what's the right thinking process to find the right tactic

I once worked on a tactics puzzle from GM Aagaard. I stared and stared at the position but nothing came. I finally gave up and read the solution. It was a couple of forced moves that climaxed with White Qd5 forking Black's castled king on g8 and his queen rook on a8 which was loose because the QB was still on c8 like your position.

Aagaard commented that any master would notice the possibility in that configuration. Well, I'm not a master and I hadn't noticed, but I do now, although not consciously. I get a tingle when I see the diagonals to the king and rook are open like that, just like I do when I see a king open to a back row mate.

But that comes from experience, not a set of conscious rules for move selection.

When looking at the position, the first thing I notice are of course the obvious forcing looking moves. If I move the bishop the queen is attacked by a rook. That is duh of course...but because of that I start to look for ways the bishop can move with some threat besides just the queen.

I also chunk together various common tactical motifs. I see that if my queen reaches d5...AND if I get the enemy queen to not defend f7, then I will win a rook or smothered mate. That whole tactical sequence is something I see a lot so its almost like one move to me. So I see that moving my bishop will not only threaten the queen, but the queen may be fairly limited in its moves.

Another tactical pattern chunk that I see...because I am interested in moving the knight because I am interested in the white queen having access to d5, is Nb5-nxc7 Rb8 Bxa7, trapping the rook. Again that sequence is very common so it is like one move to me.

Something that makes finding the tactics here rather difficult is that there are so many different tempting ideas...for example the move f5 looks tempting because I can see a lot of potential action on the f file, especially coordinated with Qd5+.

I played the first move in the puzzle but I was really kinda guessing. I wasn't entirely sure it was better than Nb5...and I still just felt like playing f5...but I could also see that black has a lot of 'things' going on himself and therefore could probably repair his position if I don't do something very forcing...nb5 and f5 just did not feel like it came with quite enough of a threat. But I did not see that Qf6 was forced. In fact, I still don't. Qd7 Ne6 re8 looks unpleasant, but I don't see the win.

But anyway after Qf6 it was easy to play Nb5 because it was something 'in the air' from the start and c7 is unprotected. But again...I don't really see why qxf4 is the 'only' move.

I did not guess 19.qxq. I wanted to play 19 Qd5+ kh8 20 Bg3  with the idea of nf7+...and I still don't see why that's wrong.

nevermind...I see whats wrong with 19 qd5 and 20 bg3....black 'sacrifices' his queen taking twice on f1 and ends with Ne3+. That's an irritating swindle. I guess its a good thing I wasn't playing. But that tactical refutation would need to be found before you have an excuse to NOT play 19 Qd5+ imho...otherwise it would be the obvious move...not qxq.

The rook trap is a nice pattern to remember. I hadn't seen that idea before. Cool game.

I won't be back on this site until tomorrow (I have a tournament in 1.5 hours from now!), but I'll answer one question quickly: Yes, I would have gone for Nb5 in a real game. Tactics override positional play, and in a real game I would have *mentally* (in contrast to trying moves in an online puzzle) evaluated the various N moves at that point for tactical wins, and presumably I would have come to the same conclusion that there was no tactical win at that point. After checking for *tactical* threats, I then would have moved on to the second priority: *positional* threats by either player, whereupon I would have seen that Nxc7 move, and evaluated the result of that, and seen the continuation.

This is a very interesting and valuable type of question that we don't see much on this site, by the way: Getting people to explain their thinking/problemsolving processes. It would be nice to see a lot more of that type of question, maybe in a dedicated thread or even a group.

for the sake of shorteness I'll try to answer to everyone in one post

1) optimissed I'm not totally sure I understand what you say...

2)Ashvapathi  you surprised me a little I looked to f5 too but honestly Nb5 was not on at the top of my candidate moves.

3) ipcress12 thanks for your kindness, I wouldn't say that I'm really worried, I happened to have up and downs before, so it's not that I've lost hope in chess.

peraphs today I recovered some point in blitz although while I was playing I forgot the coffe percolator on the hotplate, I guess you can't have everything in life... 

Seirawan says that once he understood that 90% of tactics are forks  (or a combination that lead to a fork) it made it easier for him to solve them (I guess the remaining 10 are skeewers...) if you comibne that with looking for unprotected pieces it may be easier to find the solution you needed.

as a matter of facts combining fisheyesfools suggestion (what is not doing anymore the last moved piece) + the idea of double attack, I was able to clearly see the solution of the 3rd diagram that I failed to see on the game, (I thought that Qc2 was refuted by Rc8...) so maybe for the future I should focus  on combining more elements when looking for tactics.

4) Vercingetorix75 for sure too many options didn't made it easier to calculate, I wasn't really sure to haven't miss some obscure or invisible defense in every line... 

the rook pattern was sort of new to me...

