Expert: Len Molden - 3/3/2010

Question
QUESTION: This one is hard because it is specific and I know you are only an expert.  Just wondering what your opinion is.

It seems that a new way for me to learn openings is to use a database, so I bought chessbase and have that 4.5 million super every game ever played by a master in an international tournment game edge.  ????  I wish, I can use this data but I can't make sense of it.

It seems GM's and the world's top players play lines until someone loses and then it's done.  Here's an example, a famous game I think it was from one of the world's top players from 1855, John Cochrine was the player and the variation it seems or move has never been tried again.  Yet I'm convinced the move is good and Fritz thinks it wins.



Your question was "[Event "Calcutta"] [Site "Calcutta"] [Date "1855.05.18"] [Round "?"] [White "Cochrane, John"] [Black "Somacarana"] [Result "1-0"] [ECO "C37"] [PlyCount "49"] [EventDate "1855.??.??"] [EventType "game"] [EventRounds "1"] [EventCountry "IND"] [Source "ChessBase"] [SourceDate "2006.11.23"]   1. e4 e5 2. f4 exf4 3. Nf3 g5 4. Bc4 g4  (Only 1 GM in the world has ever played this since the GM title existed, his name is GOdena, and he is thought to be weak, compared to other GM's.   5. O-O gxf3 6. Qxf3 Bh6    So here is the move, what a bad move right?  No top player would ever play this again right?  It's never been played since 1855.  


  1. e4 e5 2. f4 exf4 3. Nf3 g5 4. Bc4 g4 (NOTE: NO GRANDMASTER HAS EVER PLAYED THIS IN THE HISTORY OF FIDE TOURNMENTS except a player called Godena, Italy's best player.  Yet no one else dare's to play this at the top level.  And he is considered weak I suppose by his competition.  THIS MOVE CLEARLY DOESN'T LOSE FOR BLACK) 5. O-O gxf3 6. Qxf3 Bh6   I wanted to stop here, this looks like a great move, and it isn't in modern chess theory. I mean after this game was played in 1886 no one ever considered it again. But this blows me away isn't black fine here?  I think the point of bh6 is to simply to defend the pawn on f4, if d4 then black can simply play Qh4 covering f4.  If the pawn on f4 stays white's attack is nullified.    The game continued, black got spanked because of poor play and no master has ever dared play bh6 again.  Yet Fritz thinks it wins for black, and I too am convinced this is winning. Stange eh?   More to the point, no top player in FIDE history has played 4.... g4 as black.  Does that mean that white players of the Kings gambit shouldn't prepare for it? Does it actually lose.  I don't belive it yet it isn't played.  I guess that's the end of it.  Do you find things like this in chess often?  I thought a database would help me, but all it has done is confuse me.  I just don't get opening theory.  

  7. d4 Qf6 8. e5 Qb6 9. c3 Nc6 10. Bxf4 Nxd4 11. Bxf7+ Kxf7 12. Qh5+ Ke7 13. Bg5+ Ke6 14. Qf7+ Kxe5 15. Qf4+ Kd5 16. c4+ Kxc4 17. Na3+ Kb4 18. Nc2+ Ka5 19. Qd2+ Ka6 20. Nb4+ Kb5 21. a4+ Kc5 22. Rac1+ Kd6 23. Qf4+ Ke6 24. Qf7+ Ke5 25. Qd5# 1-0     

Okay so that's the end of the game, white wins but it wasn't because of 6... bh6 was it?  This move has never been played again by a master according to chessbase.  If you started to dig into modern chess theory it would seem that 4.. g4 as black in the king's gambit is a bad move.  


I don't believe it though and fritz doesn't get it, as it thinks black is winning.  Will this ever be played again by a GM  ...6 bh6?

Isn't it a good move in your opionion.  Yes black lost that game but how could you pin it on a move at move 6?

Sorry for the tough question.  Don't you find opening theory crazy?  This doesn't seem logical at all.


John van Leeuwen

ANSWER: Hi John,

This was indeed a very difficult question to answer and as I didn't feel that I could, I enlisted the help of a cast of much stronger players than myself.  I received the help of two Canadian masters, one FIDE FM, and one famous FIDE Grandmaster.  I hope this helps,

Regards,

Len


Here's what they had to say:


GM Yasser Seirawan
==================

4.Bc4 is not good for White because of 4...g4!

I "thought" that White is supposed to play 4.h4! and then after 4...g4 5.Ne5
d6 6.Nxg4 Nf6, is the main line.  Shirov play's the Black side of this for example.

It is because of this 4.h4 move that some GM's, trying to transpose into "Fischer's refutation" play:

1.e4 e5 2.f4 exf4 3.Nf3 h6 (!?)

