Training Ideas - repost
bobstra
I've decided to repost this on RMA rather than send it to you
individually. I hope that's ok by you.
To everyone else
This is a repost of something I wrote for mma.com. It disappeared (along
with everthing else) after I dared ask a few questions. Whatever.
I'm re-posting this for because -
(1) I wrote it to be read, not to gather dust on my HDD
(2) It's something new to discuss
(3) I'd like some feedback and ideas to improve upon the stuff
contained therein.
Cheers
bobstra
bobstra
---------------------------
DISCLAIMER NUMBER THE FIRST
I'm sure most of you won't remember, but last year I promised to share
some of my ideas in regards to training from a kinesiology point of
view. The following are some ideas that popped into my head while on
holiday (what can I say - a man can only fish for so long :-). They are
not the final word on performance enhancement, just some ideas!! And
BTW I'm just a guy sharing some stuff not a guru - everything is open to
debate and question. As the ideas came to me chaotically, you'll forgive
the stream of consciousness style I've adopted. (I tried to use headings
to clean things up)
---------------------------------------------------------------------------
There are three corollaries to the below -
(1) "If you know what you're doing, you can do what you want" - Moshe
Feldenkrais (2) Conversely, if you don't know what you're doing, you
can't do what you want!! (3) If in
doubt, look it up. I've made no deliberate attempt to trick you, but
you may doubt the veracity of what I've written. So look it up. There's
over 100yrs of research into motor skills development. Use the Net to
leach it!! Try a google search on "Motor Skills".
PATTERNS, ENGRAMS
A movement pattern is a term often thrown about by ppl trying to appear
knowledgeable. Ever considered what a "pattern" really is? Literally
that. It represent the wiring of
specific neurones into a particular shape (pattern or engram). This
wiring, when activated, brings about a particular movement.
To change one part of the pattern results in the change of the whole. It
is difficult (some argue impossible) to change one aspect of a pattern
without affecting everything else in the chain. So, if we have the
pattern A,B,C,D and we focus on a drill to change element D, a new
pattern / wiring emerges (A1, B1, C1, D1). Unless one is specifically
trying to bring about a change with the drill, the pattern may be
undesirable!! Know what the drill is for!!
FATIGUE
Learning a new method of movement involves three steps - Cognitive,
associative, automatic. At the cognitive stage, things need to be
thought about. At the associative stage things begin to fall together.
At the automatic stage... well, they're automatic.
From cognitive to automatic, the amount of conscious thought generally
decreases. However, conscious thought when fatigued (even if the learner
has automatic mastery of the material) tends to increases performance.
The argument being that conscious thought, while slowing movement down
retains the correct motor skills. Thus if you only train a
narrow band of "correct" moves, then irradiation (unnecessary muscle
contraction) at subsequent fatigue level decreases because you only have
1 pattern to choose from. In
other words, the signal:noise ratio is in your favour.
ATTRIBUTES - WHAT & WHEN
When in the first stage (Cognitive) the *physical* attributes most
useful are generic ones (speed, stamina, strength etc). This goes part
way to explaining why non weight
trained BJJ beginners see dramatic improvement in performance when
undertaking resistance training. Not only are they adding to their
attributes (becoming "fit"), but they're priming the systems necessary
to assist in the learning of movement (cutting the groove, priming
neuromuscular efficiency, whatever) Unfortunately, this only seems to be
true in the beginning. Zucatto studied the relationship of attribute :
motor ability and found only a weak positive correlation (something like
0.2). What's more, initial test scores for subjects showed very little
correlation to end scores - everyone improved at unpredictable rates. So
there is a point of diminishing returns in regards to boosting physical
attributes to improve motor learning.
In the second stage(Associative), perception attributes (ie: reflexes)
are most important. Gotta be able to "move and see fast" so as to sew
techniques together, get the timing
right etc. In third stage (Automatic) task specific skills are most
important. Am I flexible enough in this particular motion? Can I move
fast enough for this type of manoeuvre? Am I accurate enough in
controlling xyz?
Obviously, doing the right type of training at the right time will
improve your performance. Less than optimal gain will be had if trying
to boost reflexes (Stage 2) while not having the necessary attributes
(Stage 1) when just starting out / learning. That's one reason why
beginners should always start slow.