Qf6 wasn't forced but all the other lines are inferior, still out of the 3 possibilities I didn't felt confident that I missed some defence for black... 

I think you really have a good point, you need to calculate everything before going into a line, although there are long term positional sacrifice that can be played this really doesn't look like the case...

5) K_Brown thanks

i would of played 17.h3 to neutralize the knight, moving back with a piece psychologically feels horrible especially if the opponent moves forward to avoid your threat.

I try to use Jeremy Silman's method of Imbalances from How to Reassess Your Chess, though I'm not good at it and often resort to looking strictly for tactics patterns.

The first thing that most of us see is the discovered attack on the Black Q.  We also see that both Queens are undefended.  We also see that the e-file can potentially be controlled by White's 2 R's and that would allow white to post a virtually impregnable N on e6, even if Black's QB trades one off. That would put a lot of pressure on the dark squares around Black's K, and after one N is exchanged for Black's c8-B, the other N can only be attacked by a N, if it's still on the board, that has a hard time getting to a square from which it can attack - otherwise only R's and Q or K can attack it, making it very hard to move.

Without having enough time to look forward (dinner's ready!) I think I would work toward putting the g5-N on e6.

I strongly believe that 'gestalt' psychology is be best approach to chess. I am very confident entering a chaotic tactical position even against a very strong human player simply because I have a massive number of tactical 'gestalts' memorized that pop up whenever I am close to them. In this manner, I cannot compete with computers who are able to calculate far more than any human is capable of...but I absolutely CAN compete with the strongest human players. And I can potentially defeat even strong GMs in complex tactical positions because I have absorbed into my subconscious an absurd amount of tactical gestalts.

but yes...you need to also preserve some degree of caution...because you need to see what really is and is not forced.  80% of the time a tactical refutation such as qxr+ would simply not exist in that position, but 20% of the time it does...and the difference is a win versus a loss...assuming your opponent sees it.

The reason it probably does not pop up in my gestalt is because it involves ne3+...which is significant because of the 'retrograde' motion of the knight, hitting d5. Such things are so uncommon that most many strong players have not included it in their subconscious 'repertoire'...just as kramnick overlooked a mate in 1 based on retrograde knight influence.

Here's one source, though I believe I've read it in mutliple places:

(p. 6)

FORCED PLAY

A series of moves where the opponent has little or no choice of
responses is called a forcing variation, as in the following
example:

(p. 8)
Now we are in a position to define exactly what we mean by
"forcing variation": A series of moves which alternate between
a threat and a response to that threat. The defending side cannot
break this chain of alternating threats and defenses without
losing material or allowing checkmate, until the forcing
variation has come to an end.
Of course it would be wrong to assume that all activity on the
chessboard consists of forcing variations. Without a doubt, most
of the time chessplayers dream up forcing variations and
sacrifices that never actually take place, often because the
opponent takes measures to prevent them. Sometimes the
tactical picture is as quiet as the strategic one, but all the same
one must be able to calculate forcing variations to see how they
might turn out, whether for good or bad.

The creation of forcing variations is crucial to a chessplayer's
development. In a sharp position involving many forcing moves,
variations with sacrifices may be worked out several moves
ahead.

Palatnik, Sam, and Lev Alburt. 2013. Chess Tactics for the Tournament Player. New York, NY: Chess Information & Research Center. 

sorry I missed that post, good luck with your tournament.

I guess that even in the case that I had seen Nb5 I would have avoided tactical complication, I'm a sort of coward chess player, I prefer find a small positional element to grind on rather than dive into uncertainty.

I don't feel to be properly qualified to answer you but it's definitely an interesting idea, I do agree that retreat moves are hard to find, as for my self (in open positions) before considering to retreat I analize sacs, becuause I expect to have more chances for compensation there then by retreating.

II think that although in a real otb match game I might have played Nb5 and won by the same method, I didn't realise when I looked at this that this was a computer puzzle that had been solved by an engine. It annoyed me that no consideration was given to Nde6 instead of Nb5, which is a natural line that one would examine, especially since it wins outright in two variations.

<<can you find the right continuation for white without an engine? and if so what was your thinking process?>> seems very dishonest, given that this was one of those annoying puzzles that I normally avoid like the plague because they don't attach any importance to other winning methods. After all, a win is a win and what is the right way for one person may not be for another. I'm still not convinced that 18. Nde6 is fine for black but then I don't have the benefit of having a chess engine. I'll have a look at it again.

ok, now I think I start to understand what you mean, I think you just looked quickly to the post and you didn't read everything, all of the 3 different diagrams are taken from real games that I've played (and sadly for me I've lost too) 

on the first post

>"[...] this is a 3 day per move game that I've just lost:

can you find the right continuation for white without an engine? and if so what was your thinking process?"

for practical pourpose I din't put all the lines in the diagram, after 17. Bf2  both 18. Nde6 and 18. Ng6 are fine and gives a plus 1.31 and  plus 1.18 respectively, Nb5 is evaluated 2.32 so I put the best line in it, the problem is that I didn't play Bf2 in my game... 

if you click on the diagram you can access the analysis board, otherwise you will have to copy and paste the pgn from the diagram and use this link: http://www.chess.com/analysis-board-editor

P.S.: I missed so many things in that game that's embarassing... after missing the advantage with 17. Bf2 I wasted the draw because I blundered a pawn, I really missed that move, well it may happen but for the grand finale I did hang the quenn...  if you solve the puzzle you can still see the whole game...

