Life In 19x19 :: How does not playing equate to improvement?

I just don't get how taking a casual break from the game leads to sudden improvement. :scratch:

Abyssinica wrote:

I just don't get how taking a casual break from the game leads to sudden improvement. :scratch:

How do you know you had a sudden improvement ?
You are equating winning to improvement and losing to regression -- this is not true.

You can be losing many games and are still improving;
conversely, you can be winning many games are are still regressing.

More than 50% of what kyu players know is wrong. ( Don't belive this?

When you take time off, you forget things. Since the majority of what you know is wrong, when you forget stuff, the majority of what you forget will be wrong. Thus, by taking a break, you improve.

Joaz Banbeck wrote:

More than 50% of what kyu players know is wrong. ( Don't belive this?

True.

Joaz Banbeck wrote:

When you take time off, you forget things. Since the majority of what you know is wrong, when you forget stuff, the majority of what you forget will be wrong. Thus, by taking a break, you improve.

False logic.

Analogy: a beginner English student learns 100 new English words (boy, cat, dog, etc.). In the beginning he gets 20 correct and 80 misspelled. He takes a break and forgets 20%. Now he spells 16 right. It's no improvement. It's regression. Same with Go. Substitute language vocabulary with Go knowledge: basic shapes, basic life-and-death, basic tesuji, etc.

EdLee wrote:

Joaz Banbeck wrote:

Even after reaching the dan ranks, I know only a tiny bit more than kyus about basic shapes and tesujis.

Hi lemmata, I think once again it's time for Cargo cult science.

Speaking of which, Abyssinica, how is your data coming along ? Are you anywhere close to collecting a significant number of serious games (say 100) to show evidence of improvement, regression, stability, or other ?
Since it's only been 11 days since your OP, maybe too soon.

I get the feeling that they're forgetting vaguely stated pseudo-knowledge about go, which can be a hindrance to true progress.

Hi lemmata,

lemmata wrote:

What do you think?

Suppose we see a growth curve. With a break somewhere on it. To prove that whatever happens after the break is caused partially or entirely by the break,
we need to show that without the break, it would not have happened. How to do this ? It may involve a lot of data, with a lot of very carefully designed control groups.
Correlation and causality are two different things.
If someone, like Abyssinica, plays 100 serious games, and gets good feedback from reviews of those games, we'd expect this to affect the growth curve;
We're back to discussing how to improve, which is one of many perennial Go topics here and on KGS, etc.
With an endless supply of threads and discussion. Education. Non-trivial.
And of course, Cargo cult science.

lemmata wrote:

I get the feeling that they're forgetting vaguely stated pseudo-knowledge about go, which can be a hindrance to true progress.

Which is an interesting hypothesis. And which, if we are to test it, would require a lot of data, a lot of work.
But yes, to agree with you, from direct first-hand observation on KGS, especially with kyu to mid-dan reviewers,
there is an infinite supply of bull****, I mean pseudo-knowledge, about Go.

read it along with the Apology of Socrates next term. :tmbup:

Which level (grad school, undergrad, etc.) do you teach?

Author:	Abyssinica [ Tue Apr 08, 2014 5:20 pm ]
Post subject:	How does not playing equate to improvement?
Occasionally I get frustrated with Go to the point that I stop playing for a week. About this time last month in March I was on a losing streak to players in the 13-15k range who attach to almost everything move 2-5. I wrestled with the pig in the mud, got dirty, and lost a lot, so I decided to turn my kgs rank to - from 14k because I didn't need to worry about it for a little while. Throughout the month I didn't really do anything; I read a little out of A&D, the first two chapters of the direction of play, and the second book of go, not really studying. I played maybe 10 or so full-games with sdks. I watched dsaun a little bit. I did maybe 10 L&D tsumego. And now that I'm back, I've been able to beat 10/11k players in even games. I just don't get how taking a casual break (Not going cold turkey) from the game leads to sudden improvement. Does anyone else have experiences where they've stopped playing/actively trying to improve and suddenly came back stronger/better? Edit: Oh god, it actually doesn't sound like I stopped when I reread over this, but in comparison it feels like it!

