tactics and improvement

Tactics, just like chess study in general, is about quality over quantity.  It won’t matter how many tactics you study, it won’t matter how many hours a day you put in.  If you don’t thoroughly understand the problem you’re working on, it’s a waste of time.  If you aren’t understanding the theme, pawn structure, position, weakness(es), piece placement, moves that lead up to that position, you’re not going to fully understand what you’re studying.  This is why the idea that you need to solve tactics as fast as possible so your "online" rating doesn’t suffer is bull. 

If you’re serious about tactical improvement, do 3 things.

1. Set up the problems on a real board with real pieces.  If you choose to do them online.  Give yourself 2 minutes, and if you don’t solve it by then.  Set it up on a real board with real pieces. 

2.  Look for Forcing Moves: Checks, Captures, and Threats. 

3. Any problems you don’t get correct.  Take as much time as needed to ensure you thoroughly understand it. 

If i had to add a #4 it would be this.  Let the rating go.  If your main concern is an online number, youre only going to hold your progress back.   

One useful question to ask yourself when looking at a tactics problem is "What could I do IF...".

... if that defending piece was distracted.

... if I had one more piece in the action.

... if that line was open.

... etc.

Play "solitaire" chess.  Find GM games, or actually any titled players.  Play the moves for both sides through move 6 or 7.  Cover up the moves, and play the side that loses.  Try and figure out the moves made in the game.

What i do is using pen and paper, i write down the candidate moves, and lines of analysis for each move.  I then look at the move made in the game.   Its a real eye opener comparing your thought process to someone better.  And there is no better feeling when you get a move right!

Yes, tactical exercises will help your chess.  If you're under 2000, tactics should be the bulk of your training.  There are at least two big reasons for doing tactics:

1.  To improve calculation skills.  To this end, do tactics that are somewhat difficult for you and solve them completely instead of guessing at moves.

2.  To improve pattern recognition.  To this end, do lots of simple tactics and do the same ones over and over again until you can spot the answer to each of them in a second or two.  I recommend going through Susan Polgar's _Chess Tactics for Champions_ in this manner.  Do all the puzzles in the book 4 to 7 times.

I completely agree that intermediate players will learn more from studying games from the 1850s to 1920s rather than studying modern GM games.

Morphy to Lasker... chess got a lot more complex with Alekhine and Botvinnik, let alone Kasparov et al.

Maybe they're not doing it right?  It's important to do difficult tactics as well as repetitive solving of simple tactics.  Also, when doing tactics, don't move the pieces and try not to guess at moves.  You're done when you have solved all the important lines in your head.  I recommend solving tactics from books to avoid the temptation to guess at moves, although you can augment this with some computer tactics.  

Game play is also very important and shouldn't be neglected.  You're not going to get to 2000 without lots of game experience as well as excellent calculation and pattern recognition skills.  It's possible to pick up calculation and pattern recognition through games, but doing tactical exercises (in the proper way) is a quicker way to those skills.

Doing tactics will improve your chess skill, any player below 2000 will dramatically get better by by doing tactics everyday.

Lots of ways to game chess.com ratings unfortunately.  Also the same puzzles get served up all too often.  Chesstempo handles that sort of thing better--far less repetition and you get much less credit for solving a puzzle you've solved before.  At any rate, as I've mentioned, I recommend book tactics.

I don't think i've ever seen a collection of such puzzles - can you reference a book or something please?

First you get good, THEN you get fast.