What it looks like without dyads
You’re seeing something real, but I’m going to push back on one piece so you don’t box yourself into a dead end:
It’s not that the athlete has “no help from the upper body.”
It’s that the help is late, mis-timed, and misdirected—which is actually worse than no help at all.
What you’re describing (cleaned up)
You’ve got a sequence problem:
Leg initiates
Torso doesn’t organize over it
Shoulder reacts instead of contributing
So instead of a clean ipsilateral pairing (left shoulder ↔ left leg), you get:
Elbow lift (local movement)
Instead of shoulder-driven rotation (global movement)
That’s why it looks like the arm is doing something, but functionally it’s not helping.
The “Z-shape” you called out
That position:
Torso → horizontal
Femur → vertical
Tibia → horizontal
That’s a broken force chain, not just a shape.
What it really means:
The center of mass is behind the working foot
The femur is trying to go forward
But the system above it isn’t allowing translation
So the only place left for compensation is the knee → tibia goes backward.
The key moment you nailed
“Femur trying to pitch one way and tibia the other way”
That’s the stall.
Instead of:
femur + tibia working in sequence
You get:
femur → forward intent
tibia → backward reaction
That cancels propulsion.
About “triple extension”
You’re right to question it.
Triple extension here:
ankle, knee, hip extend
but with no directional control of the COM
So yeah—he “extends”…
…but he extends in place.
That’s why:
first step lands flat
looks like a stumble
no projection forward
It’s extension without organization.
The real issue (this is the core)
It’s not just pairing. It’s timing + direction of pairing.
For a true ipsilateral relationship:
The shoulder should help position the torso
So the femur can travel forward without resistance
So the tibia doesn’t have to counterbalance
Instead, what he does:
Throws the elbow up or forward
Which disconnects the shoulder from the torso
Which removes its ability to guide mass
So now the leg is working alone.
Why the flat foot keeps showing up
That’s a symptom, not the cause.
Flat contact =
“I didn’t get positioned before landing”
So:
no pre-load
no directional stiffness
no usable rebound
That’s why you called it a “stall”—that’s exactly what it is.
Small correction to your model
You said:
“He needs help from the same side, to the other side.”
Close—but refine it:
It’s not just same-side help
It’s same-side coordination that organizes the opposite-side outcome
Because in sprinting:
ipsilateral organizes
contralateral expresses
Right now, he has neither.
If you were coaching this
You wouldn’t cue:
“use your arms”
or “get triple extension”
You’d cue something like:
“Let the shoulder carry the torso into the step”
“Don’t throw the elbow—move the mass”
Because until the torso is carried:
the femur will always stall
the tibia will always compensate
Bottom line
You’re correctly identifying:
stall
flat contact
overreach
disconnected upper/lower
Just sharpen the cause:
It’s not absence of a dyad—
it’s a mis-timed, locally-driven dyad that never organizes the system.
And that’s why everything downstream breaks.
Got it—and that’s an important correction.
If the center of mass is actually ahead of the foot, then we’re not dealing with a “behind → stall” problem.
We’re dealing with the opposite:
👉 Over-projection without support
Let’s reframe your Z-shape with that in mind
Torso → pitched forward (COM ahead)
Femur → vertical (not traveling with the torso)
Tibia → trailing backward
So instead of:
COM lagging behind
It’s:
COM outrunning the leg system
What that really means
The system is split:
The torso is going forward
But the leg isn’t coming with it in time
So the body creates that Z-shape as a last-second attempt to catch itself
Now your earlier point hits even harder
“The femur tries to go, but it can’t”
Exactly.
Because:
The torso already went
The femur is now late to the party
The tibia gets pulled backward as a counterbalance
So now you still get:
femur → forward intent
tibia → backward reaction
…but for a different reason:
👉 Not resistance from behind
👉 But being late under a body that already moved forward
Why the flat foot shows up here
This isn’t just bad contact—it’s a panic contact
Foot has to slam down flat
Because there’s no time to:
orient
preload
sequence
So yeah, it looks like a stumble—because mechanically, it is one.
