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

Keeping track of L&D study results
http://www.lifein19x19.com/viewtopic.php?f=12&t=13445		 Page 1 of 2
Author:  Koosh [ Wed Aug 03, 2016 8:58 am ]
Post subject:  Keeping track of L&D study results
I don't know if a dedicated conversation has been held here before on this topic. If so, please link me to it. :salute:

I've been thinking and trying to come up with a single most efficient way of tracking my study of life and death. As I know it, reviewing problems that one has gotten wrong a number of times in a timed manner is ideal. This also leads to "hard-wiring" - a term that I've read elsewhere in L19.

The problem I face is that which comes from reading too many books at once. When I put one down for too long and try to come back to it, I lose track of the sense I had of how much of the book I've completed and I become extremely tempted to just start from the beginning again.

When you are studying Life and Death problems from a book, how do you keep track of your progress? Do you mark your attempts directly in the book? Do you date those attempts? Do you review the book in full afterward, pulling problems marked wrong, or do you keep a separate notebook? What do you write in that notebook?
Author:  jeromie [ Wed Aug 03, 2016 9:34 am ]
Post subject:  Re: Keeping track of L&D study results
I typically face the same problem, but in a few problem heavy books I have used some light pencil marks to indicate whether I got a problem I attempted right or wrong. This is a nice way to keep track of my progress, and I don't find it too distracting when I return to the book.
Author:  DrStraw [ Wed Aug 03, 2016 9:50 am ]
Post subject:  Re: Keeping track of L&D study results
The simple answer is that I never did keep track. I just did the problems and then hopefully saw the results when I was able to apply them in a game. Bottom line is that it does not matter what happens during your study if you cannot apply it when it is needed.
Author:  jeromie [ Wed Aug 03, 2016 10:02 am ]
Post subject:  Re: Keeping track of L&D study results
I just thought of something else to add. I've recently been working through The Endgame by Tomoko Ogawa and James Davies. In chapter 2 of that book, there are a set of problems where the authors ask you to keep score by evaluating how many points each move is worth. Even though they provide space in the book to write your answers, I didn't want to ruin my ability to try the problems again on subsequent read-through. So I made a bookmark that allowed me to score the problems.

You could do something similar for life and death problems if you don't want to write in your books. You could even keep the bookmarks if you want to compare your performance on subsequent readings of the book.

Attachments:
Endgame Bookmark.pdf [14.49 KiB]
Downloaded 661 times
Author:  Kirby [ Wed Aug 03, 2016 11:05 am ]
Post subject:  Re: Keeping track of L&D study results
If you want stuff repeated based on time, you can use Anki.

that's not how I do problems, though. I just start at the beginning and go through it. If a problem is too hard, I mean might told the page corner and move on.
Author:  Koosh [ Wed Aug 03, 2016 12:44 pm ]
Post subject:  Re: Keeping track of L&D study results
@DrStraw: I've also done just this to get to 1D. I went through problem books and then promptly forget about them. When the goal is to get 95% right before moving on - which seems to be a key aspect of the training required to become a high dan player - then I need to find a way to effectively track and evaluate progress.

@jeromie: I appreciate you posting this bookmark! I remember studying my way through that chapter. I need to reread this book. I first read it when I was 20k and gained little from it. The second time I read it (SDK), it blew my mind. Maybe I can use this next time I read through it.

@Kirby: I like Anki, but I don't like the time requirement needed to build cards. More importantly, if my inherent problem is that I put books down for a time and then forget where I am with them, then forgetting to do Anki daily is going to put me in the same situation.

I have an idea of how I want to do this moving forward, but I want to see if anyone else has a systematic way that they study L&D.

Here's my idea.
1) Set up an Evernote account - this allows me to access the note from a computer or smartphone whenever I want/wherever I am.
2) Create a note for each book; list the publication details.
3) Write: First Attempt Start Date: ______, First Attempt % Correct: ______, First Attempt Problems Incorrect ______.
3) Use a list to create numbers (1,2,3,4,5....)
4) Mark an "o" next to problems I get right, and an "x" next to those I don't
5) Finish the problem set, total up the percentage
7) Mark next to the first mark for consequent attempts (oo, ox, xo, xx) (ooo, oxo, ect.)

