How to get better at chess tactics?

That reminds me a bit of myself I also liked the attacking style of Alekhine and is my favorite player still. But squeezzing opponents by precise positional play has it's charms also.

But I don't have those type of problems like the kid in your example. I think my main problem is to find time for quality work, because right know what I'm doing is not quality work. Just fooling around with fast games...

You reckon you have trouble making progress?!?!

According to the site here Ive done over 26,000 puzzles.. according to the email Chess.com send me monthly I've done over 42,000. 

Thats over a period of about 1.5 years. And my rating on TT has gone up and down reguarly but overall slowly improves to I'm about 1800 now. 

Ive got a bit bored with them for now and I also find then incredibly frustrating sometimes.. so for the sake of my own sanity Im not doing them as often as I did before. 

But yeah .. maybe at 43 Im struggling to make that jump in pattern recognition and memory to get up to 2000+ I don't know. Guess I will try some more soon ;) Happy Chess to all! 

P.s I'm about at 1550 rated 5 min player .. maybe 1650 OTB.

P.p.s. anyone know why there would be a huge difference in my stats? Which one would be correct?

What do you mean by 'correct'?

If you are having fun with what your are doing then keep on doing it.  I brought the example because he is a nice young man, great parents.  And if he is happy playing "tactically" then so be it.  But to refuse to play and learn any other way, and to get upset when you cant beat players at a certain rating level is well immature.  But again he is 12.

Im not sure how many I have actually done ...

The "weekly report" email Chess.com send me says - 

Tactics: Your rating is 1803 (42062 problems over 281.3 hours)!

However - the Stats section on Chess.com says - 

@A_L_I_V_E

My rating is now pretty much beyond 1800 and if you look at my last 500 games I regurally win opponents 1800-2200. So, please just quit with the nonsense.

The main problem I obviosly had was the lack of self-confidence. I did not think I could actually do it. But when I tried I quickly moved up with my rating (currently 1941) and now regurally playing almost only 1750+ players. Some of my recent games:

http://www.chess.com/livechess/game?id=1048955380 (won vs. 2202 rating)

http://www.chess.com/livechess/game?id=1054370484 (won vs. 2136 rating)

http://www.chess.com/livechess/game?id=1059962802 (won vs. 2112 rating)

etc...

... also won large number of games against 1900-2100 players since I would not get this rating playing low rated opponents anyway. So, that's all the test I need. All this time I was probably better then I thought I am but just didn't have the confidence to play in that manner. End of story.

Acting like you know me, trying to define who I am and thinking I need to prove something to you? ... Sweet Jesus. You got some serious issues.

alive is a guy, block him and move on.

You're dealing with an idiot. Best to block and move on.

You're right. Blocked!

@tracer : you're worth 1000 times better than this frustrated anonymous idiot who is only 'alive' by trolling. Just ignore him and enjoy your chess :-)

.
Tracer2960
You won large number of UNRATED games against 1900-2100 players...

Hahaha..just unrated games. They traced from your best win on the game data record!

These are wise words.  I've recently suggested to a friend, a very positional player, to do 20 minutes of slow tactics from a book, 20 minutes of intermediately timed tactics (2 mins or so), and 20 minutes of super quick tactics on chessemrald.net (5 seconds or less) per day, and he jumped 200 points in 3 tournaments.

For my own part, I've did Michael de la Maza's training program (lots of tactical positions with increasingly quick sweeps through a cycle of 1,000 problems), and now I am doing slower, more calculated studies of combinations.  MDLM's program made me jump 250 points on most tactics trainers in 3 months.  I am noticing a steady climb as I work on calculation, board memorization, and blindfold problems.  I expect another huge jump in my ratings on tactics servers and software when I return to devoting my time towards tactics.  (I am doing a few other things right now, like "Chess Strategy for the Club Player" by Grooten, which I just finished, etc.)

This is a question I've recently started to wonder and this seems like a good place to ask it.

What is the value of fast tactics?

Currently, I do my tactics on chess tempo. I do unlimited time, and almost never, ever give up. If I get stuck, I take a break and come back. I flip the board. I take out an analysis board. Anything to get my mind going. I've found that while pattern recognition helps, the improvement in my calculation because of this has helped me a lot with my tactics. I mean, if I am really, really stuck on a problem, one of the things I do is count how many legal moves I have. This has gotten me out of many stuck problems after 10+ minutes of thinking. I literally look at every single legal move I can play, and often I'll see that there's one I didn't consider.