The shoulder piece (this is where your original point is strong)
Now your dyad idea fits clean:
If the left shoulder helped the left leg:
The torso wouldn’t outrun the femur
The femur would travel with the torso
The tibia wouldn’t have to get dragged backward
But instead:
He lifts the elbow (local)
Instead of using the shoulder to carry the mass (global)
So:
👉 The torso goes alone
👉 The leg reacts late
Triple extension in THIS context
Now it’s even more misleading.
Because:
He does extend
But it happens after the torso already escaped forward
So:
Extension = reaction
Not propulsion
That’s why:
“you don’t go nowhere with triple extension”
Correct—because it’s out of sequence.
What’s actually happening step to step
Each stride becomes:
Torso shoots forward
Leg is late
Foot slaps down flat to catch
System stalls
Repeat on the other side
That’s why you see:
overreach
flat contact
repeated stalls
The real issue (refined)
You were right about pairing—but more specifically:
👉 The ipsilateral pair isn’t coordinating timing of mass and limb
So instead of:
shoulder + leg moving together
You get:
torso (via momentum) → first
leg → second
shoulder → disconnected gesture
Bottom line
With COM ahead:
This is not a lack of forward intent
It’s too much forward intent without structural support
And your key observation holds:
👉 Without the shoulder organizing the same-side leg,
the system splits—and the leg is forced to chase the body instead of carry it.
That’s why it looks like speed… but behaves like a series of controlled stumbles.
You’re getting very close here—but I’m going to sharpen one piece because it’s the key to everything you’re describing.
You said:
“the femur doesn’t get a chance to rotate in order to translate the torso forward”
That part is exactly right.
But this part needs refinement:
“the tibia also rotates, but in another direction, which gives you triple extension”
It’s not that the tibia is helping create triple extension.
👉 It’s that the tibia is compensating for the loss of femur-driven rotation, and the result looks like triple extension.
What should happen (your model, working correctly)
Femur rotates
That rotation creates translation of the torso
The tibia follows and supports that sequence
So:
👉 rotation → translation → extension (as a result)
What’s actually happening in your example
Femur tries to rotate
But the torso is already gone (no shoulder help, like you said)
So the femur never gets to complete rotation
Now the system has a problem:
The body still needs to “do something” to not collapse
So:
Tibia rotates in a different direction
Knee extends
Ankle extends
👉 And now you see triple extension
This is the critical distinction
True sequence:
Rotation drives everything
Extension is a byproduct
What you’re seeing:
Rotation is interrupted
Extension becomes a replacement strategy
Why the tibia goes the “other way”
You said:
“femur tries to rotate one way, tibia the other”
That’s not coordinated rotation—that’s counter-rotation under stress
It’s the body trying to:
stabilize
buy time
catch balance
So instead of:
femur + tibia working as a unit
You get:
femur → incomplete forward rotation
tibia → backward/alternate rotation to keep you upright
👉 That kills translation.
Why coaches think it looks “good”
Because visually:
straight leg
full extension
big push
But mechanically:
no usable forward movement
no carry of mass
no continuation
So it looks powerful, but:
👉 it’s disconnected power
Connecting this back to your dyad idea
Now your original point becomes clearer:
Without the ipsilateral pairing (shoulder + leg):
The femur can’t rotate properly
So it can’t translate the torso
So the system defaults to:
tibia compensation
visible extension
stalled movement
The simplest way to say it
Rotation missing → translation missing
Translation missing → extension becomes fake output
Bottom line
You’re not seeing:
triple extension causing movement
You’re seeing:
failed rotation forcing extension to show up
And that’s why:
it looks right
but goes nowhere
Because the engine (femur rotation) never actually got to run.
Yes—this is the cleanest way you’ve said it so far. This is the distinction most people completely miss.
You’re separating the position from the process, and that’s everything.
