I have not failed. I've just found 10,000 ways that won't work.
Tiles of wisdom
Get link
Facebook
X
Pinterest
Email
Other Apps
In the enthousiastic violence of our debates the essence of the matter is all too often obfuscated by interesting sidesteps. In order to to keep them for posterity I have put them on tiles.
These tiles are a nice idea. Some of them are contradictionary, but I guess you know about that, too.
I liked this one very much: "deliberate practice excludes autopilot".
Am I on autopilot? Or do I memorize these puzzles? And what is the difference? The trouble is - I often dont know. If it is on autopilot, I should be still able to do it very fast after I have not seen it for a year or so.
"first do it right then do it fast". Though I know what you mean, this is probably wrong. It should be more or less be this, dont you agree?: "first look at what you did wrong. Second Step: do it right. Third Step: do it fast."
About recognition I am a bit on splitting hairs and would rather put it like this: "Recognition is about understanding, Rote learning is not". This is what I think you totally miss in our discussions, since you think my sentence implies what is in your sentence. However, to see the similar requires requires to see the difference, which requires to have a good understanding about the same. All crows looked the same to me. Then it happened that I found a young crow abandoned. Its mother did not come back, and the next day I took the very little almost dead crow home and raised it myself. Now I know, that they have faces, they all look different, and I have the impression their voices are different, too. When my crow was adult it flew away, but came back for some time. I could always tell which one it was: the one with the slightly bend beak/pecker.
I'm sure that you and Aox are smart enough to pinpoint every contradiction and idiocy in this set of tiles. But can you be even smarter? Find how to reconcile the contradictions and to ignore the idiocy without annihilating common sense? That would help to find the essence.
Have you any idea what I'm trying to say with "Recognition is about the similar, Rote learning is about the same"?
Can you forget for a moment that I don't seem to understand you? Ridiculous, of course, but just for the sake of reasoning?
You seem to be troubled by the expression rote learning. But you do you use spaced repetition, right? And spaced repetition is about committing something to memory. I call that rote learning. But maybe we should call it repetition? That should read:
Recognition is about the similar, Repetition is about the same.
I thought about it. I start to believe, that rote learning in chess puzzles does not exist.
I have a prove for this: Aoxomoxoa developed a board memorization tool. You look at the board for a few seconds. Then one piece gets taken away. Which piece is it? You would not have a chance if you dont try to quickly make some sense of the structure of the board position. Rote learning would mean we could recall all the pieces by just looking at them without trying to make some sense (understanding) of the position. Randomly shuffeled boards with even illegal positions are very difficult to remember. With rote learning it would not make any difference if the board position makes sense or not.
So: we dont rote learn any of the puzzles at all. None of them.
Our mind is efficient, though. If we first had 3 key thoughts to solve a puzzle - 2 key thought for the main variation, and 1 key thought for a side variation - we only keep the 2 key thoughts to find the main variation. It leads to "I know what the solutional move is and I remember the CT players "stupid" queen sac to avoid the mate threat. But I cant see (remember) the mate threat anymore." That means, that we will forget some information of a puzzle we desire to learn with spaced repetition. But we dont forget all of it. We will maintain enough understanding for the main variation, and by the sheer number of puzzles we solve we will cover (hopefully) the dropped of thoughts in some other puzzles. In this respect it is probably more helpful to have puzzles that only contain one (new) key thought, pointing again very much to look at easy puzzles instead of hard puzzles. For a master a 1800 rated puzzle is "easy" and the (new) key thoughts he needs to remember are less in number than they are for a weaker player with less knowlege. Improving with hard puzzles is still possible, but I guess to make up for the dropped of thoughts we need an increasingly high volume (number) of puzzles to make up (to cover) the thoughts that get lost in the side variations.
Repetition of puzzles will lead to remember the puzzle with less effort. Our brain cuts away unneccessary informations for that particular puzzle, as well as it may derive simple rules that dont even make sense in a chessability way but just makes it simpler to remember the puzzle (for instance: This is the puzzle with the tripple pawn on the a-file and the solutional move is placing s.th. on the top of this tripple pawn). As long as it makes it simpler, our brain is likely to use this in order to memorize it. Yes, that sounds like rote learning at the end of several repetitions. However, we cant rule out, that it is easiest for the brain to simply memorize the key thought for the main variation, because that IS the most simple way for our brain. Unless you train very hard complex puzzles...
you asked several question, and those who are not rethoric questions are answered.
Your question about a comparison to recognition and repetition and rote learning assumes very much the idea, that there is s.th. like rote learning - which I deny. Part if your meaning is that rote learning exists. The whole meaning gets flawed though, if this assumption does not hold true.
