This framework of understanding is very detailed and
specific -- which is good if it's actually correct. But it is not
straightforward to verify if it's correct or not -- or which parts are
correct how much.
So here I'm going to discuss why it's tricky, what's
the real basis for me believing it, which leads to the point that you'd
really have to know a certain area of basic physics very well in order
to verify it's correctness. But for those without that physics, there
are some approaches for "confirming" parts of it, without getting to
It's tricky to verify that the whole or a part of this
understanding is correct or not, for several reasons.
because it's about what sequence of moves
physically and biomechanically possible to employ to add
forward-propulsion power in skating, not what any particular skater
does in a certain situation.
because the framework is so complicated that it's
difficult to analyze or simulate.
because it's difficult for most of us non-elite skaters
to coordinate performing so many moves when we try them.
because for us non-elite skaters, some of our muscles don't
have the speed / strength / endurance to add noticeable power even if we
perform some of the moves correctly.
because some of the moves are difficult for us to observe
accurately or reliably when elite racers perform them.
basis in physics + biomechanics
The basis for the whole framework is that:
(a) the physics of how the ski transmits power to the ground for forward propulsion is understood, and that physics imposes contraints on what
kinds of forces and motions can possibly be effective for propulsion.
e.g. A force whose direction is parallel to the aiming
the ski cannot add significantly to propulsive power (unless there's a lot of
resistance in gliding on the ground, which is not a situation in consideration
e.g. A force whose direction against the ground does
not include a backward-directed component cannot add to current
forward-propulsion work (though it might cleverly add to propulsion work in a
later phase of the skating stroke-cycle).
e.g. A sideways-directed component of force to
add propulsion work must be toward the same side as the ski is currently
aimed toward (off the skier's overall forward motion direction).
(b) the biomechanical geometry of the joints and bones of the
human body allows some kinds of motions and not others, and can apply some kinds
of forces and torques and not others. Sometimes a motion or force/torque is
effective only from some starting configurations and not from others. For a given
human-body-configuration, we could make a list of all the possible
(c) some moves have bad side-effects.
e.g. a force which (in addition to a useful
force component) also includes a large component parallel to the aim of
the ski might throw the skier way out of balance forward or backward, or
at least leave the skier-relative-to-ski configuration in a state not
effective for transmitting future forces.
e.g. there might be some sharp quick moves
with the neck and head which could add propulsion power, but we think
the effects on perception and balance are not worth it.
(d) If we put together the force-transmission constraints of
skating physics in (a) with the geometrical constraints of the forces/torques and
starting-configurations of human biomechanics of (b), and remove
moves/configurations with bad side effects in (c), then we obtain a shorter list of
force/torque moves which are both possible from a given body-configuration, and
add significant positive forward-propulsion work to skating.
The moves on this shorter list will be different for
different body-configurations -- both the configuration of joints and bones and
muscles relative to each other -- and their configuration relative to the ski in
contact with the ground. The starting velocity/momentum of various body parts in
some configuration might also be relevant to which moves are effective for
forward-propulsion, or for the amount of forward-propulsion work each move can
add. All this is fully governed and understood by the laws and methods of
(e) overall sequence and timing is important, because prior
starting-configuration and momentum is important for each move. So which moves
come before and after a given move is important. And to have a repeatable
stroke-cycle, there must also be some "recovery" moves to get some of the joints
and bones back again (and again) into a similar starting-configuration for each
move. Some of these "recovery" moves might not add to forward-propulsion power,
and some might even reduce it, so it's important to choose moves and a sequence
which produce an overall net positive power over the entire repeatable
what claims are made
The claim here is that the listed set of moves done in
the given sequence of phases is the set that complies with the
constraints of (d) and (e) -- the full set of the effective moves
available for a skater to select for effective forward-propulsion
In more detail, the claims are:
if a skater executes any one these moves in its
correct phase in the sequence, from an effective starting
configuration (which might require some other preparatory move in a
previous phase), and has developed sufficient specific-muscle
strength-speed-endurance to execute the move with significant power
-- then the skater's overall forward speed will be increased (other
things being equal).