BTW, there is some evidence to show that movement patterns *might* be
cumulative, so once you master some, others are gained a little more
quickly. (Search for keyword - General Motor Program or GMP)
CHAINS, SHAPES, BACK AND FRO
A Chain is a sequence of moves. A shape is a smaller part of a chain
(say, 2 moves). Normally what happens is that when training, a chain is
shown (eg: scissors sweep) then
it's drilled from start to end. But when you're first learning (Stage 1
- cognitive) the most important thing is to analyse, understand what is
going on then memorise. For you to
learn things with the greatest of ease, follow that sequence (analyse,
understand, memorise). Unfortunately, what is all too common is that ppl
just repeat a chain over and over, hoping it'll stick in their memory so
that they can move to Stage 2. This is rote learning folks and it is no
more efficient here than when trying to memorise the state capitals or
the periodic table. Wasted time and misuse of a brilliant tool (eg: your
brain).
Still, some slightly smarter folks decided to break the chain into
smaller shapes, stack them together and hope for the best (forward
shaping). Problem is, humans have a
goldfish like short term memory (actually, pre-short term or "scratch
pad" memory). The famous 7 +- 2 rule is true in motor skills as it is in
psychology. So the problem
becomes that by the time your reach shape 7, you've forgotten 1, 8
forget 2 etc. In essence forward shaping of pattern A,B,C,D = learn 1st
step (A) + 2nd step (B) + 3rd step... wait, what was 1 again?
Still some even smarter folks cottoned on to that. What, they argued,
would happen if instead going from the start to the end we did it the
other way around? And what if we
linked each back step to the one before it, leading all the way to the
end? So for pattern A,B,C,D where D is the terminal point we do the
following = 1st step (D). 2nd step =
(C)+(D). 3rd step = (B)+(C)+(D). Please note, that the shaping should
be kept short - no more than 6-8 moves, lest you incur the same
interference problem as above.
In essence what backwards shaping does is to reduce forgetfulness,
frustration and errors while increasing learner motivation. It does this
by giving a context (eg: "This is step 3 of the scissors sweep, when
opponent is doing xyz" VS "this is step 3").
It's a little like being a kid and taking a toaster apart - you just
gotta know find out how they made that shiny toaster!! (fortunately, I
always put the toaster back together)!
So, forward shaping = good. Backward shaping= better.
BUT. Some even smarter folks (those bastards are always around) said
"Look, it's Stage 1, right. Cognitive. Shouldn't yer friggin brain come
into it someplace?"
See, even with backwards shaping, you're still learning by rote. So
some folks decided to steal an old memory enhancement trick (pegging) to
make things even faster. I won't
bore you by detailing The Peg system (see end). Suffice to say you
choose a word, link an action to it and join them into a kind of a
story. Thus scissors sweep might become
"hip- out- hands-over-screw, pull" put together in the form of a story.
To those who know the peg system know that once it's learnt in one
direction, it's immediately available in the other direction too. (BTW,
pegs can be greatly enhanced by Buzan's
"Major Method" of phonetic sound=number system. Try it!! You'll never
forget anything in the chain in and out of sequence).
Finally, the mental giants of the game came along and gently tapped
everyone on the shoulder. "'Scuse me", said they "But Stage 1 =
*UNDERSTAND* + *ANYLISE* and
only them memorise. Where's the understanding and analysis huh?" These
were the guys (like Feldenkrais) that revolutionised motor control
learning. I'm not going to
bastardise what they've contributed - see the end for some recommended
reading.
So in summary. See Spot. Analyse Spot. Peg Spot. See Spot run
backwards. Good boy, spot ;-) (woof)
TRANSFER OF SKILL AND DRILLING
The way most coaches teach, a drill has very little to do with the
skill you're trying to master. There is ample evidence of this in the
swimming world - lots of paddle drills,
flipper drills, land drills. Yet, no one's performance seems to improve
much between Olympics.
If you want the skills you learn from drilling to be of value, then : -
(1) Make sure the context and content of the drill is similar to the
task
(2) Make sure you UNDERSTAND shared components of drill & task
(3) Organise the drill so the components are kept clear. Reduce the
non-vital aspects of the drill.
(4) Represent the components as you would use them in Real Life -
remember content and context!!
Now, the interesting stuff
(1) Variation within parameters can increase performance. The theory
being that by playing around with the components a "survival of the
fittest pattern" occurs. That is, the most suitable movement elements
are chosen for a general motor program.
(2) As a caveat to the above. If the motor skill is to be performed
*exactly* as it is to be drilled (arguably a kata), then disregard
variation. If the drill is to be performed in
varying, unpredictable situations, then vary the parameters (eg:
environment, opponent etc).
(3) In regards to (1) and (2). This "playing around of components"
should probably occur when you've reached Stage 3 (automatic) as
contextual interference slows down
learning, but increases transfer (drill:skill)
(4) Feedback is *vitally* important in Stage 1, important in Stage 2
and somewhat important in Stage 3 (feedFORWARD comes to the fore here).