I think I can understand your bad mood, I tend to play quite bad otb and on long time control in live, more likely I will start to play for a team otb in the future and I'm already a bit worried, I usually play 70 or 80% of the game quite fine then I guess I run out of fuel and I make some obvious blunder... so somehow I guess I know the feeling...

Responding to the title, there is an AMAZING video on youtube done by chessedge, called "the six power moves of chess" that shows a couple of questions you should ask yourself when making moves. Most people that I've talked to have said that the biggest difference between an 1100 and 1500 is the ability to effectively use these

I think you mean this one: https://www.youtube.com/watch?v=V2v55mtUQkw

yep somenthing like that was already mentioned earlier but thanks for your contribution.

P.S.: since you like videos back in itme I really liked this one https://www.youtube.com/watch?v=EmJcUI_wSy8&t=449s it may be a little tough right now for you but it should become useful in the future, it's not all about solving tactics on the board you also have to create a position which will allow you to have tactical shots (so we're basically speaking about activity of the pieces...)

P.P.S.: I guess you already know the channel but Saint louis Chess club (the club organizing the sinquefield cup ) has imo a very fine channel on youtube from time to time I like to watch some lesson there...

OK, I'm back.

Whew. I looked at the position for about 15 minutes the first day, thought about it at work today, then maybe another 10-15 minutes tonight. If I'd been pressed for time without a good indication that White had a win then I would have just exchanged queens to save time on the clock.

There were few indications of a tactical win, especially of a mate, though after I studied the position I saw that Black had a few weaknesses that might be exploited:

(1)

Black's kingside is weakly defended while some of White's pieces are either aggressively pointing at it or in the vicinity of Black's king, so I looked at tactical shots like...

Bxh6

Ng6

Nxf7

and even...

Nd7

Nc6

...since I noticed Black's bishop is in line with White's rook on the e-file, suggesting a possible discovered attack. I also noticed that White's B covers Black's only escape square from the back rank, so I looked for back rank mate threats but couldn't find any. That d5-pawn really hinders White's B from being used in the attack. I considered ways to remove it or to distract the bishop, even crazy stuff like Qa6, but didn't find anything viable that would win. I couldn't find any fast win with any of those possibilities.

I then considered adding the Q to the attack--more force by declining the queen exchange and moving toward the kingside via Qd1 or Qc2, like maybe Qd1-Qg5 to threaten Bxh6. That latter looks most promising, but I still didn't see any forced win.

(2) 

Black has a number of related problems on his queenside, too. Black can't develop his QN normally. He loses that N immediately if he emerges at d7, and loses a P if he emerges at c6, so he would have to "rimize" it at a6 if he were to develop it at all. His QB is also blocked in by that d5-pawn. That QB is also hanging, so I considered ways I might attack it, such as Bh3, Bf1, Rc1-Rc7, but didn't see anything that would work fast enough to force a win before Black could develop. Also, Black's QR is trapped, but I didn't see any way to exploit it by getting my N to c7 or distracting the QB protecting it.

For a moment I thought I had a win from the key move Nd3 by trapping Blacks KB--Qxb4 ...Bxb4 Rec1 followed by Nd3 ...Be7 Re1 but then the B could escape to d8. That Rc1 move got me thinking along those lines, though, since I was so close to trapping that bishop if only it didn't have the extra tempo to escape to d8 I would win it. If only my R could escape the bishop's threat *and* stay on the e-file. Then I realized it could: all I had to do is play the slightly less natural (i.e., less coordinated) move Re2 to accomplish both tasks, and I'd win the bishop. But after playing it out I realized that Black still had a tempo where he could play ...Re8 since now *my* rook was hanging.

Somewhere in there, probably while debating how my N could attack the retreating B if it went to d6, I realized that Nc4 would be a terrific move under certain conditions, since Black's d5-pawn is pinned to the hanging B at b7, and if that fell the rook behind it might also fall.

The best I've been able to do so far is the following, but Black might have some better replies, and considering all those would rack up a lot of time.

P.S.--4. Rc7 looks like a faster win in my puzzle.

That's What I saw too.  In fact I couldn't even see any other continuation.  

Great post and thread!