Past problem rears it's ugly head again

Last year at the club I got really frustrated because I could not beat players in a G30 game that I could clearly outplay. I could get a winning position but would run out of time. If only I could think faster was all I could say. I solved that temporarily by playing a lot of 5 12 blitz and positional study helped me make decent moves quickly in non-tactical postitions. Tactical study and pattern recognition has definately played a part as well.

Now I am having a similar problem. I can play into a position where I have a lot of pressure on my opponent and it looks like I have an overwhelming attack but I cannot calculate to checkmate, advantage or quiet postion. So now I am trying to come up with some guidelines where a sacrifice or all out attack would be in my favor. Obviously there is no substitution for calcluation but I am lucky if I can 3 or 4 moves deep in a complicated postition.

So 'The art of Sacrifice' is sounding very interesting to me right now. Too bad I don't have it. Anyway I am trying to come up with some guidelines for when to start my attack when calculation fails.

I have a few to try out.
I am reading 'The Tao of Chess' right now. I think the book sucks as far as some of the examples they give. But it is good for me to come up with new ideas.

One idea I got from the book is a Queen and a Rook is enough to mate a King in the open. The examply sucked in the book and what exactly is 'in the open' and how many defender can be around? Ovioulsy these questions can not be answered easily. At lest the example does give some clue.

So this is how I will apply it. If I can trade down my attack to where I have a R and Q left with my opponent having 1 or less pawns left and 1 or less pieces to defend with then I will pursue my attack.

What I have come up with on my own is to count the number of pieces I have in the attack and the number of pieces he has in defense. If it is 1 greater then I should consider the next step of how many pieces can I easily bring into the attack and and how many piece can he easily bring to defense. And finally do either of us have any inactive pieces and then if things look favorable start an attack.

I am already very comfortable making a B sac on h6/3 (it makes me salivate when they put a little ear on their castled K) with the QB battery if I have some other piece to support bring into play etc. I have also learned the valuable lesson of leaving the the Q on h6/3 where King is one check away from mate. In the past I have foolishly moved my q to give check on the g file and allow his R to come into play. Sometimes that has been quite disasterous.

I would appreciate any other guidlines you guys use when deciding to attack.