Now if 4.d4 g5 5.Bc4 d6, the game has transposed into Fischer's refutation.

If White tries 4.Bc4 d6! (to stop Nf3-e5) 5.O-O Bg7, Black is supposed to be happy.


FM Bruce Harper
===============

I don't play the King's Gambit for either side, but I don't think White is forced to sacrifice the piece on f3, is he?  I think this may be a slightly weaker version of the Muzio, but White certainly has some compensation for the piece (his position is easier to play against a human).


Dan Scoones (Canadian Master)
=============================

Opening theory is found in opening books, not in raw databases.  For novices, I used to recommend Fine's book but I believe it's been overtaken by the Watson series.
And no, 6...Bh6 does not refute the Muzio Gambit.  The bots are good in the short range but except for Rybka they are all confused by positional piece sacrifices!


Anonymous Canadian Master
=========================
There are three mistaken assumptions in the question asked by John:

1) Fritz is a reliable measurement of all chess positions.
2) Positions that are not played are not considered.
3) Grandmasters choose opening moves primarily based on "best" play.

Computers by themselves have trouble analyzing positions which are exceptions to the basic principles of chess.  These kind of positions contradict a computer's assessment logic and tactical resolution is often beyond the computer's horizon.  To use a computer to analyze such a position you need to guide it further into the position.  However the effectiveness and efficiency of that guidance is dependent on the knowledge and understanding of the player.

There are an endless number of ideas that have been considered and even talked about that never get played over the board.  It is not always necessary for someone to experience a misfortune to persuade people to stay away from it.  Bh6 and Qh4 has probably been considered.  Maybe try Bilguer's Handbuch.  I do not think that this idea is advisable for multiple reasons.

Grandmasters do not play 4.Bc4 very often.  4.h4 and 4.d4 are considered preferable.  The reason is that the consensus since the early part of the twentieth century has been that the position after 4...Bg7 5.d4 d6 is highly favorable to black.  This is the basic assumption that Fischer built upon as the inspiration to the "Fischer Defense" introduced in his famous article from the 1960s declaring the Kings Gambit "busted."  This was in reaction to losing to his rival Spassky in the Kieseritzky - 4.h4.

Considering that most Grandmasters consider 4.Bc4 Bg7 to be extremely favorable for black and that 4.Bc4 g4 is extremely complex and extremely risky, it is not practical to devote time and effort to 4...g4.  4.Bc4 is very infrequently seen.  This makes 4...g4 a rabbit hole - a lot of digging, leading to no place.  The rewards simply do not justify the effort.  That is why Grandmasters pay little attention to it.  This is similar to why few Grandmasters wish to spend time on the sacrifice 1.e4
Nf6 2.e5 Nd5 3.d4 d6 4.Nf3 dxe5 5.Nxe5 Nd7 6.Nxf7!  Even Tal made the practical decision to decline to play this with 6.Nf3 when confronted with it.

---------- FOLLOW-UP ----------

QUESTION: The response by the anonymous Canadian Master directly clashes with what I think GM Yasser Seirawan was stating.

Although I find what the anonymous Canadian Master is saying interesting, certain moves are too complicated so if you can just avoid them, you should.  He calls them "rabbit holes", however Gm Seirawan stated that 4 Bc4 is not supposed to be good for white because of 4..g4!  

"Considering that most Grandmasters consider 4.Bc4 Bg7 to be extremely favorable for black and that 4.Bc4 g4 is extremely complex and extremely risky, it is not practical to devote time and effort to 4...g4.  4.Bc4 is very infrequently seen.  This makes 4...g4 a rabbit hole - a lot of digging, leading to no place." -- Anonymous Canadian Master

How can you avoid rabbit holes?, isn't that what opening theory is about?
To try to identify a move that is supposed to be bad and understand why and how to hopefully get some type of advantage?  Here black should enter the rabbit hole because then black is supposed to be happy.  (According to Seirawan)  From the small amount of work that I've put into this, I think I agree with him.

This is actually the reason why I too have always avoided the king's gambit.  I thought I would have to memorize and understand a lot of lines and exact moves for tactical reasons that aren't obvious at the board. But of course I never really looked into them. (Starting too) But this is what it's all about isn't it?

Should i spend little to no time trying to memorize refutations to lines that seldom come up?  Is that wize?  What if your opponent plays a move out of book? Or one of these lines that never often come up?  Navigating these rabbit holes are really hard then!

Wouldn't it make more sense trying to learn why certain moves are considered to be bad or slightly inferior then memorizing book-like good ones?