Don't think you need
feedback? Uh-huh. Here's a fun little experiment. Get a partner, a
blind fold, a ruler, pen and paper.
(i) Draw a bulls eye on the paper. Place paper on table.
(ii) Put on blind fold.
(iii) Get partner to spin you in a circle 3 times while blind folded,
then lead you back to your chair. Pick up pen.
TRIAL 1 - While blind folded and dizzy, try to place pen on bulls-eye.
You partner is only allowed to say "Yes" or "No", then measure how far
you are from the bulls-eye and
then writes this down. Try to hit it within 15 attempts.
TRIAL 2 - after repeating step (i)-(iii), you again try to hit the
bulls-eye. This time, you partner will say "No - you missed by X cm" or
"Yes, congratulations"
TRIAL 3 - again repeat (i)-(iii). Again with the bulls-eye. This time,
if you miss, your partner can say "No - you missed by X cm. You've
landed at 3 o'clock"
TRIAL 4 - as TRIAL 3, except after telling you by how much you've missed
and where you've landed, you partner places your hand on your mark, then
the bulls-eye the on your mark again.
Now If you look at how much you've missed by and how close you were to
the bulls-eye, which trial do you think will most quickly boost your
performance? Graph it - the
results are DRAMATICALLY improved when feedback improves. Feedback not
only reduces the magnitude of error but increases the acquisition of
skill!! It hold true for this silly little lab just as well as for
improving your grappling game. So use feedback.
RECOMMENDED READING
Kinesiology 101 - the basics
http://www.google.com/search?q=cache:8u2FeA6QFmEC:www.fap.com.au/joel/site/new/hscfiles/php-perform-matt.pdf+spaced%2Bgrouped%2Bskill%2Brepetition&hl=en
You want more details on performance enhancement (general) -
http://www-rohan.sdsu.edu/dept/coachsci/mastable.htm
More details on what I've written in particular -
http://www-rohan.sdsu.edu/dept/coachsci/vol31/table.htm
More on shapes, chains and how to apply -
http://www-rohan.sdsu.edu/dept/coachsci/vol31/wrest1.htm
Some Feldenkrais patterns to explore
http://www.nas.com/~richf/pastlow.htm
Feldenkrais in General
http://www.healthy.net/feldenkrais/
Bookwise, try -
Motor Learning Concepts and Applications Magill
Virtually anything by Human Kinetics (online at www.humankinetics.com)
Supertraining by Dr Mel Siff
Anything written by Moshe Feldenkrais or FM Alexander.
"The Feldenkrais Method - teaching by handling" by Rywerant. Rare as
hen's teeth, but gives a very good scientific insight in Feldie work via
the model/s of Cybernetics and Information Processing.
Job's Body.
A good anatomy book to understand some of the above (anatomica,
gibrowbski, martini). A copy of the 1918 edition Gray's Anatomy can be
seen at www.bartleby.com/107/ Not exactly up to date, but a good
foundation if you can't find anything else.
Good stuff on brain and learning (general)
http://www.supermemo.com/english/links.htm
Specifically on Pegging and other memory systems
http://members.aol.com/beagenius/membasics.html
Books:
Anything by Tony Buzan is a good start. Try a good Psych book cause
it'll have stuff on memory+motor learning. I like Hilgards's
Introduction to Psychology.
And for fun, watch the movie Momento to learn about memory. It'll mess
with ya!!
DISCLAIMER NUMBER THE SECOND
I'm sure there are many grammatical and syntax errors in this. If that
decreases the value of the above for you...then, I guess it decreases
the value of the above for you. I
wanted to post this before I realised I hated, edited it and ended up
with something entirely different. Besides which, it's 4am.
Thankyou.
Discuss.
ADDENDUM
TRAINING FREQUENCY
A common misconception is that if some is good then more must be
better. Thus, we get to the ridiculous extreme of ppl repeating a
movement, over and over, 2,000 x a day,
hoping to "get it down". (Don't laugh, some athletes actually do this).
This kind of thinking is based upon the rote memory approach to
learning.
Yes, it works.
Also, it's kinda dumb. Here's 3 reasons why (1) It's boring. No,
really. (2) It's time consuming (3) It's ineffective.
I'm sure everyone can figure out why it's boring, as well as being time
consuming. As for ineffective... scroll back to the first few posts of
this message. The pattern of learning
- Cognitive, Associative, Automatic. Furthermore, Cognitive = See,
analyse, understand THEN memorise. 2,000 repetitions a day is an attempt
to skip the first 3 and jump
straight into memorising.