Author:	EdLee [ Tue Apr 08, 2014 5:24 pm ]
Post subject:
Abyssinica wrote: I just don't get how taking a casual break from the game leads to sudden improvement. How do you know you had a sudden improvement ? You are equating winning to improvement and losing to regression -- this is not true. You can be losing many games and are still improving; conversely, you can be winning many games and are still regressing. Without a much higher level Go person to assess your moves, here is one way to check your level: play 100 serious games against opponents you think are around your current level, and look at the results. A mere 10 ~ 20 games are not enough data.

Author:	Abyssinica [ Tue Apr 08, 2014 5:28 pm ]
Post subject:	Re:
EdLee wrote: Abyssinica wrote: I just don't get how taking a casual break from the game leads to sudden improvement. How do you know you had a sudden improvement ? You are equating winning to improvement and losing to regression -- this is not true. You can be losing many games and are still improving; conversely, you can be winning many games are are still regressing. I just considered the fact that I was responding to bad moves with equally bad/even worse responses a sign that I'm doing really bad and should probably stop/take a break. People that seemed strong and far off/teacher-like before didn't seem that way any more after I tried to get back into it. Maybe I'm too receptive to winning/losing streaks and think every even game should be W-L-W-L-W-L

Author:	DrStraw [ Tue Apr 08, 2014 6:33 pm ]
Post subject:	Re: How does not playing equate to improvement?
I happens to people all the time. It gives you mind time to absorb new ideas without worrying about playing.

Author:	StlenVlr [ Tue Apr 08, 2014 6:51 pm ]
Post subject:	Re: How does not playing equate to improvement?
If you play a lot, taking a break is a good time to let all that playing experience sink in. I got from 4k to 2k by stopping playing for two months.

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

How does not playing equate to improvement? http://www.lifein19x19.com/viewtopic.php?f=12&t=10133	Page 1 of 1

Author:	xed_over [ Wed Apr 09, 2014 12:02 am ]
Post subject:	Re: How does not playing equate to improvement?
yeah, taking a break can often help with improvement, both real and perceived. try not to get sucked into playing your opponent's game, play your your own game instead (though you might start losing again for a little while, but eventually, you'll jump leaps and bounds above those old opponents).

Author:	Atheyon [ Sat Apr 19, 2014 12:52 pm ]
Post subject:	Re: How does not playing equate to improvement?
I have had the exact same experience. Are you me? I was stuck at 16-15 kyu for a while. I stopped playing for a month and came back as a 10 kyu. Now however, I am stuck at 10k :/. Though it appears that is slowly changing as I rid myself of the habit of playing very slow moves. That will take a long while. Maybe I should take a break again

Author:	Joaz Banbeck [ Sat Apr 19, 2014 3:12 pm ]
Post subject:	Re: How does not playing equate to improvement?
More than 50% of what kyu players know is wrong. ( Don't belive this? Ask any pro who teaches kyus. He will confirm it ) When you take time off, you forget things. Since the majority of what you know is wrong, when you forget stuff, the majority of what you forget will be wrong. Thus, by taking a break, you improve.

Author:	EdLee [ Sat Apr 19, 2014 3:55 pm ]
Post subject:
Joaz Banbeck wrote: More than 50% of what kyu players know is wrong. ( Don't belive this? True. Joaz Banbeck wrote: When you take time off, you forget things. Since the majority of what you know is wrong, when you forget stuff, the majority of what you forget will be wrong. Thus, by taking a break, you improve. False logic. Analogy: a beginner English student learns 100 new English words (boy, cat, dog, etc.). In the beginning he gets 20 correct and 80 misspelled. He takes a break and forgets 20%. Now he spells 16 right. It's no improvement. It's regression. Same with Go. Substitute language vocabulary with Go knowledge: basic shapes, basic life-and-death, basic tesuji, etc.