Something like this: http://www.evernote.com/l/AMgOl16HMCNLR7whRW8DmfBM_7PaoitM_Ks/
Author:  DrStraw [ Wed Aug 03, 2016 12:56 pm ]
Post subject:  Re: Keeping track of L&D study results
Koosh wrote:
@DrStraw: I've also done just this to get to 1D. I went through problem books and then promptly forget about them. When the goal is to get 95% right before moving on - which seems to be a key aspect of the training required to become a high dan player - then I need to find a way to effectively track and evaluate progress.


All I can say is that I got to 5 dan using the method. I guess it may not work for everyone.
Author:  Bill Spight [ Wed Aug 03, 2016 1:56 pm ]
Post subject:  Re: Keeping track of L&D study results
Koosh wrote:
{T}o get to 1D I went through problem books and then promptly forget about them. When the goal is to get 95% right before moving on - which seems to be a key aspect of the training required to become a high dan player - then I need to find a way to effectively track and evaluate progress.


If that is your goal you are not challenging yourself enough. A good time to move on is when you are getting more than 60% right.

OC, moving on does not mean that you do not continue to overlearn problems that you missed. Here is the overlearning approach that I use. To quote myself from another thread:

Bill Spight wrote:
One learning technique that I used was overlearning, by solving problems that I had already solved. IIUC, spaced repetition is a form of overlearning.

The overlearning technique that I used works this way. Suppose that it takes N tries to a problem before success. After that you overlearn the problem by adding N/2 more successes. I kept track of successes by Xs and failures by check marks. (I also crossed through a check mark to make an X, after the first success.) Sometimes I waited so long between repetitions that I failed on problems that I had succeeded on before. ;)
Author:  Kirby [ Wed Aug 03, 2016 2:27 pm ]
Post subject:  Re: Keeping track of L&D study results
I don't understand the idea of getting only 60% right. If a problem is too hard for me, maybe I don't get the answer yet, but I don't somehow think the wrong answer is right (usually).

I would say that my accuracy rate is high for even difficult problems, because I don't think I'm done attempting before I have pretty high confidence that there are no refutations.

When you say 60% correct, do you mean that you never thought you had the answer for 40% of them?
Author:  jeromie [ Wed Aug 03, 2016 2:37 pm ]
Post subject:  Re: Keeping track of L&D study results
I don't know about Bill, but I get problems wrong sometimes. This can be because of lazy reading (i.e. with more time or focus I would have realized my answer was wrong - I should eliminate these kinds of error, since they lead to lazy reading in games) or an honest oversight. It also occurs because I can have different aims with different problem sets: sometimes I practice exhaustive reading, and sometimes I work through problems to expose myself to new ideas. With the latter purpose, spending a large amount of time to make sure I have an individual problem correct can be counterproductive since I am exposed to fewer ideas.
Author:  Bill Spight [ Wed Aug 03, 2016 3:02 pm ]
Post subject:  Re: Keeping track of L&D study results
Kirby wrote:
I don't understand the idea of getting only 60% right. If a problem is too hard for me, maybe I don't get the answer yet, but I don't somehow think the wrong answer is right (usually).

I would say that my accuracy rate is high for even difficult problems, because I don't think I'm done attempting before I have pretty high confidence that there are no refutations.

When you say 60% correct, do you mean that you never thought you had the answer for 40% of them?


Give yourself a time limit. :)
Author:  Bill Spight [ Wed Aug 03, 2016 3:04 pm ]
Post subject:  Re: Keeping track of L&D study results
jeromie wrote:
I don't know about Bill, but I get problems wrong sometimes.


Me, too. :)
Author:  Kirby [ Wed Aug 03, 2016 4:16 pm ]
Post subject:  Re: Keeping track of L&D study results
Hmm. I don't feel comfortable looking at the answer if I haven't solved it. Of course, I have no proof of the best method.
Author:  Bill Spight [ Wed Aug 03, 2016 5:24 pm ]
Post subject:  Re: Keeping track of L&D study results
Kirby wrote:
Hmm. I don't feel comfortable looking at the answer if I haven't solved it. Of course, I have no proof of the best method.


You don't have to look. If you have not solved the problem within the time limit, it counts as a failure. :)
Author:  Kirby [ Wed Aug 03, 2016 5:32 pm ]
Post subject:  Re: Keeping track of L&D study results
Bill Spight wrote:

You don't have to look. If you have not solved the problem within the time limit, it counts as a failure. :)


Sure. That's what I meant when I was asking about this:
Kirby wrote:
When you say 60% correct, do you mean that you never thought you had the answer for 40% of them?
Author:  Bill Spight [ Wed Aug 03, 2016 6:50 pm ]
Post subject:  Re: Keeping track of L&D study results
Kirby wrote:
Bill Spight wrote:

You don't have to look. If you have not solved the problem within the time limit, it counts as a failure. :)


Sure. That's what I meant when I was asking about this:
Kirby wrote:
When you say 60% correct, do you mean that you never thought you had the answer for 40% of them?