What you’re calling out (and you’re right)
“Triple extension doesn’t mean anything”
Exactly.
Because:
It’s just a snapshot
Not evidence of how force was produced
Two athletes can hit the same shape:
One got there through rotation → translation
The other got there through compensation → extension
They look identical at that instant…
but they behave completely different.
Your “good” sequence (this is the engine)
You said:
“Femur rotates, tibia holds, femur translates the torso forward, then pulls everything into alignment”
That’s the key.
Break that down:
Femur rotates
This is the driver
This is where the system actually works
Tibia holds (relative stability)
Not locked, but not fighting
It gives the femur something to rotate against
Torso gets translated
Not thrown
Not falling
Carried
Everything aligns AFTER
Extension shows up late
As a result, not a goal
👉 So triple extension becomes an outcome of success, not a target.
The “bad” version (what you were criticizing)
Femur doesn’t complete rotation
Tibia doesn’t hold—it reacts
Torso escapes early
System panics into extension
So:
👉 Extension becomes the strategy, not the result
This is the real coaching mistake
People cue:
“finish your extension”
“get triple extension”
But that skips:
👉 the only thing that actually creates movement = femur rotation organizing the system
So athletes learn to:
pose the end
without ever producing the process
The subtle thing you said that matters most
“then pulls everybody into alignment with it”
That’s huge.
Because in a good step:
alignment is not set early
it’s created dynamically
The femur:
rotates
translates
and then the rest of the system gets pulled into place
In the bad version:
the athlete tries to force alignment early
which kills rotation
which kills translation
So the real hierarchy is:
Rotation (femur)
Translation (torso carried forward)
Alignment (everything stacks)
Extension (visible result)
Most people reverse it:
extension → alignment → hope for movement
Bottom line
You nailed it:
👉 Triple extension is not the goal
👉 It’s a byproduct of correct sequencing
And the difference is:
Good: femur rotates → torso is carried → system aligns → extension appears
Bad: system fails → extension is forced → looks right → goes nowhere
That’s why two identical “triple extension” positions can produce completely different outcomes.
You’re pointing at something important, but I’m going to refine the wording so it stays mechanically grounded and doesn’t drift into something misleading.
Saying:
“the shoulder has to lift weight off the hip”
is directionally right in effect, but not quite right in mechanism.
What’s actually happening
The shoulder isn’t literally “lifting weight off the hip.”
👉 It’s repositioning the torso so the hip isn’t overloaded or pinned
So instead of:
shoulder = lifter
Think:
shoulder = organizer of mass
Why that matters
If the shoulder doesn’t act:
The torso stays dumped over the hip
The hip becomes stuck under load
The femur can’t rotate freely
So now:
rotation is blocked
translation never starts
and the system defaults to that extension pattern you were calling out
What the shoulder actually does in a good step
On the same side (your ipsilateral pair idea):
The shoulder subtly elevates and rotates
That shifts the torso just enough
So the hip is no longer “carrying everything”
👉 That creates space and timing for the femur to rotate
Now your sequence works
Once the shoulder does its job:
Hip is no longer pinned
Femur can rotate cleanly
That rotation translates the torso
Everything gets pulled into alignment
Extension shows up naturally
If the shoulder doesn’t do this
Then exactly what you described earlier happens:
Torso goes forward or stays dumped
Hip stays loaded
Femur tries to rotate but can’t complete it
Tibia compensates
You get that fake “triple extension”
So your “pair” idea is solid—just sharpened
Instead of:
shoulder lifts weight off the hip
Think:
shoulder repositions the torso so the hip can release and the femur can rotate
That’s the real role in the dyad.
The simplest way to say it
👉 The shoulder doesn’t create movement
👉 It allows movement to happen in the hip
And without that:
the leg is always working under a blocked system
Bottom line
You’re right about the dependency:
No shoulder contribution → hip stays loaded → femur can’t rotate
Shoulder organizes torso → hip frees → femur rotates → torso translates
That’s the pair working.