So I answere your question, but probably not in a way you liked to hear it. Reminds me of the joke: Are you temposchlucker or are you not temposchlucker? Answer me with a simple "yes" or "no"!
(or the other way round: "are we going to survive or are we going to die?" --> "yes!")
@Munich, If I translate that to an answer that reads as "you have no idea what I mean by it".
A few days ago I said that both Aox and you, and me are like bishops of a different color. We live in a different universe.
That started me thinking about the posts of the past months. I realized that I always wondered that your comments had no relation to what I meant to say. The comments are interesting enough themselves, don't get me wrong, they simply had no relation with the essence of what I meant.
Both you and Aox are integer guys, so I realize that you made an honest attempt to understand what I mean and that you not just wanted to make my life miserable.
The same is true for me. I really want to help you to understand what I'm saying. I don't mean to offend you.
So I decided to write the essence of my posts from the past months down in the form of tiles of wisdom. And I intended to cut off any distraction in your reactions which lead away from this essence.
But now I see that there is a fundamental problem that I simply don't come across.
I see you falsifying an assumption of what you think I mean. But your assumption is wrong. That way we continue to make itchy comments over and over again until we become irritated. I don't want that to happen.
We live in a different universe and I don't know how to bridge the gap. I hope that the non commenting readers of my blog don't have the same problem. The few I speak via e-mail don't seem to suffer from it.
I think we live in the same universe, its the universe where the results counts. I am not interested in wisdom ;) i want to improve in tactics ( and then use the method to improve in chess ) I dont care of "the colour of my bishop" if i can mate with him.
So folks: tell me you methods, then tell me your results and then i try to brew my own recepie for improvement :) ( you see i am one of the nasty parasites.. )
At THIS moment SEEMINGLY only munichs method works for me.
MAYBE Tempos "new" method will work too, but i just cant see any indications for that right now
Aox, I did improve, no doubt about that anymore. Even within the last 3 months I improved. If I extrapolate the last 3 mothns into a year, than I currently have an improving speed of 133 CT Blitz rating points per year.
Here my analysis about my performance/improvement in detail, it looks very promissing indeed:
I don't find the figures convincing at all. Munich improved about 150 points in 21 months. If someone is new to tactics he will allways improve about 150 points during the first time. Regardless the method.
It's evident we don't come any further in this discussion. Let's make an agreement that we don't talk about it anymore until I have surpassed the 2000 hallmark. Official Dutch rating or FIDE rating, that is.
If you comment, I will place it, but I will not react if it is about this specific subject.
My first 823 puzzles (under Blitz mode, I started with Standard) with the end date ...13.02.2011 had an average CT Blitz rating of 1738. These I am willing not to count but to assume that any improvement here might have been caused by getting used to Blitz mode.
But dont you underestimate that, tempo! You did so far just 726 puzzles under Blitz mode, so please do another 100 and then we can talk that you have a start rating.
From 13.02.2011-13.03.2011 I did another 1423 puzzles with an average rating of 1813. It includes my All-time-high of 1857, which was reached with an RD of close to 70. In fact, only a few reading had been above 1850. (But it is enough to consider it as an active rating).
Up to the date of 20.04.2011 I did another 648 puzzles. These 648 puzzles had an average of 1797.5. That is 15.5 points less than the average before, probably due to the duplicate reward reduction setting allready in at that volume.
I started my training based on deliberate practice, tag-themed, easy ranged, aiming at solving times below 20 seconds on the 21.04.2011, and since then I did not change much anymore. Before this I trained differently.
Till this very date I did 2798 puzzles. You are missing 2000 puzzles until you reached this number. After 2000-3000 puzzles, most matured adults dont improve any more.
But I did. About 100 CT Ratig points from average to average with my new training. Within 16 months.
I hardly know of anybody, who is matured above the age of 30 who achieved this. But if you find somebody, I am curious about his training. I found somebody, though. As it turned out, he had a similar training to me.
My chesscube rating rose due to my training about 150 points. In the club I am stronger, too. I started to be able to compete against the other experts in Blitz games, but have no experience in OTB games. But anyway, I was not used to win against expert players before, so this is an indication, that my improvement in CT tactics translates into real chessability.
I am curious how you will think about any CT blitz rating improvement after you did 2000 puzzles. Then we talk again...
These tiles are a nice idea.
ReplyDeleteSome of them are contradictionary, but I guess you know about that, too.
I liked this one very much:
"deliberate practice excludes autopilot".
Am I on autopilot? Or do I memorize these puzzles? And what is the difference? The trouble is - I often dont know.