Author:	StlenVlr [ Sat Apr 19, 2014 4:29 pm ]
Post subject:	Re: How does not playing equate to improvement?
Unlearning whatever bad habits you have learned tends to be more difficult than learning new things. For that reason, taking break from go tends to work to your benefit.

Author:	lemmata [ Sat Apr 19, 2014 6:53 pm ]
Post subject:	Re: How does not playing equate to improvement?
EdLee wrote: Joaz Banbeck wrote: More than 50% of what kyu players know is wrong. ( Don't belive this? True. Joaz Banbeck wrote: When you take time off, you forget things. Since the majority of what you know is wrong, when you forget stuff, the majority of what you forget will be wrong. Thus, by taking a break, you improve. False logic. Analogy: a beginner English student learns 100 new English words (boy, cat, dog, etc.). In the beginning he gets 20 correct and 80 misspelled. He takes a break and forgets 20%. Now he spells 16 right. It's no improvement. It's regression. Same with Go. Substitute language vocabulary with Go knowledge: basic shapes, basic life-and-death, basic tesuji, etc. Hi Ed, But most kyu players don't know much about basic shapes, basic life-and-death, basic tesuji, etc. So they're not forgetting much of that. I get the feeling that they're forgetting vaguely stated pseudo-knowledge about go, which can be a hindrance to true progress. Even after reaching the dan ranks, I know only a tiny bit more than kyus about basic shapes and tesujis. This demonstrates how important even a tiny bit of knowledge about basic shapes and tesujis are, but a person can only forget what he knows and I think that kyus would be dans if they had that knowledge to forget in the first place. What do you think?

Author:	Abyssinica [ Sat Apr 19, 2014 7:48 pm ]
Post subject:	Re: How does not playing equate to improvement?
lemmata wrote: EdLee wrote: Joaz Banbeck wrote: Even after reaching the dan ranks, I know only a tiny bit more than kyus about basic shapes and tesujis. I don't know where I heard this from, but someone said "When I was 8k, all I thought I had to do was learn more joseki, learn more tesuji, get better reading, and that's all I need to get passed 1d and the rest of the dan ranks. But when I got to 2d I realised I was way off and there was a lot more to it." Obscenely paraphrased, of course.

Author:	EdLee [ Sat Apr 19, 2014 8:58 pm ]
Post subject:
Hi lemmata, I think once again it's time for Cargo cult science. Speaking of which, Abyssinica, how is your data coming along ? Are you anywhere close to collecting a significant number of serious games (say 100) to show evidence of improvement, regression, stability, or other ? Since it's only been 11 days since your OP, maybe too soon.

Author:	Abyssinica [ Sat Apr 19, 2014 9:55 pm ]
Post subject:	Re:
EdLee wrote: Hi lemmata, I think once again it's time for Cargo cult science. Speaking of which, Abyssinica, how is your data coming along ? Are you anywhere close to collecting a significant number of serious games (say 100) to show evidence of improvement, regression, stability, or other ? Since it's only been 11 days since your OP, maybe too soon. Here is my rank graph; the start after the gap is about the same day I made this thread. I have played 29 ranked games and somewhere between 29-35 "serious" games since. I am trying to kick the habit of playing on autopilot and severely underestimating my opponents. (That is where the drop happened)

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

Author:	moyoaji [ Sun Apr 20, 2014 12:08 am ]
Post subject:	Re: How does not playing equate to improvement?
I notice that your graph looks like it simply continues on as if there was no interruption in play. Taking a short break seemed to neither hurt nor help your progress. I think there is something to be said for taking a break from any skill. You can over-practice something and so giving your mind a rest can be like resting muscles after exercise. Or maybe think of it like sleeping - your brain is still working, just not consciously. Although sometimes my brain's work on go is conscious. Many times I will think about go when I am bored. I might be on a bus or in a waiting room (or maybe someplace where I'm supposed to be paying attention but just aren't ) and I'll conjure up board positions. Think about life and death scenarios. Stuff like that. It is like my brain is always working on the game somewhere. So even when I've taken a break from go I never really escape the game. It is always with me.

