"So according to Rybka, the best move is 1...Nd4 (-0.27), the one we wrote off as strategically bad."

 

I know how to interpret Rybka. Rybka is saying Nd7 and Qd7 result in a better position than d5 or Nd4, assuming best play by both sides.

 

"Rybka has really told us little decisive in terms of the BEST move, and even less about the BEST plan."

 

It's provided a wealth of information if you read carefully. It quickly confirmed our analysis of Nd4, it found a serious hole in our analysis of d5 and Qd7 followed by Bxh3, and it opened our eyes to a new candidate--Nd7.

"Basically, I'm saying computers aren't the be-all-end-all of chess analysis."

 

Nothing is... you... me... Paolo... Rybka. But Rybka at 3050 ELO outrates all of us by at least 800 ELO. It was written by an IM and it's better positionally than most chess engines. When it picks d5 after a second, Qd7 after 10 seconds, and Nd7 from 20 minutes to 24 hours, its best move deserves serious consideration and absolutely should have been one of our candidate moves. I think the important things to focus on are:

 

1. Did you see d5 as a candidate? Did you see 1...d5 2.exd5 Nxd5 3.Re1? Did you see how White would gain an advantage from Black's replies?

 

2. Did you see Qd7 as a candidate? Did you see that White doesn't need to defend against Bxh3 at all, and even playing h6 doesn't improve matters?

 

3. Did you see Nd7,Nc5 as a candidate? If not, why not since Black had no clear winning plan and these moves improve Black's minor pieces?

Well firstly, I wouldn't trust a computer in a puzzle where the problem is to find the best plan.

 

Secondly, the difference between Rybka's analysis of those moves is 0.09, less than a tenth of a pawn, making any move playable. Rybka has really told us little decisive in terms of the BEST move, and even less about the BEST plan.

 

And unless Rybka works differently then other computers, then the lower the value, then the better the move is for Black, thats us. So according to Rybka, the best move is 1...Nd4 (-0.27), the one we wrote off as strategically bad.

 

Basically, I'm saying computers aren't the be-all-end-all of chess analysis. But I would instead use them for giving value to the positions and the end of our analysis. Especially those variations which are very tactical, perhaps the lines containing ...f4 at some point.

 

As the original poster has not responded, I asked the Silicon Monster to teach us something about this position. It thought for a long time (about a day) then said:

 

Rybka, depth 22: d5 is -0.25, Qd7 is -0.19, and Nd7 is -0.18.

  

1...Nd7 is better than all our moves. Unfortunately, it didn't explain why.

 

1...Nd7 2.Be3 Nc5 3.Bd5 Bf6 4.a4 Qd7 5.Re1 Rad8 6.Bc4 Bxc4 7.dxc4 Nb4 8.Bxc5

1...Qd7 2.Bd2 Bxb3 3.axb3 Rfe8 4.Re1 a5 5.Be3 Nb4 6.d4 exd4 7.Bxd4 b6 8.Na4

1...d5 2.exd5 Nxd5 3.Re1 Bc5 4.Qd2 Bd4 5.Nxd4 exd4 6.Nxd5 Bxd5 7.c4 Be6 8.Qf4

 

Understanding

 

1...Nd4 (-0.27) and Black is worse after 2.Nxd4 exd4 Ne2 as explained in earlier posts.

 

1...d5's (-0.25) problem is easy to understand. After 1...d5 2.exd5 Nxd5 3.Re1 the e5-pawn is attacked. How would you defend?

3...f6? is met by 4.d4! The e5-pawn has three attackers and only two defenders, the pawn is fixed (exd5 and e4 are disastrous), and White can apply more pressure to the pawn than Black can.

3...Bf6 4.Ne4 Be7 and White's position looks better.

3...Qd6 4.Nb5 Qd7 and White can capture the e5-pawn.

3...Bf6 4.Ne4 Be7 Ba4 and Black doesn't lose material, but White's position certainly looks stronger than in the original position.

3...Bc5! is surprisingly Black's best continuation... would you find it? If White tries to win the e5-pawn, Black can sacrifice his bishop on f2 and counterattack.

 

1...Qd7 (-0.19) (or 1...h6 2...Qd7) makes a concrete threat that puts White on edge. We identified 2.Ng5 as a defense, but it turns out that's not even necessary. Even if 2.<pass> Bxh3 3.gxh3 Qxh3 4.Ng5! ends the attack. And even if Black gets in h6 for free, the fortress Kh1/Nh2 can't be broken down with so few Black pieces in the vicinity. Qd7 is not a bad move, but even against a class B opponent it's not a great move.

 

1...Nd7 (-0.18) - Perhaps it's quite simple. After calculcating that neither of our forcing moves worked well, and checking White's forcing moves, we should have turned to simply improving our pieces. After Nd7, Nc5 both our dark-squared bishop and our knight are more powerful and harmonious than they were in the original position. 

