Latest Updates:
Page Index Toggle Pages: 1 2 3 [4] 
Topic Tools
Very Hot Topic (More than 25 Replies) Smith and Tikkanen's Woodpecker Method (Read 16698 times)
Stigma
God Member
*****
Online


There is a crack in everything.

Posts: 3032
Joined: 11/07/06
Gender: Male
Re: Smith and Tikkanen's Woodpecker Method
Reply #6 - 07/20/18 at 15:23:34
Post Tools
barnaby wrote on 07/20/18 at 14:29:11:
I just went up 120 rating points last 12 months and did nothing other than play in events. 

All these claims of ratings gains due to books and working with specific books in certain ways  is absolute marketing nonsense.

Congratulations, that's impressive progress!

The conclusion doesn't follow at all though. It's almost a truism that you did something better during those 12 months than previously. If it wasn't studying anything in particular, it could be more interest, better focus, better time management, better sleep, better health, more time to prepare before rounds, fewer other things on your mind... anything.

I can attribute real progress to specific books i studied, and mostly due to pattern recognition after studying well-chosen examples repeatedly, not due to techniques like the "Silman thinking technique" and whatnot.

Authors/publishers like Silman and Aaggard readily admit that there are many ways to improve in chess, and their books are just good options among many other possibilities. But don't discount the value of following a specific book or program for the automatic motivation and structure that provides. I've learned most from chess books when I've committed to going through them cover-to-cover even though not all the material was equally useful for me at the time.
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
barnaby
Senior Member
****
Offline


The night is dark and
full of terrors.

Posts: 345
Joined: 01/09/12
Gender: Female
Re: Smith and Tikkanen's Woodpecker Method
Reply #5 - 07/20/18 at 14:29:11
Post Tools
I just went up 120 rating points last 12 months and did nothing other than play in events. 

All these claims of ratings gains due to books and working with specific books in certain ways  is absolute marketing nonsense.
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1042
Joined: 05/17/10
Gender: Male
Re: Smith and Tikkanen's Woodpecker Method
Reply #4 - 07/20/18 at 00:25:09
Post Tools
Some great points there, particularly about not having to calculate in the successful reviews. Spaced repetition, though, is much more efficient than the Woodpecker method as described, particularly in that using spaced repetition you are predominantly redoing problems that you got wrong and are (arguably) reviewing mainly to identify those. With that change of emphasis,  using the same problems makes more sense than otherwise--even math teachers make you redo problems that you got wrong. Indeed, if you get a tactics review position wrong, a fortiori you would likely get a merely related position wrong, for you likely don't remember the mechanism involved.

But there really is no evidence on Woodpecker vs Stappen or the like and it would require a prospective study. Surely the Woodpecker will help; I just wonder about Woodpecker's "value over replacement" in the field of tactics study.
« Last Edit: 07/20/18 at 17:29:21 by ReneDescartes »  
Back to top
 
IP Logged
 
Stigma
God Member
*****
Online


There is a crack in everything.

Posts: 3032
Joined: 11/07/06
Gender: Male
Re: Smith and Tikkanen's Woodpecker Method
Reply #3 - 07/19/18 at 20:37:20
Post Tools
ReneDescartes wrote on 07/19/18 at 15:25:17:
The chess idea is that drilling exactly the same positions imprints "patterns" on the mind and improves "pattern recognition," but the more closely you examine the latter concept, the blurrier it gets! A pin with a White Bb5 against a Black Nc6 and Ke8, with d7 unoccupied, is a pattern, but in my opinion the placement of a pawn on h3 or h2 is not part of it. So what is the actual pattern, or the part of it that is to be "recognized," since you will not see in your own game the position that it is so important to drill over and over, but only a similar one? And why should we think that recognizing identical positions rather than similar positions in drill is the best training for recognizing similar positions but not identical positions in play?

I don't see the blur here. The actual chess pattern isn't the entire position, but just the parts of it that are essential for the relevant idea, whether tactical or positional, to work (though of course that may be the entire position in some cases).
I haven't thought about it before, but one advantage of the positions drilled being identical is that you don't have to spend a lot of time stopping and checking whether the pattern really works - after a few runs through you already know it does. Thus you can drill the patterns into your mind more quickly.

Of course it's also important to train with positions where it's not clear whether a tempting pattern works, or which of several tempting ones. But that's a different kind of training - one we can get on the first pass through a tactics set and also during serious calculation training, but not when drilling patterns. [Edit: Actually, with problems I repeatedly get wrong, "interference" from another tempting pattern is often the culprit. But that can be useful information in itself - "Note: Study these two types of pattern more deeply in order to distinguish them better" etc. From a woodpecker perspective, such errors may also mean the interval between tries of the same problem was too long, i.e. I tried to cover too many problems in the time I had available.]

ReneDescartes wrote on 07/19/18 at 15:25:17:
I would love to see this method tested scientifically, against the same amount of time spent on (1) detailed analysis of of complicated positions and (2) the Stappenmethode, which is more like a traditional math workbook, and, say (3) slow play. Such a test has never been done to my knowledge, though Smith says his friend Hagen improved faster after switching to the same-positions method.