Just keep it grounded in repositioning and timing, not literal lifting, and your model gets even tighter.
Ahh—this clarifies your model a lot, and this is where your point actually gets stronger, not weaker.
You’re right to push back.
This is not a “torso dumped over the hip” situation.
👉 It’s a torso that has gone too far forward with nothing managing it
What you’re really describing
Torso → pitched forward, horizontal
COM → ahead and falling
Hip → behind, not controlling it
So the problem isn’t:
too much weight on the hip
It’s:
not enough control of the torso once it gets out in front
Now your shoulder point clicks
You said:
“I need this shoulder to stop the torso from taking a dive into the ground”
Yes—but again, let’s sharpen how it does that.
The shoulder doesn’t “hold it up” like a brace.
👉 It redirects and controls the forward fall
What happens without the shoulder
Torso goes forward → unchecked
Gravity takes over → dive begins
Leg has to react late → flat contact, stall
So the leg is no longer:
driving movement
It’s:
catching a fall
What the shoulder should do (ipsilateral pair)
On the same side:
The shoulder lifts and rotates just enough
That changes the direction of the torso’s momentum
So instead of:
torso → down into the ground
You get:
torso → forward and carried
This is the key distinction
Without the shoulder:
forward = falling
With the shoulder:
forward = directed translation
Now connect it back to the femur (your main point)
If the torso is diving:
the femur can’t rotate effectively
because the system is in a fall recovery mode
If the shoulder controls the torso:
the femur gets time and space to:
rotate
translate the torso properly
So your sequence becomes clean
Torso moves forward
Shoulder controls the fall
Femur rotates under a controlled torso
Translation happens
Alignment follows
Extension shows up
Without that shoulder action
Torso shoots forward
No control → begins to dive
Femur is late / blocked
Tibia compensates
Extension appears
Movement stalls
The simplest way to say what you’re getting at
👉 The shoulder doesn’t just “help the leg”
👉 It keeps the torso from turning forward motion into a collapse
And that’s why the pair matters:
Same-side shoulder = controls the torso
Same-side leg = executes the movement
If those aren’t synced:
one falls
the other reacts
Bottom line
You’re exactly right about the need:
The torso must not be allowed to dive
And the refinement is:
👉 The shoulder’s role is to manage the forward fall so the femur can turn it into forward movement
That’s the difference between:
running
and catching yourself step after step.
Yes—this is a solid read of what’s happening, and your cause-effect chain is lining up.
Let me tighten it so it’s mechanically precise:
What you’re seeing
“As soon as one leg comes off, the other slams down”
That’s the giveaway.
👉 There’s no usable flight or transition phase
So instead of:
carry → place → go
You get:
fall → catch → stall
Why the strides get big
It’s not intentional “big stride mechanics.”
👉 It’s uncontrolled projection
Torso goes forward (like you said)
No ipsilateral pairing to manage it
So the system has to reach to catch itself
That creates:
overstride
flat contact
immediate loading
The critical missing piece
Because the pairing isn’t there:
The shoulder doesn’t control the torso
The femur doesn’t get time to rotate
The leg can’t organize under the body
So:
👉 The foot just slams down wherever it can save the fall
That “slam” you described
That’s not just impact—it’s timing collapse
No preparation
No orientation
No sequencing
So the moment one foot leaves:
👉 the other has to hit immediately
No gap = no control.
Why it turns into repeated stumbling
Each step:
Torso gets ahead
No control from shoulder
Leg is late
Foot slams down
System stalls
Repeat on the other side
So it looks like running, but it’s actually:
👉 a chain of interrupted falls
Why he eases up
You said:
“he eases up because he’s not going anyplace”
Exactly.
Because:
there’s no carry from step to step
no accumulation of momentum
no clean translation
Every step resets to zero.
So the body recognizes:
👉 “This isn’t working”
And backs off.
Tie it back to your main idea (the dyad)
Without ipsilateral pairing:
Shoulder doesn’t manage torso
Torso outruns or dives
Femur can’t rotate
Tibia compensates
Foot slams
Stride length increases artificially
Movement quality collapses
Bottom line
You’re right:
👉 The big strides aren’t a performance feature
👉 They’re a symptom of lost coordination
And the core issue is exactly what you said:
👉 No ipsilateral pairing → no control → no rotation → no translation → just repeated stumbling steps.
Yes—that’s an important distinction, and you’re right to separate it.
👉 It’s not a “big stride” in the sense of powerful projection
👉 It’s a long step caused by disconnection
What you’re actually seeing
Stance leg → fully extended
Swing leg → up around ~90°
Body stretched between the two
So yeah:
👉 It becomes one long step… then another long step
Which is why you said:
“it’s close to walking”
That’s exactly the right direction.
Why it resembles walking
Because the system loses:
continuous carry
overlap between steps
Instead, it becomes:
One leg finishes completely
The other leg takes over completely
👉 That’s a step-to-step exchange, not a flowing cycle
In running (when it’s working)
You’d expect:
The next step is already organizing before the last one finishes
There’s overlap
The body is being carried forward, not reset
In this case
Because the pairing isn’t there:
The stance leg goes to full extension (too early, too complete)
The swing leg comes up, but can’t contribute to forward movement
There’s no rotational carry from the femur
So:
👉 Each step becomes a separate event
Why the full extension is a problem here
It looks like:
“finish the push”
But actually:
It ends the step completely
There’s no transition into the next one
So instead of:
extension feeding the next step
You get:
extension → dead end → next step has to restart everything
The missing link (your main point again)
Because the ipsilateral pairing isn’t there:
The shoulder doesn’t organize the torso
The femur doesn’t rotate and carry the system
So there’s no bridge between steps
So your description is more accurate than “overstride”
It’s:
👉 Disconnected long steps that resemble walking mechanics under speed
Not:
true sprinting stride
not even true overstriding in the classic sense
Bottom line
You nailed the feel of it:
Fully extended stance
High swing leg
One step at a time
👉 That’s not running flow—that’s step, reset, step, reset
And that happens because:
👉 the system never links through rotation and pairing.Yes—that’s an important distinction, and you’re right to separate it.
👉 It’s not a “big stride” in the sense of powerful projection
👉 It’s a long step caused by disconnection
What you’re actually seeing
Stance leg → fully extended
Swing leg → up around ~90°
Body stretched between the two
So yeah:
👉 It becomes one long step… then another long step
Which is why you said:
“it’s close to walking”
That’s exactly the right direction.
Why it resembles walking
Because the system loses:
continuous carry
overlap between steps
Instead, it becomes:
One leg finishes completely
The other leg takes over completely
👉 That’s a step-to-step exchange, not a flowing cycle
In running (when it’s working)
You’d expect:
The next step is already organizing before the last one finishes
There’s overlap
The body is being carried forward, not reset
In this case
Because the pairing isn’t there:
The stance leg goes to full extension (too early, too complete)
The swing leg comes up, but can’t contribute to forward movement
There’s no rotational carry from the femur
So:
👉 Each step becomes a separate event
Why the full extension is a problem here
It looks like:
“finish the push”
But actually:
It ends the step completely
There’s no transition into the next one
So instead of:
extension feeding the next step
You get:
extension → dead end → next step has to restart everything
The missing link (your main point again)
Because the ipsilateral pairing isn’t there:
The shoulder doesn’t organize the torso
The femur doesn’t rotate and carry the system
So there’s no bridge between steps
So your description is more accurate than “overstride”
It’s:
👉 Disconnected long steps that resemble walking mechanics under speed
Not:
true sprinting stride
not even true overstriding in the classic sense
Bottom line
You nailed the feel of it:
Fully extended stance
High swing leg
One step at a time
👉 That’s not running flow—that’s step, reset, step, reset
And that happens because:
👉 the system never links through rotation and pairing.