Tactics on books versus Tactics on the board

Is there any problem with studying tactics on books and translating the patterns to over the board play or Should we, to achieve better results, set every problem, even the easy ones, on physical board ?

Don't waste the time recreating a book position on the board! Solve it by looking at the book position, then go immediately to the next problem.

Thank you. I see your rating in tactics here is pretty high. Don't you face any difficulty spotting tactic motifs on the real board during games ?

"Chess is 99% tactics", said Richard Teichmann. To this I might add, "tactics is 99% pattern recognition". 

It all comes to practice. Chess.com's Tactics Trainer, the website chesstempo, books and Convekta's CT-ART tactics CD are the tools I used to improve. It did translate into skill on the real board, as I was able to beat the guys at the park fairly regularly.

For an amateur, the fastest way to improve is to invest time in tactics training, IMO. Studying openings & endings may come later on.

You can do it in two steps: first try to visualize everything in your mind, only using the diagrams presented in the book as reference. Then play it out on the board, to see what you missed and why. Rinse and repeat. I would recommend doing it only for harder problems though, otherwise you will burn out.

If you solve problems easily then don't need to OTB practice but for hardest one OTB is best for everyone.

i prefer OTB since am a beginner and need to familiarize with myself how to calculate OTB

I disagree with CrazyChessPlayah. For the best results from training, make sure it mirrors tournament practice as much as possible. Therefore, when you work on tactics, unless they're extremely simple, always set up the position on the board and use a clock to time yourself.

How many puzzles you should do in a session is entirely dependent on the level of difficulty and the book you're using. I'll give an example of my own training when I use 'Combinational Motifs' by Maxim Blokh (Alexander Grischuk's first trainer), the best tactics puzzle that's ever been published in my view. This book contains 1,200 puzzles, divided into motifs, with six positions per page, many of which have solutions for both White and Black ( so in effect nearly 2,400 puzzles), which is great because you're forced to look at what both sides are up to. Most of them are taken from genuine games (or sidelines/analysis from genuine games). The level of difficulty is rather high and typically increases with every position on each individual page. He uses a point system from 1 to 12 to indicate the level difficulty. The way I do it is this: I sum up the points on the page I'm going to work and the set chess clock accordingly, adding an extra six minutes (1 min/position) for setting the positions up [I leave the clock running. Clearly you can stop the clock between each, but I use this as another form of training, piece cluster pattern recognition. (For this, it's important that the positions aren't composed, as opposed to chess problems and some endgame studies which can be rather 'outrageous' in terms of divergence from proper games.)] For anything up to 24 points (basically low level of difficulty) I set the clock at 30mins. For every extra 5-point interval, I add another 30mins on the clock. A session normally varies between 30mins and 2h30mins. Then I work on the positions, with great focus (as if they were games) one by one, noting down the solution on paper. I always try to find the solution, but also I try avoid 'losing' on time, so sometimes I skip a position, in order to return to it once I've solved the others. Then I check against the key.

In my view, this is great practical chess training.