Another reason why the standard approach is ineffective is due to the
nature of the materials, that is to say the nervous system. The nervous
system has particular properties,
particular speeds. One odd thing about the way the body learns (that
is, the NERVOUS SYSTEM including YOUR BRAIN) is that it needs a bit of
time to sift things through.
Think of things this way. Your whole nervous system is a laptop
computer running Windows. Every now and again, when the data stack gets
too high, you see an hourglass
symbol. Meaning data is currently being sorted and would you please
mind not doing anything else for just a tick.
Now, this laptop has a screwy hard-drive (let's say that the tracks and
sectors are made from wax). If you try to write too much data all at
once, things get messed up. Meaning
there's a relative speed limit as to how fast things get written.
Unfortunately, wax tends to denature after a while, so the data needs a
periodic re-write. If you re-write too
often, then the wax just gets worn into a groove and the data isn't
available. If you don't rewrite often enough, the tracks erode over time
and disappear.
The question now becomes, how often do I need to re-write things?
Now we get into a grey area..
Very clearly, spaced re-writes are better than massed re-writes /
repetitions.
Why? For one, massed re-writes are likely to encourage rote learning
Spaced re-writes however, tend to promote *exactly* those process
necessary in Stage 1 Cognitive (see,
analyse, understand, memorise). What's more spaced repetitions are
easier to track (ala feedback) and are much, much less time consuming
than the "2000 reps a day"
approach.
I gonna depart from the conventional wisdom and recommend that you try
the supermemo algorithm. (infact, download the program it's free). On
paper, it looks a little like this (quoting from SM)
I(1)=4 I(n)=I(n-1)*1.7 where: I(n) - interval after the n-th repetition
expressed in days. (if you don't understand mathematics then look
further) Fractional intervals should be
rounded to the nearest integer.
The formula indicates that the following intervals should have the
following lengths: 4, 7, 12, 20 days, then 1, 2, 3, 5, 9, 16 months,
then 2, 4, 6, 11, 18 years, etc.
So, if you memorize a page on Jan 1, 1990 (the first repetition), the
second repetition should follow in 4 days (on Jan 5). The third
repetition should follow the second by one
week (Jan 12). The subsequent repetitions should be in: 12 days (Jan
24), 20 days (Feb 13) 1 month (March 13), 2 months (May 13), 3 months
(Aug 13), through 1990..., Jan
13, 1991, Oct 13, 1991, Feb 13, 1992, etc.
Bear in mind that for this to work, you must follow the "see, analyse,
understand THEN memorise" pattern. Why? Because practice makes
PERMANENT, not perfect. Don't
rush through the analyse etc stages. Learning something may take more
than one day. This is where incremental learning comes in. (you can
apply the ideas from http://www.supermemo.com/help/read.htm)
Also, no one says that you can't learn more than 1 thing a day, or that
you can't learn new info as you are reviewing older info.
I strongly suggest you scope out the supermemo.com site for the
specifics. In fairness, here's a link explaining the standard approach.
Try both. SM works for me., YMMV
http://www.google.com/search?q=cache:TCfkRV7iw5UC:www.unlv.edu/Colleges/Healt
h_Sciences/Kinesiology/classes/KIN750/meta.pdf+massed+vs+spaced+repetition+kinesi
ology&hl=en
A final note. When my repetition date comes around, I will perform the
necessary
"trick" for x number of repetitions (lets say, 10 a side). That's
probably just me being
paranoid. Furthermore, intermittent repetitions inevitably occur during
class ( I manage to pull off a scissors sweep every class). Don't be
worried that you aren't getting enough
practice and feel free to practice a little in between.
OK. That's pretty much it guys. I'd really appreciate some feedback on
this plus any other constructive comments. I know for a fact there men
and women on this forum who
know a great deal more than I on this topic.
Many thanks
bobstra
Batman
<3C729312...@operamail.com>:
>This is a repost of something I wrote for mma.com. It disappeared (along
>with everthing else) after I dared ask a few questions. Whatever.
Except, theres still the tiny little problem of ... mma.com is media marketing
assessment. If you tried to post training ideas up there, I'm really not
surprised that they canceled the post. It would be slightly off topic.
-jeff (FSG)
bobstra
Batman wrote:
>
> bobstra <bob_...@operamail.com> wrote in
> <3C729312...@operamail.com>:
>
> Except, theres still the tiny little problem of ... mma.com is media >marketing assessment. If you tried to post training ideas up there, I'm >really not
> surprised that they canceled the post. It would be slightly off topic.
>
Heh, I suspect that that's not too far from the truth for some of the
forums. However, the S&C q&a seemed to be good, as did the bjj and
wrestling ones.
Regardless, for the reasons you've hinted at (and also the pettiness of
a certain individual), I think I'm coming back to rma.