Hello Jim et al,
Thanks for this. I thought Nene might be White. Nene, I'd love to know when you
spotted 17 Na7!
Stockfish is stronger than Deep Rybka but it all depends on how long you leave
it to analyse. The top level, as Jim hints, may well take several hours before
producing a result. Even then engines can miss things or make wrong assessments.
Just to clarify my name. It is Andrew Tyson Mordue. I never use the Andrew. On
the FIDE Elo list I'm down as A Tyson Mordue, or Mordue, A Tyson depending on
what the filter is. Unfortunately my internet providers are a bit fussy about
my user name so it comes out as the clumsy Mordue Andrew. I have complained but
it's too much bother to sort out. No offence taken Jim.
For the sports fans amongst you: I'd just like to reassure you that the books
by a certain British boxer do not reflect my state of mind! Tyson Fury in case
you were wondering.
Regards,
Tyson
On 19 April 2021 at 21:48 Jim Slagle <dmarc-noreply@xxxxxxxxxxxxx> wrote:
Hi,
Thank you Mordue Andrew!
Congratulations Nene, as white!
Lichess uses Stockfish, which can be strong.
In Lichess, you can use is at levels between 1 and 8.
1 is beginner.
8 is grandmaster.
I assume that it was set at a low level, since black played poorly.
The problem, I believe, is that it analyzed at a low level.
This is practically a bug.
The analysis should be at a much higher level, for example, 7.
Level 8 kmight be too slow.
Is there any way to tell Stockfish to analyze at a higher level?
If not, maybe the Lichess programmers could be asked to do this.
I understand that they are extremely helpful in general.
Best wishes,
Jim Slagle
On Monday, April 19, 2021, 12:58:53 PM EDT, mordue andrew
<dmarc-noreply@xxxxxxxxxxxxx> wrote:
Hello everyone,
Sorry that I've taken a while to get to this one but I've been busy on
many fronts, some of them cold ones coming in from the Atlantic.
This is intriguing. I'm being asked to comment about the computer's
comments. I suppose there's a certain irony in that but here goes.
The engine's comments are actually very shallow and in at least one
instance wrong. The main drawback is that the engine doesn't necessarily tell
you why a move is an error, nor does it necessarily explain why its
suggestion is an improvement. On the other side there is something definitely
missing. This is actually a very good game by White and there is no credit
given at all by the engine to the White player. So we can hardly call the
comments even-handed.
Let's see if we can do better. N.B. I'm not entirely sure who the players
are so I'm going to stick with White and Black. For that reason I'm omitting
all the preliminary details and just going straight to the text (game) moves.
1. d4 d6
The engine's comment "Inaccuracy. e6 was best."
I know precisely what Chris (Ross) would say to this and it is with some
difficulty I refrain from saying the same. Also I have to bear in mind that
Black could be a human who is a member of this group.
This move is definitely not inaccurate. It's move one and there are many
different ways of playing, many different systems, formations, styles of play
etc. Most moves are acceptable here. The bad ones are 1...g5?? blundering a
Pawn, 1...a5 and h5 which are true beginner's moves with the intent to
develop the Rooks on a6 or h6, and probably lose them to a Bishop capture.
Virtually all the other Pawn moves avaiable are recognised systems other than
1...f6, and even all the Knight moves including to a6 and h6 are recognised
lines.
2. Nf3 g6
3. Bd2
My own comment here. This is a very modest move. The London system with 3
Bf4 and the Torre Attack with 3 Bg5 are far more potent. The text together
with White's next mean that the only square for the b1 Knight is a3.
It so happens that later this Bishop proves to be on the right square to
take advantage of an opportunity.
3...Bg7 4. c3 a5?!
The engine's comment: "Inaccuracy. Nf6 was best."
This is probably true. 4...Nf6 is the most obvious move. If Black wants
to play on the Queenside then 4...b5 is better and this is similar to a
Polish Defence which is usually 1 d4 b5.
5. a4!?
This interesting counter may raise some eyebrows yet the engine fails to
comment! The simple point is that if Black can spend a tempo to play on the
Queenside, because the follow-up should be ...a5-a4 at some point, then White
can also spend a tempo there to stop this. Note that Black now has to be
careful about playing ...c7-c5 because it cedes the b5 square to White. In
the game this becomes relevant because later Black has to recapture ...c6xd5
and loses control of b5 anyway.
5...f5
Taking the game into a sort of Dutch Defence, normally 1 d4 f5 or 1 d4 e6
2 c4 f5.
6.Qc2
Maybe this is a little early to define the Queen's position. I prefer 6
g3. See note to White's 7th.
6...Nd7
The engine's comment is "Inaccuracy. Nc6 was best."
Again it's a question of taste as to where to develop the Knights. I'd be
playing 6...Nf6 and 7...0-0 and determining the Queenside development later.
There is a slight problem with the text in that e6 may become a weak square.
We'll come back to this.
7. e3?!
Once again no comment by the engine but this isn't a good move. It makes
the d2 Bishop look like a big Pawn. I much prefer 7 g3, 8 Bg2 and 9 0-0. A
Kingside fiachetto is a traditional way of playing against the Dutch Defence.
7...Ngf6
8. Be2?!
Once again this Bishop development is too modest. 8 Bc4 controlling the
whole a2-g8 diagonal is much better. We've just had Easter and Bishops are
used to making statements around then!
Incidentally Deep Rybka evaluates the position as absolutely even after
8...Ne4. 8 Bd3 instead would have prevented this,
8...c6
No comment on this either but an instructive moment is missed. I looked
at 8...0-0 thinking it was bad after 9 Qb3+ Kh8 10 Ng5 with the threats of 11
Ne6 and 11 Nf7+. However, after 10...Nb6 11 Nf7+ Rxf7 12 Qxf7 d5 Deep Rybka
claims that Black is better!
The point is that the White Queen is trapped with no obvious get out.
Black has the option of ...Qd6 and ...Be6, or ...Qd6, ...Bd7 and ...Rf8 to
snare Her Majesty. Deep Rybka suggests 13 e4 giving up a Pawn - White is
still an Exchange up - followed by 14 Bg5. Well worth looking at.
9. O-O b6?!
The natural moves here are 9...0-0 and 9...e5. Clearly the text is played
with the intention of ...Ba6 but I'd give priority to one of the other two
moves. Ultimately Black pays the price for not castling. The text runs into a
little problem.
10. Ng5! Nf8
Previously Ng5 could have been answered by ...Nb6, but last move Black
blocked that square. Hence he is forced into this rather awkward move.
11. Bf3?
Again no comment but another opportunity missed. Strong is 11 Qb3
threatening 12 Qf7+ winning material. Black must either reply 11...d5 which
weakens e5 and robs his Pawn position of some flexibility, or 11...e6 when
his third rank is looking very soft. White should then play 12 f3 aiming to
open the position up with 13 e4 while Black's King is still in the middle.
11...d5
Now White's Bishop is not a good piece. Also Black threatens to take an
initiative on the KIngside with 12...h6 13 Nh3 g5 and White must retreat the
f3 Bishop to avoid the ...g5-g4 fork and there's litle to do about that.
Probably White's best move here is the immediate 12 c4 to pressure the
centre. The b1 Knight can be developed on c3.
12. Na3 Ba6?!
This looks tempting and simple but 12...h6 pushing the g5 Knight back
first is better, while 12...Ne6 is a clear improvement that gets the poor f8
Knight into play.
13. c4!
Finally getting to grips with the fixed Black centre. Black needs to be
careful here as he has lots of weak squares, he is behind in development and
his king is uncastled.
13...e6?
Once again the engine says "Inaccuracy. h6 was best."
Deep Rybka agrees that 13...h6 is a better move, but it prefers 13...Ne4
because that reduces the pressure on d5. The text ironically creates a target
on e6 which, amongst other things, means the f8 Knight can't move.
14. cxd5!
This is a fine move with an excellent follow-up as well. The engine fails
to praise this move. The point is that 14...Bxf1 15 Qc6+ Ke7 16 d6+! Qxd6 17
Qxd6+ Kxd6 18 Bxa8 regains the Rook and leaves Black threatened with 19 Kxf1
and 19 Nf7+ forking. If Black had forced Ng5-h3 with an earlier ...h7-h6 this
wouldn't have been a big problem.
Black is forced to recapture with the c6 Pawn or lose that for nothing
with check.
14...cxd5
15. Qc6+ Ke7
Interposing with the Queen left the a8 Rook en prise with check, while
putting either Knight in the way simple allowed 16 Nxe6.
This is a good moment to take stock for a couple of reasons, First of
all, material is still level although the positional factors are that Black
is behind in development, he has lost his castling rights and his Pawn
structure is full of holes. However, Black has two threats here: the obvious
16...Bxf1 and the not so obvious 16...Rc8 trapping the White Queen.
Before we proceed to some analysis I just want to take this opportunity
to make a point about the rules of the game. Suppose White chose to retreat
with say 16 Qc2 and Black in reply went 16...Ke8. This is apparently the same
position after Black's 14th move. Now assume that Play proceeded 17 Qc6+ Ke7
18 Qc2 Ke8. This is now the same position for the third time. Question: Can
Black claim a draw by repetition before playing 18...Ke8? (N.B. He loses the
right to claim if he actually makes the move.) In his turn can White claim a
draw by repeition by declaring his intention to play 19 Qc6+?
The answer is in fact no! The reason why is that while all the static
factors all the same (everything in the same places) a dynamic factor has
changed. At move 15 Black could still legally castle in subsequent play, but
at moves 17 and 19 he can't. This would invalidate a draw claim. No less a
player than Anatoly Karpov was once caught out by this.
Back to real life. As I wrote earlier I have no idea who the players are
here. But it looks like White saw a lot further into this position than Black
did. Certainly as late as prior to playing his 15th move he would need to
have seen his smart 17th and 18th moves.
16. Nb5
Interposing against the Bishop's attack on the f1 Rook and introducing
another piece into the attack. If now 16...Qe8?? 17 Qd6 is mate. Note that
the Knight on g5 controls f7. Now what would you play for Black?
16...Rc8?
The engine note says "Mistake. [16...]Bxb5 was best."
This time the engine is absolutely right. Also playable is 16...h6 17 Nh3
and then 17...Bxb5. The point is that the b5 Knight is immediately
eliminated. However, it is so easy to miss White's excellent reply and the
horses start to run riot.
17. Na7!!
A fabulous move. White leaves his Queen en prise and his f1 Rook is again
under fire, but Black can't take either. If 17...Bxf1 then 18 Nxc8+ not only
levels the material but Black's only legal reply is to surrender his Queen.
Alternatively if 17...Rxc6 18 Nxc6+ Ke8 or d7 19 Nxd8 Bxf1 20 K or Rxf1 Kxd8
21 Nf7+ and 22 Nxh8 leaves White an Exchange up with an easily winning ending
after b2-b4 opening more lines for the Rook.
Now the Black Rook on c8 is en prise to the Knight. There are three moves
possible other than taking the White Queen:
a) 17... Ra8 18. Bxa5! Rxa7 (if 18...bxa5 19 Qxa6 and 20 Nc6+ or 20 Qb7+
are coming) 19. Bxb6 Qa8 20. Qc5+ winning the Rook.
b) 17...Rb8 18 Qc3! (18 Bxa5 is also good. The text threatens a triple
fork on c6 hence ...) 18...Bb7 19 Qa3+ Ke8 (if 19 ...Kd7 20 Nf7 forking) 20
Nb5! Qe7 21 Nd6 Kd7 22 Ngf7 Rg8 23 Ne5 Kd8 24 Rdc1 and now there is no
defence to 25 Ne2 and 26 Nef7+ Kd7 27 Bb5+. Analysis by Deep Rybka. I'd have
settled for 18 Bxa5 as indicated.
Finally there's the game move but that drops the b6 Pawn immediately.
17...Rc7
18. Qxb6 Bxf1
Saving the Bishop and nabbing a Rook but White's next is going to
guarantee him being at least two Pawns up.
19. Bxa5
Ironically proving that d2 was the right square for the Bishop!
19...Kd7
20. Nf7!
A crushing move. Despite being temporarily a Rook up Black has no reply
to this invasion. His f1 Bishop, both Rooks and his Queen are all en prise.
The priority is saving the Queen but there is no square that it can safely go
to that continues to defend the c7 Rook. This means major material loss with
the King still completely exposed. Understandably Black resigns.
1-0.
An excellent attack by White, whoever it was. Poor analysis by the engine.
Regards,
Tyson
> > On 13 April 2021 at 22:25 Nene <dmarc-noreply@xxxxxxxxxxxxx>
wrote:
Hello, I want to share my game that I analysed using the lichess.
It's interesting to hear Tyson's comment on this analysis.
I played white and black resigns on this game.
Here it goes:
Moves
1 delta 4, delta 6Inaccuracy. e6 was best.,
2 knight foxtrot 3, golf 6,
3 bishop delta 2, bishop golf 7,
4 charlie 3, alpha 5,"Inaccuracy. Nf6 was best.
5 alpha 4, foxtrot 5,
6 queen charlie 2, knight delta 7Inaccuracy. Nc6 was best.,
7 echo 3, knight golf foxtrot 6,
8 bishop echo 2, charlie 6,
9 short castling, bravo 6,
10 knight golf 5, knight foxtrot 8,
11 bishop foxtrot 3, delta 5,
12 knight alpha 3, bishop alpha 6,
13 charlie 4, echo 6Inaccuracy. h6 was best.,
14 charlie takes delta 5, charlie takes delta 5,
15 queen charlie 6 check, king echo 7,
16 knight bravo 5, rook charlie 8Mistake. Bxb5 was best.,
17 knight alpha 7, rook charlie 7Blunder. h6 was best.,
18 queen takes bravo 6, bishop takes foxtrot 1,
19 bishop takes alpha 5, king delta 7,
20 knight foxtrot 7
Regards,
Nene
>