Maybe it's strictly correct to call the woodpecker method "untested", but I find it a bit misleading nonetheless. The woodpecker method is part of a family of learning strategies where spaced repetition is the most reputable one, and that has been tested heavily, with better results than most other strategies. I look forward to reading Smith and Tikkanen's discussion of historical predecessors to the "woodpecker", but to my mind the methods are similar enough that when one works very well, the other should at least work quite well. And if one doesn't want to use a computer program for spaced repetition, the woodpecker method is extremely easy to implement since there is very little math involved in deciding which problem to do next and how to score it (Though there are other simplified methods: For instance I've been trying to implement a simple version of the "Leitner boxes" using just a tactics book and a pencil).

Anecdotally, while I have never used the woodpecker method strictly as prescribed, there are two-and-a-half tactics books I've gone through so many times that the effect must be very similar. Ever since I started doing that, tactical pattern recognition has been a relative strength in my play.

I don't see anyone arguing that your points 1)-3) shouldn't be useful - 1) and 3) are virtually indispensable to becoming a strong player and not to be thought of as alternatives to the woodpecker method, but complimentary. But all these three do constitute the traditional ways of training chess. If someone has tried them for years and reached a plateau, it's reasonable to add or switch to repetition-based strategies.

ReneDescartes wrote on 07/19/18 at 15:25:17:
During the de la Maza craze, a number of class players tried something close to the Tikkanen method, and it produced around 200 points improvement rather than the much greater number of points that was claimed. On the other hand, 160 or 100 points by working six to ten hours a day is all that Smith claims for it.

Just to state the obvious, 200 points does not equal 200 points - it depends a lot on the starting level. The higher you get rating-wise, the harder further improvement becomes. How much did Tikkanen improve in actual playing strength during his crucial test of the method - 50 points or 100? Yet that was enough to give him three GM norms. That's probably a more difficult improvement to achieve than, say, going from 1200 to 2000, where the possible improvement paths are well-known.
« Last Edit: 07/19/18 at 21:37:28 by Stigma »  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
gillbod
Full Member
***
Offline


I Love ChessPublishing!

Posts: 150
Joined: 03/26/13
Re: Smith and Tikkanen's Woodpecker Method
Reply #2 - 07/19/18 at 19:56:48
Post Tools
ReneDescartes wrote on 07/19/18 at 15:25:17:
in most other areas of learning--in mathematics instruction, for example--we have students drill with similar, but not identical exercises where the same ideas appear in somewhat different contexts (in reciprocal form, , or in geometric form, or where a different variable in the formula is unknown, etc.).


I like your parallel to mathematical education. I am into maths, and I posit that the short end of improving at mathematics and chess is the same: incessant and active thinking.

The art is choosing problems so that you can actually maintain this. From my perspective, the following three should apply:

1) Choosing problems you find interesting (enough to fuel your enthusiasm)

2) Choosing problems which are non-trivial (forcing you to think actively)

3) Choosing problems easy enough so that you can think of more than a single way to try and solve it (to avoid the pitfall that conditions 1 and 2 can lead you to trying problems which become demoralisingly hard).

However, I venture from the original intention of the post...
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1042
Joined: 05/17/10
Gender: Male
Re: Smith and Tikkanen's Woodpecker Method
Reply #1 - 07/19/18 at 15:25:17
Post Tools
There is an untested hypothesis, with some adherents such as Smith and Heisman and (ahem) de la Maza, that rehearsing the same set of exercises over and over produces better results than drilling with exercises similar, but not identical, to each other. The hypothesis is untested, so no one knows, but I am not a fan; in most other areas of learning--in mathematics instruction, for example--we have students drill with similar, but not identical exercises where the same ideas appear in somewhat different contexts (in reciprocal form, , or in geometric form, or where a different variable in the formula is unknown, etc.).

The chess idea is that drilling exactly the same positions imprints "patterns" on the mind and improves "pattern recognition," but the more closely you examine the latter concept, the blurrier it gets! A pin with a White Bb5 against a Black Nc6 and Ke8, with d7 unoccupied, is a pattern, but in my opinion the placement of a pawn on h3 or h2 is not part of it. So what is the actual pattern, or the part of it that is to be "recognized," since you will [i]not[i] see in your own game the position that it is so important to drill over and over, but only a similar one? And why should we think that recognizing identical positions rather than similar positions in drill is the best training for recognizing similar positions but not identical positions in play?

I would love to see this method tested scientifically, against the same amount of time spent on (1) detailed analysis of of complicated positions and (2) the Stappenmethode, which is more like a traditional math workbook, and, say (3) slow play. Such a test has never been done to my knowledge, though Smith says his friend Hagen improved faster after switching to the same-positions method.

During the de la Maza craze, a number of class players tried something close to the Tikkanen method, and it produced around 200 points improvement rather than the much greater number of points that was claimed. On the other hand, 160 or 100 points by working six to ten hours a day is all that Smith claims for it.
  
Back to top
 
IP Logged
 
gillbod
Full Member
***
Offline


I Love ChessPublishing!

Posts: 150
Joined: 03/26/13
Smith and Tikkanen's Woodpecker Method
07/19/18 at 12:50:09
Post Tools
a book on 'the woodpecker method' is coming soon from quality chess, with the ebook is out already. i was hoping to have a short discussion on it.

for the benefit of those who don't know, the method is roughly repeating the same set of 1000 tactics 'til the end of time.

i am left wondering what more this book offers over a regular tactics book. perhaps someone here can enlighten me?
  
Back to top
 
IP Logged
 
Page Index Toggle Pages: 1 2 3 [4] 
Topic Tools
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo