King and queen vs king, bishop and knight fortress

If white achieves this position it's a drawn as the bishop and knight are creating a barrier around the king (a3, b3, c3, c2, c1) but is there any simple pattern to achieve this position and defending it as white? It looks kinda hard

If you play Qh1+ Kb2 Qd5 followed by Qc4 and the King's march? Zugzwanging the white pieces? I believe Knight+Bishop vs Queen is theorically winning for the strong side without any fortress. They are some fortresses with 2 Bishop vs Queen and 2 Knights vs Queen but it's pretty rare.

With knight and bishop you can make this fortress https://www.youtube.com/watch?v=SOzdwycf8y0

Huh?

Ok I didn't remember that, thank you to show me my error.

It would be a fortress if the bishop were on b2.

Sería una fortaleza si el alfil estuviera en b2.

no tienes ni idea

The 2 bishops lose to the queen.  The 2 knights draw without much difficulty.

There is a theoretical fortress of two bishops against a queen.

Please visit my thread https://www.chess.com/forum/view/general/draw-trap 

jerry even says that you will never get this in your game so i wouldnt spend 2 much time on it

Yes, there are fortresses for the two bishops or the knight and bishop, but they are tough to reach, and impossible to reach from most positions with the two bishops.  The two knights hold the draw much more easily.

I understand that this position is an unbreakable fortress. What is the strategy for black to hinder white achieving this fortress and how can black win? I like to know that. Can anyone share a link? It is not explained in the book "100 endgames you must know".

Thanks

This analysis is wrong. White's 2.Kb2 fails to maintain the fortress because it blocks the bishop's access to b2. He should have played 2.Ka2 (see alternative line).  However 3...Qd5 4.(something) Qc4 is also then wrong for Black, in fact 3...Qd5 is already wrong. The fortress can be reestablished. Instead 3...Qf1, after which Black can break up the fortress. This is obviously well and truly broken by the end of the sequence shown (Stockfish as White in this sequence). Nalimov gives the final position as mate in 36. I didn't attempt to play it out because I haven't practiced this endgame much and it probably wouldn't have turned out to be mate in 36 (leastways not for Black).

I can't really answer that, but I've played the bishop and knight against Nalimov from a number of positions and my impression was that unless you can jump straight into one of a fairly limited number of positions that are more or less already in the fortress you can't make it. See the variation starting 2...Qf1 in the previous post.

As far as winning with the queen goes, in general terms the strategy is to take the bishop and knight (or the knight at any rate), but the tactics are not easy against Nalimov and you have to be constantly alert that he doesn't turn it into a bishop and knight endgame, as he would here for example:

I would say also impossible to reach from most positions with the bishop and knight. With a black queen Nalimov gives the percentage of theoretical Black to play draws as 1.71% and if you remove the positions where the queen is attacked by the knight or on the same diagonal as the bishop (to take out positions where there is already a fork or skewer) that drops to 0.05%. By comparison, similarly adjusted percentages for two bishops and two knights are respectively 0.46% and 4.21%. 

I would say that two knights hold the draw much more possibly rather than much more easily.

For example what should White play in each of these two knights v pawn positions to avoid losing? 

And how should play continue for the next 50 moves?

Not totally relevant, but endgame manuals usually describe two knights v queen as generally drawn. Fine (and probably many others) also describes two bishops v queen as generally drawn. 

Both probably are generally drawn, but the reason is that when they occur, in almost all cases, neither player has the slightest idea what to do.

The endgame manuals usually reach this conclusion by producing a fortress position and saying, "there you go - obviously drawn". This ignores @ram7000's question. He asks, "how can Black win (in KBNKQ)". The only answer I've been able to arrive at so far is, "with difficulty".

I don't too much like computers poking their noses into chess, but in this respect I'm inclined to  believe the EGTBs. Almost all queen v two minor piece positions are theoretical wins for the side with the queen. (It's just not so easy against an accurate opponent.)