Tactics Training

I have been enjoying Tactics training and have a quesiton about speed for advanced players. 

At the tactics training, I can solve most all of them, but not within the time frame to increase ratings points. 

How important is speed in improving overall chess play?

At the chess tactics server, I can't break 1400 as I solve, but lose .2 for example, due to time elapsed. 

How valuable is this training once the patterns are recognized?

thanks

You might get a lot of different answers on this. My answer would be, accuracy is much more important than simply rushing through the problem.

A lot of people completely misunderstand the concept of "pattern recognition" and think that spending 2-3 seconds on every problem, getting around 45% of them wrong, still improves your tactical ability since you're seeing "many different patterns." I think that's complete nonsense.

In my online tactics training (not on this site), I try to keep the accuracy level at 90% (that is, I try to get 1 wrong problem maximum for every 10), and try to spend whatever time that takes to reach that percentage.

If speed is the area of chess that your are trying to improve then trying to recognize the pattern faster is what you should focus on.  Otherwise, accuracy is more important.

I think accuracy is much more important. Which is why I don't really use the tactics trainer here. There is one on another site where you can also do the problems without any timing involved (and get a separate rating from the mode in which time counts).

I think the timing is nice and it's helped my quick play by hurrying to look for hardline tactics, but I don't really pay too much attention to the speed. Rather make the best move...

Volume is important.  If you are not exposing yourself to enough patterns, it will take you too long to solve it.  And if you solve too slow, you are not exposing yourself to enough patterns. That's how these GM's play simuls.  They recognize all the positions amateurs throw at them without brute-force-solving. If you want to get faster at solving the positions, you need to expose yourself to more patterns and spend less time trying to brute force solve it.  One is training your mind in series (calculation), the other in parallel (pattern recognition).  Pattern recognition will always beat out calculation because it makes calculation 10x easier.

Calculation and pattern recognition are two different things. The amount of time you spend will decide whether you are training the former or the latter.

To go further, calculation will let you more accurately execute a combination in a game. Pattern recognition will keep you from considering a move that sets up your opponent to execute one.

I think that to train pattern recognition it is better to play through games quickly. Then your mind sees the patterns it's supposed to train (pattern recognition isn't confined to tactics, and if you fail a problem you don't even see the pattern happen).

kwaloffer--what site do you use?

You don't have to be fast at start but as you are solving problems you will evolve and you will start to recognize tactical patterns faster.

would consensus of advice for someone like me (under 1400) be to (on the Tactics Trainer site):

attempt both accuracy and speed----keep trying to increase tactics rating

slow down for accuracy, allow tactic training ratings numbers to go down

With speed--I am getting lots of problems, more than 100 per day.  I now do recognize a few patterns.  

Downside:  I am getting some answers right, but it is a quick quess --such as attack the invading queen, with no idea of follow up....

thanks for taking the time to answer a low rated player...the higher rated players are kind to take the time to dispense advice and I appreciate it.  

Don't move until you see all the moves. Guessing will help in Tactics Trainer but you can't risk it in a match. You won't gain anything by guessing.

Seamus, I think that at your level it's worth discovering the basic patterns with a book before using the various tactics trainer available. The difference is that online training doesn't allow you to drill the very basic patterns from repetition, as the tactical themes are mixed.

I would recommend going through these two books first :

- Bain's tactics for students

- Seïrawan's Winning tactics

I've been too lazy about this "book first" method. Repetition is key, and although I love to solve many Tactics Trainer problems a day on this site, I am not getting enough repetition to remember all of the patterns in a game. TT is most likely improving my board vision, while learning a few fork or checkmate patterns along the way.

I am going through Dan Heisman's Back to Basics: Tactics

When I finish with Heisman's book, I'll go through the following:

1. Chess Tactics For Kids by Murray Chandler - It's not just for kids.

2. Learn Chess Tactics by John Nunn - Really good!

3. Checkmate Tactics by Garry Kasparov - A new publication with great, color illustrations.

I think that when you first start doing tactical exercises, accuracy if far more important. When I teach chess, I teach pattern recognition. Tactical imporvement is based on the ability to recognize patterns. With Knight forks, there are key spacial patterns that, once recognized, can make setting up a fork easy. Obviously, if you're playing a timed game, timing is important. However, if you're simply trying to improve your tactical ability concentrate on accuracy.

I saw my book "Chess 5334 Problems Combinations, and Games"  by Laszlo Polgar advertised for $189.00 too. Strange, since they would have to know no body's going to pay that. I got it new from Bargain Books for $8.00.

Think of it this way:

You can't calculate forever, especially in seemingly innocent (tactic free) positions. So as you consider a possible move, and an opponent's response, and your response, etc., when do you cut it off? 

Pattern recognition is what allows you to recognize when the fifth ply in the series you are considering will end in a tactic for your opponent, without having to calculate the sixth, seventh, eighth and ninth ply. This is why it is a vitally important skill, and why you have to try to train yourself to see tactics instantly, not just be able to calculate them.

That's true : pattern recognition also helps you to determine if you need to look for tactics or not in a position, and in which direction to look.

by posting chess piece to strategically positions, learned tactics recognition may appear on the chess board.tactics trainer problem solving may increase our appetite for looking to tactics that maybe applied to a particular chess position.

How do you go about teaching pattern recognition? What curriculum do you use? I'm all ears. 

I use three books for building checkmate pattern recognition:

1. How To Beat Your Dad At Chess, Including the 50 Deadly Checkmates by Murray Chandler

 2. Checklist of Checkmates with Exercises by James Stripes (self published)

3. Chess , 5334 Problems, Combinations, and Games by Laszlo Polgar  

In Polgar’s book, the first 4462 problems cover mates in 1, 2 and 3. How many of those checkmate problems would I need to commit to memory in order to reach Class C (1400-1599), not counting all the other stuff, like other tactics and endgame technique?  The number 4462 is big, and intimidating!

Has anyone  gone through 1001 Brilliant Ways to Checkmate by Fred Reinfeld?  Would you recommend this book to a beginning chess player?