Decision making in chess

ArtNJ

The really interesting thing about this position is how many people of what rating guessed that the central untouchable knight more than made up for the material deficit?  Although maybe it doesn't, unless your a cool as a cucumber engine willing to play a5.  

Farm_Hand

?

To me a5 is an extremely natural and human move. Not engine-like at all. First of all it pretty much wins a pawn by force, but also this sort of formation happens in all sorts of openings (only key players shown below) and a5 is a key move for black

(b5 pawn is not necessary, only that the white pawn on b4 can't escape the pawn break by advancing)

 

The argument against a5 is that you're opening more files for the rooks (which white has 2 of and you only have 1 of).

But since there's no play anywhere else for black, this argument also asserts you have a fortress which is always really risky to go for in the middlegame. If in doubt, I'd go with a5.

 

Yes the centralized knight is worth pointing out. It's the star minor piece for sure. Also worth pointing out IMO is how ineffective white's bishop is, just hitting air.

TwoMove

Like ArtNJ I find it funny how many people are certain of the evaluation of the position, advantage for black, based on a number from software,  In fact it doesn't seem a very stable position, and with most reasonable moves, the software evaluation changes when walk down the line becoming roughly equal, Black can prevent be7-g5 with h4, and Bf4 followed by g4 is a resource in many lines. In human chess, a slight tactical inaccuracy from black can lead to a big white advantage, so in a least practical terms white has the advantage.

When first look at the postion not very attracted to Bd6 exchanging good bishop, but not very sure what I had played.

Lyudmil_Tsvetkov

Interestingly how wrongly people are assessing this position and what unoptimal moves they are choosing, even SF.

Black has SIGNIFICANT advantage here, maybe it's not a win, maybe it is, no time to check further.

The correct plan is to pawn-storm on the king side by ...g6-g5 and ...f5-f4, simultaneously relying on some tactical resources ( the weak d4-pawn).

This will build a powerful pawn roll on the king side, very difficult to neutralise.

Bringing up the rook to the king side is to be considered to strengthen the attack via either f8 or g8.

All black should watch for is white rook penetration on c8 after possible doubling.

The white pawns on the queen side are more or less stuck.

So yes, ...g6-g5 is by far the best move here.

Daniel1115

a5 is your only chance, anything else just let's white improve while you have to sit. I disagree with the computer evaluation after a5 from a practical point of view. Whites rooks get very active and will have some threats/pressure vs The black king. Black will have to tread carefully. Bd6 is pretty bad, you are exchanging one of whites pieces that at best restricts your position, while also strengthening the exchange he is up. Without a5 you don't have enough to attack while white can easily improve. It's one thing to keep the position closed when your opponent cant do anything, it's another thing when sitting leaves you with no play while your opponent strengthens his position.

Farm_Hand

I wouldn't expect to win as black (IIRC the engine was saying it's equal-ish) but I'd be too afraid to try and fortress. I'd rather stir something up on the queenside so that white's kingside play and c file don't start to generate something dangerous.

Lyudmil_Tsvetkov
Farm_Hand wrote:

I wouldn't expect to win as black (IIRC the engine was saying it's equal-ish) but I'd be too afraid to try and fortress. I'd rather stir something up on the queenside so that white's kingside play and c file don't start to generate something dangerous.

Give it more time.

Play ...g5, or ...Be7 first and then ...g5, let SF run for an hour, and then draw your conclusions again.

Even top engines are not very good at more subtle/complex positions.

I have been investigating similar setups a lot, and I can vouch Black is either winning, or having large advantage above (no time to check now).

But you should play ...g5 and then ...f4

A major principle is you should attack where you are strong.

Where is Black stronger - on the Q- or K-side?

Of course, on the king side.

It's 5 vs 4 pawns there, and those black pawns on the king side are all Quality pawns, connected and advanced.

Advance them a bit more, and white will start cracking.

Btw., ...g5 also prepares a nice safe haven square for the black king on g6, where he is perfectly fine.

Later, black can continue with ...h5-h4 depending on circumstances.

White will probably push h2-h4 as best preventive try, but black is fully ok after ...h6

You need 30-40 moves to be fully certain how this plays out, seems to be a borderline position between win and draw.

You know how many times SF evals have started at 0.0 and ended at +500cps!

Or vice-versa.

So, careful with engines.