How do you get 1337 at tactics?

Let's say I wanted to become really really, I mean really, good at tactics. How should I approach the problem. Let's face it, the weakest link in most of our games is tactics.

Do lot's of tactics trainer? Like 1 hour per day?

Or work on some tactics workbook or similar? Or both?

Play tactical blitz games?

How would your training program look like?

Get de la mesa's pep talk book (with it's 5 pages of study plan) and then follow his 7 circle thing to the letter.

Actually I think you can find the meat of it (those few pages) for free in .pdf form.  Anyway you basically get a set amount of tactics (say 1000 problems) and in the end you'll have solved them each 7 times.  The final day/final circle you do all at once and you have to solve tactics until "you sweat blood from your forehead" which if you're not willing to do, according to de la mesa your a wimp and should stop being such a cry baby or just go home now.

Anyway that's what you have to do, work hard = big pay off.

Yes, I've been thinking about this. 1000 problems doesn't seem much. Even in tactics trainer I've done so many that I simply solve positions because I recognize it and not because I actually calculate the tactic. With 1000 problems wouldn't this be a bigger problem? Or maybe it isn't a problem?

Well consider a book like Sharpen Your Tactics by Lein, you get around 1100 tactics but they're all different levels with the hardest ones tending toward the back of the book.  Just looking at your rating I have guess, but the first 300-400 will be fairly easy, problems 400-700 you'll probably be right at home/can't quite solve but then from 700-1100 you'll be able to solve maybe 10% of them.

So you get a review/too easy, then you get to exercise for a few hundred then you're basically going to memorize new types of tactical patterns/ideas in the last 400-500 or so.  Of course the first few times you wont be able to solve them, but as you repeat you'll slowly recognize the idea, then the first few moves but can't see the end, and by the 7th time you'll have memorized most of them (I'm guessing).  And I think this was De La Mesa's idea, through rigerous repetition you engrave tactical patterns deep into your brain so they pop out at you in your own games.

It's a tremendous amout of work for questionable overall returns, but it will certainly improve your tactics.

By the way "Sharpen Your Tactics" seems to deal a lot with sacrifices and not the more quiet tactics like working them into a pin it and win it type of situation.  The harder ones in SYT seem to be sacrifice for a king hunt.

So I'm not sure what set of problems you should use.  The tactics trainer on here is timed, so a lot of the "harder" ones I'm guessing aren't hard at all, it's just that you get a higher TT rating because to solve it in 5 seconds is so difficult.  You would be better off getting a book, although I'm not sure which, and De La Mesa doesn't recommend one.

Actually I might recommend "The Ultimate Chess Puzzle Book" by Emms which has many different types and some are quite difficult.

How about this one: http://www.amazon.com/Chess-5334-Problems-Combinations-Games/dp/1579125549/ref=sr_1_1?ie=UTF8&s=books&qid=1273271285&sr=8-1

I've seen this one but don't own it -- I have no idea the quality of the puzzles but you're getting more for the money and it's highly rated on Amazon / lots of good reviews so I don't see why it wouldn't be a good one to get :)

On Amazon I like to read a few of the 5 star reviews then read a few of the lowest rated reviews.  Sometimes you get a better picture that way.

Ok so let's say I settled for a tactics book. Books are good because it is untimed and you can practice your calculation skill. Would you still use the timed tactics trainer to drill patterns anyway? If so, would you split up the time used 50/50 between solving tactic books and doing tactics trainer online or in some other way?

Actually recognizing a tactic is not a bad thing. Being able to recognize thousands of patterns is not something to avoid. It's a goal. 

Hmm.  Well I guess there are two ways to do tactics.  In a book, you take your time and be sure to get it right.  I even write out my answer on a piece of paper so I can see exactly what I missed (or got right!).  With this method you're practicing calculation/visualization and when you can't solve them you're learning new patterns, both of which are important.

Solving under a timer is really different, and you're testing/developing your pattern recognition for patterns you already know.  Maybe you could also call this your tactical intuition.  In my opinion the untimed way is a lot better / leads to more improvement while the timed stuff refines what you already have to improve speed/intuition.

By the way, it is possible to do untimed training online at chesstempo.com

Maybe my explanation wasn't clear. I mean that I recognize the position on the board and know that the first move was move X. I don't see the tactic but I just remember the move. I'm not sure that I would see the right move that fast if for example you would take a position I know and change everything on the board except for the tactic (so it's basically still the same move that is the answer). So that is why I'm concerned that once I start remembering board positions and from there remember the first move that it doesn't actually help my tactical improvment.

This ofcourse is great for my opening knowlege. I have really easy remembering broad and deep opening lines even just seeing them once or twice.

Yes good points. I agree. I usually do my 20 endgame problems per day on chesstempo to keep my endgame technique sharp. Haven't done any tactics there but I was thinking of it because it is untimed unlike here on tactics trainer. Do you think there are enough problems on chesstempo for tactics that I could even just skip the book?

On the internet 1337 > 2500 :-)

1. I don't have it, just read the pdf's.

2. Yeah I got atleast 1 hour time and energy per day. Think 1 hour is extreme? More gives diminishing returns you think?

3. Yeah you'll never see me without a chess book in my hand on the buss or bathroom ^^

Also nice link. Thx.

If it's not, how come you can't beat a computer set for 100% tactics, 0% strategy and no opening book nor endgame tablebase?

Jeremy says so.

Just passing through and just in less than the past year have started playing chess again.  I've done more than a little looking around and spent more than a couple of dollars.

My guess is that you have tunnel vision.

There are chess visualization courses where you have a board setup in front of you and then are given x number of moves. What's the answer? Now how do you get checkmate?  A blank board is setup in front of you.  What color is f5?  What is the best way to move a knight from f3 to g7?  If you can't do this quickly let alone at all then this is one of your problems.

My rating is lower than it should because I'm learning and shouldn't be teaching but tactics is but a part of the game of chess.

Take it or leave it.  Look outside of the box

AFAIK, the top move in a Chess Tempo problem has to be something like +3 (pawns) better than the second-best move (at least for the first move of the problem), so I don't believe your criticism has much weight.

In my opinion, the problem with Emrald is that the whole interface is bad, and the problem with Tactics Trainer is that there's no untimed rated mode.  If it wasn't for Chess Tempo, I'd just use books.

I definately think visualization is important and I've tried to practice this by playing blindfold games. I manage maybe 10 moves before it starts getting blurry. I have no huge problems playing with 3-4 pieces on the board blindfolded either. But I definately could get better at it, it still requires ALOT of concentration.

Do you think I should train this every day aswell along with tactics? How would you split up an hour of training time per day between blindfold training and tactics training?

I'm not sure if you misunderstood my problem above. I was more worried that I remember positions instead of the actual tactic.

From what I understand there are two skills when it comes to tactics:

1. Visualization skill (=calculation skill)

2. Pattern recognition (=chunking large pieces, moves together)

I think one needs to improve in both. I'm intrested to hear how people would train to gain both. How to split up that daily 1 hour workout to get better at both.

Jeremy from purepwnage: http://www.purepwnage.com/home_outside