 

It was six men of Indostan,
To learning much inclined,
Who went to see the Elephant
(Though all of them were blind),
That each by observation
Might satisfy his mind.

The First approach'd the Elephant,
And happening to fall
Against his broad and sturdy side,
At once began to bawl:
"God bless me! but the Elephant
Is very like a wall!"

The Second, feeling of the tusk,
Cried, -"Ho! what have we here
So very round and smooth and sharp?
To me 'tis mighty clear,
This wonder of an Elephant
Is very like a spear!"

The Third approach'd the animal,
And happening to take
The squirming trunk within his hands,
Thus boldly up and spake:
"I see," -quoth he- "the Elephant
Is very like a snake!"

thanks to likeforests for his explanation about bishops' goodness and badness; this way it looks a very democratic attribute since depending from whereto the majority of pawns will decide to settle from time to time, leaving then room to possible alternances as soon as such majority should change. 

Coming to the serious, 1...Qd7 2. Ng5 Nd4 looks fine but I have to agree with akuni that 3.f4 balances it and furthermore I think there is a psycho-strategical side to be considered too. Though we do not know the sequence of moves that led to the situation proposed, we can guess it: we are at the completion of the 8th move and all pieces but WlsB are in natural development squares;  it got there very likely from c4 and the only apparent reason for it is that black had played Be6.W preferred in all evidence to lose a tempo than trade a B, although presently "bad". Why ? Certainly not to trade it in the following 2 moves but possibly to keep fightinf for d5 control. If it is so, the only W next moves  that can prove quickly useful are Bg5 or Ng5. Both prevented by 8....h6 for which I confirm then the preference shown in my yesterday message, reserving to play ...Qd7 next move.

If there are no further contribution to this very interesting debate, may be it is hight time for the proposer of the problem to unveil his own solution.

Thanks and regards to all of you.

    

 

I like Nd4

After 1...Qd7 2. Ng5 Nd4 3. f4 here are some lines I found

3...Nxb3 4. axb3 (f5 threatened) exf4 5. Bxf4 h6 (this again, but the Knight is dangerous. 5...Rfe8 for example 6. d4 (7. d5 threatened) Qd1 7. Nxe6 fxe6 8. e5 Nd5 (Nd7 eventually loses a pawn) 9. Nxd5 exd5 and White is better) 6. Nxe6 Qxe6 (6...fxe6 lands us an isolated pawn or isolated pair after 7. e5)  7. Nb5 Qd7 and White can try to win the pawn, but even if he spurns it, White is a bit better off.

So not Nxb3, what about 3... exf4 4. Nxe6 fxe6 5. Bxf4 Nxb3 (to unpin the e-pawn) 6. axb3, and what now, I can't find any good moves for Black.

1...Qd7 2. Ng5 Nd4 3. f4 exf4 4. Nxe6 fxe6 5. Bxf4 Nxb3 6. axb3 looks like the line with best play by both players (starting with 1...Qd7 that is), but it fails to capitalize on Black's current lead in developement. I'm definently considering 1...h6  because that Knight is very dangerous in most of the variations, but its placement on f3 stops the f4 thrust.

Akuni, 2.Ng5 is a line I didn't consider... excellent find! But I think Black can do much better than 2...h6. After all, the knight threatens nothing but Nxe6 (since Qg4 and Qh5 are impossible). Why spend a tempo forcing the move?

 

How about 1...Qd7 2.Ng5 Nd4! It seems like a natural central outpost for the knight, and it's the square left unprotected by Ng5.

 

What can White do now?

 

3.Ba4, attempting to keep the bishop pair is met strongly by 3...b5 and White must return to b3 again and Black gains a tempo. 

 

3.Nxe6 is met by fxe6. Black gets his half-open file and extra central pawn, and he can reclaim the bishop pair if White spends a tempo forcing it.

 

3.Nd5 Bxd5 4.Bxd5 Nxd5 5.exd5 Bxg5 6.Bxg5 f6 and while we didn't get a half-open file, White's weak pawn give him winning chances.

 

3.Be3 Nxb3 4.axb3 d5 - This is better now than before because d5 has one less defender... we will control the center. If 5.Nxe6 we get the situation we were wanting all along, and if White proceeds to make massive exchanges his half-open a-file becomes an endgame weakness rather than a middlegame strength.

"Finally, is there any doctrinal definition for "good" and "bad" bishop?"

 

Yes. Usually your pawns are mostly arranged on either light or dark squares. Your bad bishop is on the same colour as your pawns. Your good bishop is on the opposite colour. This definition makes sense from the perspective of mobility. Of course, do not put too much stock in labels or generalities. In some positions your bad bishop might actually be a very strong piece, so use your eyes to check if that's the case.

 

For example, I wrote above:

 

"White's bishops are better than Black's bishops. That's because Black's bad bishop is locked inside his pawn chain while White's bad bishop is free "

 

And here's a position where, as Black, I used my "bad" bishop to obtain a strong bind on the light-coloured squares and dominated my opponent. Notice how none of my pawns hindered its mobility along the b1-h7 diagonal:

 

 

first of all I have to thank paolodm for submitting a very teaching challenge as it is proved moreover by the interesting likeforests+akuni analysis.

