read attacking chess by josh waitzkin. he explains the tactics and how to use them. once you know of certain tactics you know what to look for. i studied openings and middle games and neglected tactics never did any puzzles. i was 1600 by doing that. then i hit middle games and endgames and it launched me to 1800 within 2 months.
You know, i've been reading these forums for years, mainly about how best to improve your chess, especially tactics. Some say read books, some say do puzzles on computer, some say just play blitz, or get a teacher, etc. Here's an idea: Why doesn't somebody with a chess background do an actual psychological study with groups of players, using different improvement methods, with some kind of ability measurement. I remember reading something in Scientific American by a chessplayer, but it was more of a compilation of anecdotal info with the conclusion that effortful study was the best method. The key word is "effortful." I suppose that means if you are really uncomfortable mentally, either because of the subject matter or the speed you are working at, then you are doing yourself some good. Who knows? But without some objective information no one can really say for sure what is the best method. Maybe there are several, all equally good. Maybe playing lots of blitz is just as good as studying tactics in books. Of course, it's fun to discuss, but i just wish there were some more objective information.
That would be nice, but it's probably hard to pull off in practice because people tend to use so many different haphazardly-chosen techniques to improve. Most people are not going to agree to use only one particular technique for a year just so you can study their improvement. You could gather some information by taking surveys about what people did in retrospect, but that may not be that accurate (people may neglect some factor that they thought was unimportant but actually was, their memory may be faulty, etc.)
There are several blogs by people who have followed the Rapid Chess Improvement/de la Maza technique pretty faithfully and tracked their ratings, so that may be of some use (see the other thread).
I'd like someone to explain to me how playing blitz would improve tactical skills for a novice.
I don't see how a beginner will suddenly discover 3 or 4 move combinations in a 3 minute game.
They may see the odd fork or skewer, but mostly they'd be either pre-occupied with their own simple attacks or not leaving pieces én prise!
Very few people suggest starting out learning tactics playing blitz for this simple reason.
Common sense dictates that you study the patterns, test yourself, then use your new skills in a practical way - ie play games.
When it comes down to how best to instill the knowledge & how to get the pattern recognition to sink in, it's much a case of horses for courses. Some prefer puzzle books, some books and a board, while others (like me) prefer tactics training software.
Logically, longer time controls are the way to start.
Blitz games may well have a place once these tactical skills are reasonably sharp as the shorter the time control the greater the bias toward tactics.
"...do an actual psychological study with groups of players, using different improvement methods, with some kind of ability measurement".
Ahh, there's the rub. You'd have to use *human* subjects... 😉
People have different learning styles and preferences- one plays bullet after midnight, another solves endgame puzzles while riding the bus. How would you control all these factors so that your results have validity?
I like the way humans are all messy and complex, myself.
Originally posted by buddy2 You know, i've been reading these forums for years, mainly about how best to improve your chess, especially tactics. Some say read books, some say do puzzles on computer, some say just play blitz, or get a teacher, etc. Here's an idea: Why doesn't somebody with a chess background do an actual psychological study with groups of players, using different impr f course, it's fun to discuss, but i just wish there were some more objective information.
Good idea, but rigorous studies take a lot of time and energy to do properly. I'm guessing the issue just hasn't gotten high enough up on the radar screen of the high-powered researchers. But yeah, I'd like to see something like that, too.
Here's a paper that I came across and posted in another thread a while back. It's somewhat relevant to your question.
Like other subjects of tests, you'd have to pay them, but i don't know how you'd stop them from "cheating" by playing a few games of blitz, or reading a book. People are messy and prefer to be that way.
I think Tactics can be divided up into 3 seperate (but interlinked) parts.
intution ("just knowing" X is a good move)
visualization (seeing [in your head] the combination)
pattern reconigition (knowing the position is good/bad.
now, the importance of each greatly depends on your goals, Blitz master? -- intuition is the way to go, OTB master visualization & patterns, corr master pattern reconigtion.
I think the best and fastest way to improve is simply to find a variety of exercises to train each.
Blitz is obvoiusally good for intuition, so are puzzles when the emphasis is on time (eg chessemerald)
vizualisation is (probably) a more global skill, in that there are a variety of ways to train it.
my hypothesis here is that If you took aside a checkers GM and told him to learn chess he would improve faster than average Joe (even when factoring in inteligence, determination and other varibles) because checkers has already trained his visualization.
if that hypotheisis is correct then I should imagine lots of different things can be done to train visualization -- perhaps something like math (eg. working out what 6321*23 is in ones head)
of course, solving chess puzzles would be more effective because you would be training intuition and pattern rec at the same time.
as for pattern rec -- going over games, playing games, puzzles, books etc could all be used.
Mad Rook mentioned some interesting studies above. I'm reading one now. The author mentions how Kotov is dead wrong about looking at one of the branches or possibilities only once. In real life, patzers and grandmasters go over branches several times, checking for errors. This makes sense to me. There's some other interesting stuff to be garnered from the articles.
07 Aug '07 19:51
how do you get good at tactics?
Back to Top