At a certain level this stuff just can't be avoided? I wonder what it's like to be a strong master or Grandmaster.  The work and preparation..... (Sorry off topic)

And yes you made some good points, even Tal made practical decisions to avoid some lines, okay I'll take your word for it, it makes sense.  But a lot of things can't be easily avoided.  Especially if they can play a move on move 4 or 6 such as in this example.

So opening theory, isn't that all about these complications and rabbit holes?  Yes you can't prepare for everything but you would obviously try to learn some of this stuff if you were going to play the King's gambit right?

Thank you everyone again.  I'm just curious and I'm learning a little,
John

Answer
Hi John,

Here's his answer to your follow-up question:

"For most people the primary motivators in playing chess are fun and competition.  Fun is a matter of personal perspective.  Competition is about trying to win.  From a competitive perspective, practical application is far more important than playing for "truth."  

A player can play an opening that is suspect and have good results and a player can play an opening that is "correct" and have bad results.

Chess is an extremely complex game where "truth" is often elusive and outcomes are infrequently determined by playing one best move.  Top players more often try to choose openings that they think will be effective competitively rather than choosing openings that they think are objectively best.  Emanuel Lasker was known for his ability to steer opponents into positions they were uncomfortable with.  Mikhail Tal was known for his ability to see through the fog in complex and unbalanced positions and so was frequently willing to play speculative moves to steer his opponents there.  

Gary Kasparov known for his preparation would sometimes use openings that he surely knew were suspect, like:

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 e6 5.Ndb5 d6 6.c4 Nf6 7.Nc3 a6 8.Na3 d5?! which he used against Karpov in one of their matches.  In all of these examples, opening choices are being made based on the premise that the result will be determined by the choices a player makes on the spot, over the board and under pressure regardless of the "truth" of the position.  In opening play there is more gray than black and white and the fact that Yasser and I disagree about the assessment of the Muzio Gambit is evidence of this.  If there is an absence of clear answers, from a competitive standpoint, controlling the direction of play offers greater reward than trying to play a "best" move.

Avoiding "rabbit holes" is a matter of making practical choices.  A very common method of "practical" opening application is to use move order to restrict your opponent's options while steering the position to where you would like to go.  Transpositions of moves allow short cuts.  You can reach the same or similar positions from different move orders and this offers the possibility of limiting the number positions and different styles of positions that can be reached.  

Here is example in the Kings Gambit from one of Yasser's games to illustrate what I am talking about:

Spassky-Seirawan, Montpellier 1985
1.e4 e5
2.f4 exf4
3.Nf3 Ne7!?
(Avoiding the complications of 3...g5!  :)  An innovative choice by Yasser.  At the time 3..Ne7 was almost unheard of.  Alternatively 3...d5 was and is a popular option which can result in similar pawn structures and piece placement.  3...d5 was popularized by Soviet players in the post war era and became widely imitated as a result of such famous games as Bronstein-Botvinik and Kronstein-MacDonald.)
4.d4
(Significant is that 4.Nc3 d5 5.exd5 Nxd5 is a transposition to 3...d5
4.exd5 Nf6 5.Nc3 Nxd5 and 4.Bc4 d5 5.exd5 Nxd5 is a transposition to
3...d5 4.exd5 Nf6 5.Bc4 Nxd5.)
4...d5
(This is the point.  After 3...d5 4.exd5 Nf6 the move considered best at the time which had been played by Spassky in previous games was 5.Bb5.
Because of the move order white cannot achieve this because black will immediately recapture with the knight on d5.)
5.Nc3
(Avoiding transposition.  5.exd5 Nxd5 6.Bc4 is the same as 3....d5
4.exd5 Nf6 5.Bc4 Nxd5 6.d4)
5...dxe4
6.Nxe4 Ng6
7.h4 Qe7
8.Kf2 and Yasser had some difficulties.

Although the opening innovation did not turn out well for Yasser in this game, in recent years 3...Ne7 has become a popular choice.  The improvement is 6...Nd5 where 7.c4 can be met by 7...Bb4.  After success with 6...Nd5 white players including Morozevich and Short attempted to thwart this plan with 4.Qe2 indirectly pinning the knight to the king.
However this attempt has to date been unsuccessful since it was discovered that black can play in gambit style with 4...Ng6 meeting 5.exd5 with 5...Be7 and 6...0-0 leaving the white queen and king awkwardly placed in the centre - See Short-Bluvshtein.  As a result, for now, white players have been left with little choice but to allow the transpositions back to the 3...d5 line, thereby losing the option of Bb5.  In summary, this move order, thanks to Yasser's pioneering efforts allows black to limit white's options.  While reaching the same positions 3...Ne7 would require less study than 3...d5.

When studying chess, before you spend time and effort analyzing a position, you should first think about why you are in that position to begin with and consider carefully whether you want to be there."

Regards,

Len