As to me I would not venture now as black in ...Qd7 since I would instead be eager as white to play Ng5.

To the point of wondering whether the best move for black would not be the plain ..h6 to keep away any visitor from g5 and then possibly to consider  ... Qd7 in a later time.

Finally, is there any doctrinal  definition for "good" and "bad" bishop ?

Thanks for replying

I think in these situations  any thing you do can be a good mouve if it helped to open or make a check

For me I'll mouve the pawn in d4 a square to open the game  . . .

Looks like i dont need to explain a thing thanks to our good friend likeforests i couldnt explain it any better..... 

I prefer 1...d5 because as Likeforests says there areonly 3 good looking moves (Nd4, Qd7 and d5) and 1...Nd4 can be excluded quickly due to the resultant pawn weakness. But there is an interesting variation of the 1...Qd7 line that should be looked into.

1...Qd7 2. Ng5 (instead of 2. Bxe6 and guarding h3) h6 (2...Bxb6 3. axb6 looks bad) 3. Nxe6 fxe6 4. f4  and from here there are a couple possible moves I see.

1) 4...exf4 5. Bxf4  and here the tables have turned, now White can threaten sacrifices against h6 and our Bishop is poor and White has two Bishops, 5...d5 to alleviate some of these problems looks bad after 6. e5 Nh7 7. Qg4! with dual threats of Bxh6 and Bxd5 (g and e pawns are both pinned). It's +- I think.

2) 4...d5 5. f5 (threatening 6. fxe6  (Qxe6?? 7. exd5 which wins) Qe6 7. dxe5 with two extra pawns) and Black is in very bad shape, 5... exf5?? 6. exd5 Nb4 7. d6+ Kh8 8. dxe7. Even worse is 5...dxe4?? 6. Bxe6+ Qxe6 7. fxe6. One possible try for Black is 5...Ne8 but the defense is still insufficient and Black cannot recapture the pawn. So 4...d5 clearly fails.

So I have to say 1...Qd7 doesn't look good in light of this variation. I favor Ilikeforests analysis of 1...d5 for the best move.

Material is equal. We are ahead in development.

 

The pawn structures are symmetric except for White's h-pawn, which prevents Bg4 but weakens the kingside maybe allowing Bxh3 at some point. An f-pawn advance is impossible now but a d-pawn advance is possible.

 

White's bishops are better than Black's bishops. That's because Black's bad bishop is locked inside his pawn chain while White's bad bishop is free.

 

The knights, kings, rooks, and queens are obviously equal. 

 

Candidates:

1. d5 (threatens d4, frees out bad bishop, maybe keeps our good bishop on board, breaks symmetry when we're better developed) 
2. Qd7- developing our queen and threatening Bxh3 gxh3 Qxh3.
3. Nd4 - Seizing a good outpost, but I don't have high hopes since White can duplicate or immediately capture.

Quickly Rejected Candidates:

1. Captures: Nxe4, Bxh3 - These seem to lead nowhere.
   
2. Bxb3 axb3 - We trade off our good bishop for his "bad" bishop, we create doubled pawns but open his a-file (double-edged). This is rather mediocre, especially since Bxe6 would be good for us (we get a half-open f-file to attack on plus an extra central pawn).

 

1...Qd7 2....Bxh3 3.gxh3 Qxh3 4.Bg5 Ng4 5.Bxe7 Nd4 6.Bxf8 Nxf3 with mate to follow. There's a chance I miscalculcated, but 1...Qd7 could lead to a scary and likely winning attack and so probably White would prevent it with Bxe6 fxe6 and we have a desirable position: more pawns in the center and a half-open f-file!!

 

1...d5 2.exd5 Nxd5 3.Nxd5 Bxd5 4.Bxd5 Qxd5. After f6 we have a better pawn structure and our queen is central, so Black is better, but massive simplifications usually benefit the defender more than the attacker. There's no need to pursue these lines further since 1...Qd7 seems better than this line.

 

1...Nd4 2.Nxd4 exd4 3.Ne2 - This d4-pawn though advanced and central is also weak. I don't like this too much. If now 3...Qd7 the potential attack is even stronger but after Bxe6 fxe6 we have the same advantages as in the 1...Qd7 line but with the additional disadvantage of the weakness on d4. So 1...Qd7 is stronger.

 

1...Qd7 2.Bxe6 fxe6 (else 2....Bxh3 3.gxh3 Qxh3 4.Bg5 Ng4 5.Bxe7 Nd4 6.Bxf8 Nxf3 with mate to follow.) and Black has more central pawns and a half-open f-file. The position now has an imbalance that's favorable to Black and that I know how to play.

Hey littleman,

Thanks for your response. Ok let's say you play d5... can you see what white may play next?

Paolo 

 

i like d5 myself break open the position