If it is on autopilot, I should be still able to do it very fast after I have not seen it for a year or so.
"first do it right then do it fast".
Though I know what you mean, this is probably wrong.
It should be more or less be this, dont you agree?:
"first look at what you did wrong. Second Step: do it right. Third Step: do it fast."
About recognition I am a bit on splitting hairs and would rather put it like this:
"Recognition is about understanding, Rote learning is not".
This is what I think you totally miss in our discussions, since you think my sentence implies what is in your sentence.
However, to see the similar requires requires to see the difference, which requires to have a good understanding about the same.
All crows looked the same to me. Then it happened that I found a young crow abandoned. Its mother did not come back, and the next day I took the very little almost dead crow home and raised it myself.
Now I know, that they have faces, they all look different, and I have the impression their voices are different, too. When my crow was adult it flew away, but came back for some time. I could always tell which one it was: the one with the slightly bend beak/pecker.
I'm sure that you and Aox are smart enough to pinpoint every contradiction and idiocy in this set of tiles. But can you be even smarter? Find how to reconcile the contradictions and to ignore the idiocy without annihilating common sense? That would help to find the essence.
ReplyDeleteHave you any idea what I'm trying to say with "Recognition is about the similar, Rote learning is about the same"?
ReplyDeleteCan you forget for a moment that I don't seem to understand you? Ridiculous, of course, but just for the sake of reasoning?
You seem to be troubled by the expression rote learning. But you do you use spaced repetition, right? And spaced repetition is about committing something to memory. I call that rote learning. But maybe we should call it repetition? That should read:
Recognition is about the similar, Repetition is about the same.
Any idea what I mean by that?
ReplyDeleteI thought about it. I start to believe, that rote learning in chess puzzles does not exist.
I have a prove for this:
Aoxomoxoa developed a board memorization tool. You look at the board for a few seconds. Then one piece gets taken away. Which piece is it?
You would not have a chance if you dont try to quickly make some sense of the structure of the board position.
Rote learning would mean we could recall all the pieces by just looking at them without trying to make some sense (understanding) of the position.
Randomly shuffeled boards with even illegal positions are very difficult to remember. With rote learning it would not make any difference if the board position makes sense or not.
So: we dont rote learn any of the puzzles at all. None of them.
Our mind is efficient, though. If we first had 3 key thoughts to solve a puzzle - 2 key thought for the main variation, and 1 key thought for a side variation -
we only keep the 2 key thoughts to find the main variation. It leads to "I know what the solutional move is and I remember the CT players "stupid" queen sac to avoid the mate threat. But I cant see (remember) the mate threat anymore."
That means, that we will forget some information of a puzzle we desire to learn with spaced repetition.
But we dont forget all of it. We will maintain enough understanding for the main variation, and by the sheer number of puzzles we solve we will cover (hopefully) the dropped of thoughts in some other puzzles.
In this respect it is probably more helpful to have puzzles that only contain one (new) key thought, pointing again very much to look at easy puzzles instead of hard puzzles. For a master a 1800 rated puzzle is "easy" and the (new) key thoughts he needs to remember are less in number than they are for a weaker player with less knowlege.
Improving with hard puzzles is still possible, but I guess to make up for the dropped of thoughts we need an increasingly high volume (number) of puzzles to make up (to cover) the thoughts that get lost in the side variations.
Repetition of puzzles will lead to remember the puzzle with less effort. Our brain cuts away unneccessary informations for that particular puzzle, as well as it may derive simple rules that dont even make sense in a chessability way but just makes it simpler to remember the puzzle (for instance: This is the puzzle with the tripple pawn on the a-file and the solutional move is placing s.th. on the top of this tripple pawn).
As long as it makes it simpler, our brain is likely to use this in order to memorize it. Yes, that sounds like rote learning at the end of several repetitions. However, we cant rule out, that it is easiest for the brain to simply memorize the key thought for the main variation, because that IS the most simple way for our brain.
Unless you train very hard complex puzzles...
This has no relation with my last question.
ReplyDeleteyou asked several question, and those who are not rethoric questions are answered.
ReplyDeleteYour question about a comparison to recognition and repetition and rote learning assumes very much the idea, that there is s.th. like rote learning - which I deny.
Part if your meaning is that rote learning exists.
The whole meaning gets flawed though, if this assumption does not hold true.
So I answere your question, but probably not in a way you liked to hear it. Reminds me of the joke:
Are you temposchlucker or are you not temposchlucker? Answer me with a simple "yes" or "no"!
(or the other way round: "are we going to survive or are we going to die?" --> "yes!")
@munich,
ReplyDeleteRecognition is about the similar, Repetition is about the same.