Author:	EdLee [ Sun Apr 20, 2014 12:35 am ]
Post subject:
Hi lemmata, lemmata wrote: What do you think? Attachment: Kheldragar-en_US.png [ 16.86 KiB \| Viewed 11961 times ] Suppose we see a growth curve. With a break somewhere on it. To prove that whatever happens after the break is caused partially or entirely by the break, we need to show that without the break, it would not have happened. How to do this ? It may involve a lot of data, with a lot of very carefully designed control groups. Correlation and causality are two different things. If someone, like Abyssinica, plays 100 serious games, and gets good feedback from reviews of those games, we'd expect this to affect the growth curve; We're back to discussing how to improve, which is one of many perennial Go topics here and on KGS, etc. With an endless supply of threads and discussion. Education. Non-trivial. And of course, Cargo cult science. lemmata wrote: I get the feeling that they're forgetting vaguely stated pseudo-knowledge about go, which can be a hindrance to true progress. Which is an interesting hypothesis. And which, if we are to test it, would require a lot of data, a lot of work. But yes, to agree with you, from direct first-hand observation on KGS, especially with kyu to mid-dan reviewers, there is an infinite supply of bull****, I mean pseudo-knowledge, about Go.

Author:	Abyssinica [ Sun Apr 20, 2014 11:29 am ]
Post subject:	Re: How does not playing equate to improvement?
In the end, does anyone truly have accurate knowledge about the game?

Author:	lemmata [ Sun Apr 20, 2014 11:52 am ]
Post subject:	Re: How does not playing equate to improvement?
EdLee wrote: Hi lemmata, lemmata wrote: What do you think? Attachment: Kheldragar-en_US.png Suppose we see a growth curve. With a break somewhere on it. To prove that whatever happens after the break is caused partially or entirely by the break, we need to show that without the break, it would not have happened. How to do this ? It may involve a lot of data, with a lot of very carefully designed control groups. Correlation and causality are two different things. If someone, like Abyssinica, plays 100 serious games, and gets good feedback from reviews of those games, we'd expect this to affect the growth curve; We're back to discussing how to improve, which is one of many perennial Go topics here and on KGS, etc. With an endless supply of threads and discussion. Education. Non-trivial. And of course, Cargo cult science. lemmata wrote: I get the feeling that they're forgetting vaguely stated pseudo-knowledge about go, which can be a hindrance to true progress. Which is an interesting hypothesis. And which, if we are to test it, would require a lot of data, a lot of work. But yes, to agree with you, from direct first-hand observation on KGS, especially with kyu to mid-dan reviewers, there is an infinite supply of bull****, I mean pseudo-knowledge, about Go. I agree with you on the principles of sound arguments and scientific inquiry. I was throwing out a lazy alternate theory, something that might be described as entertaining cocktail party conversation fodder, which I happen to believe at the moment because I have no stake in its truth/untruth. By the way, thanks for that link. That was thoroughly entertaining. I bookmarked it and might make my students read it along with the Apology of Socrates next term.

Author:	EdLee [ Sun Apr 20, 2014 12:24 pm ]
Post subject:
lemmata wrote: read it along with the Apology of Socrates next term. Hi lemmata, that's nice. Which level (grad school, undergrad, etc.) do you teach ? Is it philosophy ? Dr. Richard Feynman is one of my all-time favorites.

Author:	lemmata [ Sun Apr 20, 2014 6:14 pm ]
Post subject:	Re: How does not playing equate to improvement?
EdLee wrote: Which level (grad school, undergrad, etc.) do you teach? It's a college course. I'll PM you the brief details since they are irrelevant to the thread.