One thing I learned as a psych major is that learning tasks with about a 50-50 chance of success are generally preferable. I figured that if your chance of success is 60%, you could move on to somewhat more difficult problems. (OC, guaging the difficulty of go problems is hardly exact. :) But they are rated, FWIW.)
Author:  sparky314 [ Wed Aug 03, 2016 9:30 pm ]
Post subject:  Re: Keeping track of L&D study results
I don't keep track of problems, but get a feel for how many I've gotten wrong. I keep track of how many times I've gone through a book (completely).

If I only went through it half way, and put it down for a while, I'll start back at the beginning. If it hasn't been that long, and I have a bookmark there, then I'll pick it up where I left off (Get Strong at Endgame, I'm looking at you).

Then I just go back and re-read it after it's been a few months. It's a system thats worked well for me.
Author:  Koosh [ Thu Aug 04, 2016 8:21 am ]
Post subject:  Re: Keeping track of L&D study results
@Bill Spight: Thanks for sharing your thoughts. Your logic makes a lot of sense here. I also want to point out that the preface in Graded Go Problems for Dan Players suggests 70% as a good "goal", but I sometimes wonder about whether this figure is geared toward the casual player or serious student (and whether there is even a difference between them).

Giving myself a time limit is also a good idea. I am going to do this moving forward to keep things uniform. Also, your suggestion with regard to how many times one should overlearn a problem is a good guideline! I'll be adding that to my process. Thanks!

To your comment about how learning tasks with about a 50-50 chance of success are generally preferable - if you attempt a Go problem once or twice, and you don't see those variations in your head, that sounds to me like you have a 0% chance of success aside from simply guessing. I think what you're talking about here is with the selection of problem sets (a 5kg book versus a 3d book), and I completely agree with that.
Author:  Koosh [ Thu Aug 04, 2016 8:22 am ]
Post subject:  Re: Keeping track of L&D study results
Separately, I've found that there are many guides written about how to become shodan, but I've only seen a few written about going from shodan to high dan. I was able to dig up one of them. https://forums.online-go.com/t/repost-to-become-a-master-of-go-is-not-easy-but-to-become-an-amateur-5d-or-6d-is-not-hard/789 -- this is where I got the 95% number from.

Quote:
You must choose a proper book for you, don’t choose a very very hard book, if you only can resolve 10% to 20% problems, just abandon it. You should choose a book which you can resolve 60% to 80% problems, do it repeatly until you can easily resolve (clear every variations) more than 95% problems, then you should change a book.


Moving on when one gets to about 65% right in a book seems like a good idea (like Bill suggested), but if one does not properly track one's progress until getting 95% right while working through a new book, it seems likely to lose the edge that comes with being extremely disciplined and accurate with your problem solving (in game or out).

This would not apply to someone who just solves problems for fun. I fully recognize and respect those who solve tsumego because it's a fun activity. :mrgreen:

I started working with Graded Go Problems for Dan Player (Vol 1) again, from the beginning, using this process with one modification. I'm marking problems with a O(!) for problems I got right but missed a variation in my mind. Here are the results so far.
http://www.evernote.com/l/AMjUCaw69oVN1oxyJVEhPKM6iKLQc06eJ14/
Author:  sparky314 [ Thu Aug 04, 2016 8:44 am ]
Post subject:  Re: Keeping track of L&D study results
Koosh wrote:
[...]but if one does not properly track one's progress until getting 95% right while working through a new book, it seems likely to lose the edge that comes with being extremely disciplined and accurate with your problem solving (in game or out).


I don't track it on a per-problem basis. I work through books in pages. I work through all problems on a given page, reading each one completely (or so I think). Then I go back through each problem on the page take sure I have a clear picture of that solution in my mind (and re-reading the problem if I can't see the solution), before looking at the answers for those problems. It makes it easy to estimate success - did I miss 2 of 18 problems? I'm close to 90%. Miss 5 of 12 problems? I'm only hitting 60%.

And its a lot quicker than tracking specific problems. :mrgreen:
Page 1 of 2 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/