
positions possible after Bg5 and then the Botvinnik or anti-Moscow gambits, it has been shown White
can never gain a fundamental advantage against such defensive variations. Also with e3 instead
fof Bg5, the anti-Meran, Black can hold on it's own and there are a number of good books
by Thinkers Publishing describing in detail such d4 variations (eg. ref 4)
After the (fundamentally) -slightly-better move for Black 1...Nf6! the openings can transpose
to the QGD or SemiSlav positions, but in addition there now also are options with ...g6, socalled
black fianchetto openings as Gruenfeld (or the slightly worse Kings Indian with d6). While Kings
indian used to be an imbalanced defensive with strategic resources for Black in particular a
king's attack, it nowadays has been analyzed in enormous detail (yets also in a book by TP) and
thus at GM level isn't played so often anymore. The Gruenfeld defense (with ...d5), however, remains
extremely solid, and 'fundamentally'' the best for Black , at least at higher levels because Black needs
to know quite precisely how to play. Have a look at the Modernized Gruenfeld, for example (5).
Concluding, also with 1.d4 there's not fundamental opening advantage for White.
So maybe White should try 1.c4, the English instead of 1.d4 to gain an opening advantage ?
Well not really, after 1...e5 the reversed Sicilian, Black can maintain equality. Not always
easy, because some of the reversed Dragon positions with a later g3 for White are
not so easy to handle for Black, but nevertheless. Note that Sicilian in itself already
is a strong defense for Black, thus a reversed Sicilian even with an extra half tempo White
in itself is not sufficient for a fundamental advantage for White. see books such as in ref 6:
Then, maybe after the amazing opening choice 1.Nf3!! the Reti, avoiding the Gruenfeld
(because after 1.Nf3 g6?! is not so good for Black anymore) White may still be able
to gain a slight opening advantage ? ok, I will discuss this in the next posting.
(hint: while the Reti can avoid the -solid- Gruenfeld, it also forsakes any advantage
in the Catalan because now Nf3 has been played before Bg2, thus making the little
trick against the Bogo-Indian with Nbd2! (as with 1.d4) impossible....
Thus usually the Reti will transpose to QGd, unless Black plays 1...c5! after which
White can play 2.e4, transposing to Sicilian, or 2.c4, the Symmetrical English.
In both situations Black can hold the position(s), believe it or not (**).
references
1) https://thinkerspublishing.com/product/ ... zo-indian/
2) https://www.amazon.nl/Complete-Opening- ... 9464201932
3) https://qualitychess.co.uk/products/1/2 ... is-avrukh/
4) https://thinkerspublishing.com/product/ ... -pavlovic/
5) https://thinkerspublishing.com/product/ ... d-defense/
6) https://qualitychess.co.uk/products/1/7 ... hardcover/
(**) but isn't it possible that suddenly at move 114 in the Najdorf Sicilian with Bg5 after a few new moves
a new position is arising which suddenly turns out to be a win eg. checkmate for White in 187 moves ?
Well no, not really. When investigating whether the Earth isn't flat it might be also of course 'possible' that
suddenly at Antartica people find a whole (not found with the 3d computer map) after which you can
gain access to hollow earth and meet some aliens who have lived there for a million years and subsequently
telling you your 3D model is a hallucination and you actually live within 6d space with four spatial dimensions
and two time dimensions. Yes that is not impossible. But just as implausible as finding a White win
in all lines from the start position. In fact even more plausible, but now i'm digressing into Ufology (lol)
-
Jouni
- Posts: 3655
- Joined: Wed Mar 08, 2006 8:15 pm
- Full name: Jouni Uski
Re: so what's the best chess opening ?
Addition. It's surprising, that 1.f4 is draw also! Fritz book has "?" for move. Main line 1. f4 d5 2. Nf3 Bg4 3. e3 Nf6 4. h3 Bxf3 5. Qxf3 e6 draw.
Jouni
-
jefk
- Posts: 1025
- Joined: Sun Jul 25, 2010 10:07 pm
- Location: the Netherlands
- Full name: Jef Kaan
Re: so what's the best chess opening ?
1.f4?! is moving towards the (probable) Black kings position but
gives up center control.
It can be an interesting move sometimes for gambiteers however.
(with the awesome P3 engine i've seen quite some amazing games
with 1.f4 d5 2.c4!? the awesome Sturm gambit !
Sturm und Drang !
gives up center control.
It can be an interesting move sometimes for gambiteers however.
(with the awesome P3 engine i've seen quite some amazing games
with 1.f4 d5 2.c4!? the awesome Sturm gambit !
Sturm und Drang !
-
Peter Berger
- Posts: 751
- Joined: Thu Mar 09, 2006 2:56 pm
Re: so what's the best chess opening ?
I have no idea what a "best opening" is supposed to mean exactly, other than in the context of a proof that chess actually is a win for white.