Any idea what I mean by that?
last attempt (to satify your question):
ReplyDeleteIt should either read: "Recognitioin is about the same, recognition is about the same."
OR
"Repetition is about the same, repetition is about the similar"
(For a more thorough analysis to really put things right see my lenghty post above.)
@Munich,
ReplyDeleteIf I translate that to an answer that reads as "you have no idea what I mean by it".
A few days ago I said that both Aox and you, and me are like bishops of a different color. We live in a different universe.
That started me thinking about the posts of the past months. I realized that I always wondered that your comments had no relation to what I meant to say. The comments are interesting enough themselves, don't get me wrong, they simply had no relation with the essence of what I meant.
Both you and Aox are integer guys, so I realize that you made an honest attempt to understand what I mean and that you not just wanted to make my life miserable.
The same is true for me. I really want to help you to understand what I'm saying. I don't mean to offend you.
So I decided to write the essence of my posts from the past months down in the form of tiles of wisdom. And I intended to cut off any distraction in your reactions which lead away from this essence.
But now I see that there is a fundamental problem that I simply don't come across.
I see you falsifying an assumption of what you think I mean. But your assumption is wrong. That way we continue to make itchy comments over and over again until we become irritated. I don't want that to happen.
We live in a different universe and I don't know how to bridge the gap. I hope that the non commenting readers of my blog don't have the same problem. The few I speak via e-mail don't seem to suffer from it.
I think we live in the same universe, its the universe where the results counts.
ReplyDeleteI am not interested in wisdom ;) i want to improve in tactics ( and then use the method to improve in chess )
I dont care of "the colour of my bishop" if i can mate with him.
So folks: tell me you methods, then tell me your results and then i try to brew my own recepie for improvement
:)
( you see i am one of the nasty parasites.. )
At THIS moment SEEMINGLY only munichs method works for me.
MAYBE Tempos "new" method will work too, but i just cant see any indications for that right now
Give my tiles of sucess!
Aox, I did improve, no doubt about that anymore.
ReplyDeleteEven within the last 3 months I improved. If I extrapolate the last 3 mothns into a year, than I currently have an improving speed of 133 CT Blitz rating points per year.
Here my analysis about my performance/improvement in detail, it looks very promissing indeed:
http://chesstempo.com/chess-forum/training_diaries/munichs_training_diary-t2537.0.html;msg32646#msg32646
I don't find the figures convincing at all. Munich improved about 150 points in 21 months. If someone is new to tactics he will allways improve about 150 points during the first time. Regardless the method.
ReplyDeleteIt's evident we don't come any further in this discussion. Let's make an agreement that we don't talk about it anymore until I have surpassed the 2000 hallmark. Official Dutch rating or FIDE rating, that is.
If you comment, I will place it, but I will not react if it is about this specific subject.
My first 823 puzzles (under Blitz mode, I started with Standard) with the end date ...13.02.2011 had an average CT Blitz rating of 1738.
ReplyDeleteThese I am willing not to count but to assume that any improvement here might have been caused by getting used to Blitz mode.
But dont you underestimate that, tempo! You did so far just 726 puzzles under Blitz mode, so please do another 100 and then we can talk that you have a start rating.
From 13.02.2011-13.03.2011 I did another 1423 puzzles with an average rating of 1813.
It includes my All-time-high of 1857, which was reached with an RD of close to 70. In fact, only a few reading had been above 1850. (But it is enough to consider it as an active rating).
Up to the date of 20.04.2011 I did another 648 puzzles. These 648 puzzles had an average of 1797.5.
That is 15.5 points less than the average before, probably due to the duplicate reward reduction setting allready in at that volume.
I started my training based on deliberate practice, tag-themed, easy ranged, aiming at solving times below 20 seconds on the 21.04.2011, and since then I did not change much anymore. Before this I trained differently.
Till this very date I did 2798 puzzles. You are missing 2000 puzzles until you reached this number.
After 2000-3000 puzzles, most matured adults dont improve any more.
But I did. About 100 CT Ratig points from average to average with my new training. Within 16 months.
I hardly know of anybody, who is matured above the age of 30 who achieved this. But if you find somebody, I am curious about his training. I found somebody, though. As it turned out, he had a similar training to me.
My chesscube rating rose due to my training about 150 points.
In the club I am stronger, too. I started to be able to compete against the other experts in Blitz games, but have no experience in OTB games. But anyway, I was not used to win against expert players before, so this is an indication, that my improvement in CT tactics translates into real chessability.
I am curious how you will think about any CT blitz rating improvement after you did 2000 puzzles. Then we talk again...