if a move is not in this set, and a skater tried to
execute the move, then it will not repeatedly or sustainably add to
overall motion speed, or it will have significant undesirable side
effects on balance, perception, or on other propulsive moves.
the overall phase sequence is more effective for
forward-propulsion speed for healthy human skaters than other
possible overall sequences. If some additional move is discovered,
then it can be incorporated into this phase sequence with at most
Since the moves and sequence are about what's
physically and biomechanically possible, it's hard to imagine a
way to verify them without a very sound working understanding of
Newtonian mechanics like is typically taught in a first
American-college-level physics course. (But my suspicion is that many
American students do not emerge from such a course with a "very sound
working understanding", so simply finding somebody who did well in the
course is not sufficient.)
Little or no physics is needed to falsify or contradict
Find a new move which repeatably increases the
sustainable speed and your buddies (with careful measurement), and
which is not on the list.
Find a new overall sequence which produces higher
Falsification has actually been done by elite
racers on inline skates -- because they effectively use at least
one additional propulsive move, and use one or two additional phases not
included in the sequence given here.
If you find some new moves that work, I'd love
to hear about them.
If you find a whole new overall sequence which
is substantally different, that would be wonderful. I'd love to try it
out (even if it's not actually faster or otherwise more effective).
Some ways to support or confirm the value of this
framework, without getting to the level of proof:
feel it personally: For some of the moves,
when I do them in the right sequence and configuration, I find that
I can simply feel the force, the additional force being
applied between my body and the ground.
The main problem here is that some of the
individual moves only work after there's a basic foundation of sound
skating in place. Also that some moves require a high degree of
neuromuscular coordination which might take weeks of careful
(well-coached?) practice to achieve.
The main problem with this is that many of the
key moves are subtle and difficult to detect in actual race videos with
unknown camera angles and distances -- probably need to understand a lot
about physics and the overall sequence and the camera-angle process, in
order to get reliable specific observations.
technique-demonstration videos: Observations
in controlled environments have the potential to be more accurate
and helpful than actual race videos -- with clothing marked for easy
observation, carefully-selected camera angles (and best of all
together with force-sensors in the boot-binding-ski interface).
A key test of this framework is: Can it be
made operational in a controlled lab environment with video +
position-velocity sensors + force-torque sensors? Is it straightforward
to accurately determine when and where each identified move starts and
The main problem here is that racers may skate
differently in a controlled situation than in an actual race. My special
concern is that racers in a technique-demonstration video try to follow
a simplistic theory of movement taught by their coaches that fits into
the limitations of the conscious rational mind. So we miss out on the
more complex movements that really work for winning -- which can only be
controlled by the full computational power of the unconscious
neuro-muscular control systems of the winning racers in the heat of real
My favorite approach: Try out the feelings of
many different movements -- have lots of fun.
go beyond this understanding
I think the most important ways to go beyond this
framework is to get quantitative.
The starting questions are like: How many
hundredths of second does each move (or phase) take? How does this
change in different situations with different skiers and equipment
and different performance objectives?
More interesting is: How much work is done
by each move (or phase)? How much of this work is effective for
Since physical work is force times distance,
one way to estimate this is by using force sensors and position sensors
in a controlled lab environment with a human skater. Since work can also
be defined as the difference between energy states, sometimes it can be
estimated by having ways to measure and calculate kinetic and potential
What's the proportional contribution of each
move to total work per stroke-cycle?
How does the contribution of this move
interact with other moves to reach a calculation of the average rate of
power over the whole stroke-cycle?
Which segments of each move deliver the
most or least propulsively-effective work?
How does these proportions change in different
Which moves or segments to eliminate if
trade-offs must be made?
Which moves and muscles to give priority to in
Very difficult, I expect.
back to main Leg motion