Same as you I suspect that chess is a draw though, and that all white moves besides 1. g4 will lead to a draw with optimal play. And in response black has to choose some setup that doesn't lose, so there should be several "best opening"s.
In practice this looks a bit different though. You'll like to choose openings that maximize your EV. In these days of strong chess engines you'll probably want to choose from openings mainly that are not lost, but other than that optimal choices will differ between different chess entities IMHO.
There are some eternal discussions among the strongest human chessplayers with the aid of powerful chess engines on good and worthwhile setups FOR THEM, but let's face it: we are not in a position to really participate.
It is somewhat easier when it is about pure chessengines, as we can try out things at home and judge by results. And as it is a reasonable assumption that most correspondence players consult Stockfish for their moves, we can probably draw some conclusions that would apply to correspondence chess, too.
But even in correspondence chess there are practical considerations. An opening setup like the Jobava London which most likely is not the very best conceivable may still look attractive, just as the best setup for black is not quite clear, so that players and engines may have a harder task to counter it.
Same as you I suspect that chess is a draw though, and that all white moves besides 1. g4 will lead to a draw with optimal play. And in response black has to choose some setup that doesn't lose, so there should be several "best opening"s.
In practice this looks a bit different though. You'll like to choose openings that maximize your EV. In these days of strong chess engines you'll probably want to choose from openings mainly that are not lost, but other than that optimal choices will differ between different chess entities IMHO.
There are some eternal discussions among the strongest human chessplayers with the aid of powerful chess engines on good and worthwhile setups FOR THEM, but let's face it: we are not in a position to really participate.
It is somewhat easier when it is about pure chessengines, as we can try out things at home and judge by results. And as it is a reasonable assumption that most correspondence players consult Stockfish for their moves, we can probably draw some conclusions that would apply to correspondence chess, too.
But even in correspondence chess there are practical considerations. An opening setup like the Jobava London which most likely is not the very best conceivable may still look attractive, just as the best setup for black is not quite clear, so that players and engines may have a harder task to counter it.
-
Graham Banks
- Posts: 44620
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
-
jefk
- Posts: 1025
- Joined: Sun Jul 25, 2010 10:07 pm
- Location: the Netherlands
- Full name: Jef Kaan
Re: so what's the best chess opening ?
The move 1.g3 (Benko) is sound but also gives no fundamental advantage as made clear in a book by eg. GM Mikhalshev (1).
Best after Benko is probably 1...e5! Then with 2.c4 you get into an English but after 2...c6 Black can equalize.
After 1.Nf3 (**) in my huge Scid database 1.a4 also has a good scoring percentage but this is apparently
after some games in recent years in the titled tuesdays blitz on chess.com
While there is no opening which can result in a positional advantage in all variations (*), the move
1.e4 is often said to be best opening in practice (at least below expert) , zie eg. youtube (2):
Note that also in the chessmood opening courses, also 1.e4 is recommended.
(*) a fact which is well known nowadays by experts, eg. within Quality Chess, ask e.g. GM Jacob Aagaard.
Usually, the 'best' opening is thought to be that one which leads to the sort of positions you like
to play, but then in practice this means the best opening is the one you know the best (with human
-online- and otb chess i usually played 1.d4 avoiding Sicilian after 1.e4 but recently changed that
and now simply play a Bb5 (or sometimes a gambit) against the Sicilian if it occurs (the rest is secret
(**) Black can also go again for a system with g6, the anti-Gruenfeld, not so easy to play at lower
levels but solild nevertheless, as also described by Mikhalshev (1)
1) https://qualitychess.co.uk/products/1/2 ... khalevski/
(2)
(3) https://chessmood.com/course/rock-n-rolling-with-white
Best after Benko is probably 1...e5! Then with 2.c4 you get into an English but after 2...c6 Black can equalize.
After 1.Nf3 (**) in my huge Scid database 1.a4 also has a good scoring percentage but this is apparently
after some games in recent years in the titled tuesdays blitz on chess.com
While there is no opening which can result in a positional advantage in all variations (*), the move
1.e4 is often said to be best opening in practice (at least below expert) , zie eg. youtube (2):
Note that also in the chessmood opening courses, also 1.e4 is recommended.
(*) a fact which is well known nowadays by experts, eg. within Quality Chess, ask e.g. GM Jacob Aagaard.
Usually, the 'best' opening is thought to be that one which leads to the sort of positions you like
to play, but then in practice this means the best opening is the one you know the best (with human
-online- and otb chess i usually played 1.d4 avoiding Sicilian after 1.e4 but recently changed that
and now simply play a Bb5 (or sometimes a gambit) against the Sicilian if it occurs (the rest is secret
(**) Black can also go again for a system with g6, the anti-Gruenfeld, not so easy to play at lower
levels but solild nevertheless, as also described by Mikhalshev (1)
1) https://qualitychess.co.uk/products/1/2 ... khalevski/
(2)
(3) https://chessmood.com/course/rock-n-rolling-with-white