Life In 19x19
http://lifein19x19.com/

Slacking off on my calculations
http://lifein19x19.com/viewtopic.php?f=12&t=15944
Page 1 of 1

Author:  Glummie [ Fri Aug 03, 2018 1:38 am ]
Post subject:  Slacking off on my calculations

Hi, i keep slacking off in my calculation. eg over reliance on my instincts. I did not calculate the variations or put much thought into my moves and felt like im kind of stuck onto SDK level. How do i force myself into calculating every move and possibility given the limited time.

Author:  Bill Spight [ Fri Aug 03, 2018 2:20 am ]
Post subject:  Re: Slacking off on my calculations

Why?

Author:  Calvin Clark [ Fri Aug 03, 2018 8:37 am ]
Post subject:  Re: Slacking off on my calculations

Glummie wrote:
How do i force myself into calculating every move and possibility given the limited time.


You can't. Even the top AIs can't. Have you seen how few initial candidates some of them explore? Let's say you have 40 minutes main time, typical for many tournaments in the US (although I am envious that the Europeans often have longer time controls.) Byoyomi sucks. Let's just face it: it's little more than a way to decide the result randomly when the players are getting too pokey. (Unless you're Cho Chikun.) So let's not even bother counting it. Assuming 240 moves in a game that goes to counting, 120 moves per player, that's 40 * 60 / 120 = 20 seconds per move, average. Shoot for finishing the game before byoyomi. Sometimes bad stuff will happen, and you can’t but try.

When people say they want to read more, they don't really want to read more. That's work. What they usually want is to either avoid blunders or find sharper punishments, but usually the former. In open, lower temperature positions, I think it's best to play faster than the 20 second per move average, in order to save time for harder, more critical positions where it takes time to read things out. While it may be good to set a minimum of, say, 5 seconds to avoid following your opponent around or making a different kind of blunder, I try to play many of the open position moves in under 10 seconds and save time for the critical fights. If no fight appears, it probably means you have to spend time estimating the score, which can be as time consuming as reading if not more.

I am finally accepting that I need to trust my intuition in open positions. Time and again, it turns out of that if I spend more time on such moves, I play something worse that if I had just played the first thing that popped into my head. And if my first feeling is wrong, then a) more thinking wouldn't have fixed it because it's probably a lack of knowledge or experience with that kind of position and b) even the mistake is probably not that big.

Even in tactical positions often that first thing I see is best, but then it can also be a horrible mistake, so reading is required to verify.

But there are other problems. One problem is spacing out and simply not trying hard even when the time is available. If that's the issue, well, I need help with that sometimes, too.

Keep at it and good luck!

Author:  Bill Spight [ Fri Aug 03, 2018 11:50 am ]
Post subject:  Re: Slacking off on my calculations

Calvin Clark wrote:
When people say they want to read more, they don't really want to read more. That's work. What they usually want is to either avoid blunders or find sharper punishments, but usually the former.


Good point. :)

Quote:
In open, lower temperature positions,


Open position are high temperature, where usually 13 - 16 pts. or so are at stake. (What most people call 26 - 32 pts.)

Quote:
I think it's best to play faster than the 20 second per move average, in order to save time for harder, more critical positions where it takes time to read things out.


That is similar to Rin Kaiho's advice, where he recommends saving around ⅓ of your time for one or two difficult positions. That does not mean playing the opening quickly. OC, if you play the opening quickly you can get into a lost position by move 50, which will save you time later on. ;)

OTOH, an 8 kyu will seldom benefit from taking more than 20 sec. for a move, even in difficult positions. Obviously, if your first play is wrong, followed by other inferior plays, deep reading does not do much good. Yes, you may find out that the result is not good, but then where is the mistake?

Quote:
I am finally accepting that I need to trust my intuition in open positions.


Developing their intuition is one of the best things that 8 kyus (Glummie et al.) can do.

Author:  Kirby [ Sat Aug 04, 2018 3:39 am ]
Post subject:  Re: Slacking off on my calculations

My thought is that practice off the board (problems, replaying games, or whatever) will improve your intuition so you can play quickly when time matters.

Author:  Bill Spight [ Sat Aug 04, 2018 5:14 am ]
Post subject:  Re: Slacking off on my calculations

To second Kirby

Some people think that you develop your intuition by playing quickly. Well, maybe so, but that doesn't mean that you are developing good intuition. ;) To develop good intuition you have to practice making correct plays, or at least good plays. As Kirby says, you can do that using problems, pro games, whatever. You get practice making good plays by review, by making plays that you think are good and by getting immediate feedback, for instance, from a book or from an sgf file.

One thing that I enjoy is to work on a problem that has an sgf file of the solution, perhaps with variations, and when I am ready, to play through the main line of the solution at a pace of about 5 sec. per move. At that pace, I find that even if a problem has been too hard for me, after a move or two I will usually see the rest of the solution. :)

If you replay a pro game at a pace of 5 sec. per move it will take maybe 20 min. If you try to guess the next move at that pace you won't knock yourself out. Then if you replay the game the following week, you will get more plays right. :)

I have never been a fast player myself, but Bruce Wilcox used to recommend playing a game in 15 minutes. At his go camps some SDKs would gain 4 stones in two weeks! OC, they weren't just playing fast games. ;) His stuff is still available online, according to recent posts here.

----

Since you are interested in improving your reading, here is a way of practicing it. There are at least three skills in reading, 1) choosing candidate plays, 2) calculating variations, and 3) evaluating the resulting positions. The last skill can be difficult, so you might practice on problems with a clear solution, such as tsumego, tesuji, or endgame problems.

First, choose what you think are the three best candidate moves, even if you think one or more is bad. That's OK. This is for practice reading. Then order the three moves and remember the order. Play the first move in your head. Now, do the same for the other player. Pick the three best moves, order them, and play the first move in your head. Do the same again. At this point you have made three moves in your head. Evaluate the position.

Then back up one move and play the second candidate move in your head. Evaluate that position. Then back up and play the third candidate move and evaluate the resulting position. Next you back up two moves and repeat the process. Continue in this fashion until you have played all 27 sequences in your head and evaluated all the resulting positions. (There will probably be fewer than 27 positions to evaluate.) Now check the solution of the problem, to get feedback on your evaluations and choice of candidate moves.

This exercise should bolster your confidence in your reading ability. :)

Author:  gaius [ Sat Aug 04, 2018 5:22 am ]
Post subject:  Re: Slacking off on my calculations

Quote:
An 8 kyu will seldom benefit from taking more than 20 sec. for a move, even in difficult positions


I agree that this is probably true, but I think a major exception is tsumego. Most games have one or two tsumego coming up during the game, and I think it really helps to sit down at those moments for a couple of minutes and see if you can read it out.

On the other hand, many situations are just too difficult to read out in the limited time that a tournament game offers you. Something that I keep having to remind myself of is that in many cases I'm just never going to find all the variations that I'd like to read out in the limited time that I have, and in those cases it's best to just trust general principles and good shape, maybe erring on the side of aggression rather than on the side of fear. Which can be really scary if you're in a neck-to-neck life-and-death situation :)

Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/