Whenever I play blitz, it is after I have done my "real" chess, and is just for fun. People say "blitz helps your fast pattern recognition." No it doesn't, at least not for me. All it does is test it. I've seen my fast pattern recognition improve much more by slow games and slow tactics. As a musician, I see it as someone saying "Playing through Moonlight sonata as fast as you can helps you play moonlight sonata fast." No no no no no! You have to play it slowly first to develope the habits, if you try to play it fast immediately, you just ruin your chance of learning it forever! To me, I see people who only play bullet akin to pianists who never get past Beethoven sonatinas because they aren't willing to practice slow. If you want to play Moonlight Sonata, you have to play it slowly first, and if you aren't willing to do that, you won't improve.

If I limited myself to 2 minutes on tactics, I would just use up my two minutes, then have to go to the next problem and never learn anything because I never force myself to think.

So I would like someone to explain to me the value of blitz to you pattern recognition and fast tactics to your tactic training. I don't challenge the value of it, it works for a lot of people, I just don't understand its value or the idea behind its value.

Xman720, you've just given a great explanation of why everyone should do slow practice and slow games first. But that doesn't mean you shouldn't go faster later.

You should google Michael de la Maza and see the training program mentioned by Milliern. He recommended getting a good set of 1000 tactics puzzles, and doing them 7 times, speeding up on each pass. So the first time through, you give yourself lots of time for each puzzle, though at some point, you do have to give up and look at the solution if you aren't getting it, just to keep the training moving along. I think he said to give yourself 5 minutes per puzzle, but a little more probably doesn't hurt, if you want to go for 10 to be more thorough. On subsequent passes, you should recognize them faster just from having seen them before. Eventually, you'll be able to do all 1000 puzzles in one marathon session in one day.

Personally, I think that's a bit extreme, especially since de la Maza recommended avoiding other types of chess study, which is kinda controversial. But I've done something similar with books of 250-400 tactics puzzles, and it helps a lot. Dan Heisman recommended this approach in one of his earliest Novice Nook columns.

The point is that you're memorizing common patterns and working on pattern recognition, not just your thought process. During an actual game, even if it's a slow game, you want the easy tactics and common patterns to jump out at you, rather than spending time and possibly missing some of them. And when calculating a deep combination, it's a lot easier if you can spot the smaller building block tactics that make it up fast and accurately.

That is partially my plan for slow tactics.

While a tactics puzzle takes me an average of 5 minutes now, my idea is that if I keep practicing, I eventually will get faster. It's just a matter of time and practice. I don't think there is any need to rush my learning by doing blitz tactics or giving up after a small amount of time. If I just keep practicing and keep forcing myself to think, I will get faster. Eventually my 5 minute average solve will turn into a 4 minute avergae solve and then a 3 minute avergae solve, and maybe one day a 15 second average solve. In that way, I think unlimited time control tactics practices my fast pattern recognition just fine.

Mostly, I agree with you, but I will make one suggestion.

You don't mention what types of tactics you're working on. If you're doing random puzzles from an internet site or something, and it feeds you puzzles based on your puzzle solving rating, then the puzzles will just keep getting harder, so you won't be solving them faster.

This is partially why it's recommended to go through the same set of puzzles over and over, so you get used to them and eventually spot the solutions nearly instantly. When you're done with one set, move on to a tougher set.

I've actually thought about that before. There are a few reasons why I decided to not worry about it.

1: A "harder" chess puzzle is, fundamentally, just a chess puzzle that combines two "easier" ones. So I'm still getting the practice if easier chess puzzles. A skewer in 1 is an easy chess puzzle, but in a skewer in 3 after 2 exchanges and a coersing check is a harder chess puzzle. As you can see though, no individual move was a difficult tatic, the puzzle is hard because it combines all of them.

Because of that, even as the puzzles get harder, I will never stop practicing basic pin/skewer/fork/mate threat/remove the defender etc.), they will just be combined in more and more complicated ways.

So in that way, I don't abandon patterns before I have the time to properly learn them. The patterns I learned 3 months ago in 1100 rated tactics don't dissapear in 1630 rated ractics, they are just combined and obscured. 

I like the idea of going through the same sets of puzzles though. I'm going to see about doing a program like that. I don't ever want to stop improving the difficulty of puzzles I'm solving, but on the side it may help to have a baseline of 2000 memorized patterns in the form of puzzles I've done back to back so many times I know them like the back of my hand.

First, you see more patterns.  Seeing more patterns helps you familiarize yourself with the abstract pattern from concretes, and then they become intuitive to you.  Looking back at my tactics ratings and progress from when I was USCF 1200, it amazes me that I see those tactics I used to miss --and I see them instantly now.  They are intuitive.  My tactics radar/aentenna pick up the pattern, and I don't even need to calculate.  I definitely agree with most of what Fromper says.

Yes, it is extreme.  However, I don't think that disqualifies it from being a successful training strategy.

I have graphs of my results on my website: http://milliern.com/2015/03/19/long-is-the-way-and-hard/ 

I also have another post on the plan to do it and logistics.