百者
Styles Philosophy Masters Training

Throwing Principles — a Programme That Trains Only the Grip-Independent Part

Kuzushi transfers across styles, nothing else does. A programme that trains exactly that part — and makes it testable through a geometric condition.

kuzushi throwing principles training doctrine judo wrestling sacripanti training plan balance
Contents

What this programme is — and its unevenly distributed basis

A programme of the training doctrine category, but the basis here is unevenly distributed, and that belongs up front:

  • The physics is solid. In 1987 Attilio Sacripanti presented a classification reducing all 77 throwing techniques of Kōdōkan judo to two Newtonian principles. That is mechanics, not training doctrine.
  • The training doctrine is not. That one should practise these principles in this order, that there are four stages, that the work begins without a throw — no study supports any of it. That is my arrangement.

A programme that transferred the solidity of the physics onto its own construction would be dishonest. For everything below the physics the same applies as in the Punching Structure Programme: origin, reasoning, falsification condition, counter-positions.

The two principles

The physical lever. The thrown body rotates around a fixed point the thrower creates and holds — hip, leg, foot or shoulder. The block is decisive: no fixed point, no lever.

The couple of forces. Two opposing forces at different heights create a torque with no pivot at all. The body rotates around its own centre of mass, because it has nothing else to rotate around. Foot sweeps, major outer reaps, every suplex.

There is no third.

And beneath them lies something simpler still

Kuzushi is nothing mystical but a geometric statement: as long as the projection of the centre of mass lies inside the area spanned by the feet, a person stands. Once it leaves that area, they fall — regardless of their strength.

That is why throwing systems are the only field in which a considerably lighter person reliably controls a considerably heavier one. And it is why this programme is possible at all: a geometric condition can be tested without anyone winning.

What this programme is allowed to train

Throwing Principles Compared separates sharply between what stays the same across all seven systems compared and what hangs on the rule set:

Kuzushi transfers across styles, nothing else does. The entry movement is tied to what the clothing offers and must be relearned when the style changes.

Subject: kuzushi · telling the two principles apart · building what each requires (for the lever the fixed point, for the couple the two heights).

Not subject: a catalogue of throws, grip fighting, groundwork, sacrifice throws, competition tactics. All of it rule- or grip-bound — and therefore the business of the system you train in.

That is little. But it is the part you take with you when the system changes.

The falsification condition

Three tests, getting harder as they go.

1 · The step. Kuzushi is geometric, therefore visible:

Clean kuzushi forces a step. If the partner stands without stepping, the projection was never outside — then they were moved, not unbalanced.

The best part of this programme, because it works without a throw.

2 · Naming the principle. After every entry: lever or couple? For the lever also: where exactly was the fixed point? Whoever cannot show it did not have one.

3 · The grip change. Change what the clothing offers — jacket, no cloth, belt, trouser grip — and do the same again. What collapses was tied to the cloth; what remains was kuzushi.

How you notice it is not holding: no forced step · the principle cannot be named, or the named fixed point does not exist · nothing survives the grip change.

What the test cannot do: it says nothing about whether this works against someone who throws back.

The problem this programme inherits

At the makiwara a device answers. Here a human answers — and humans are polite.

Kuzushi that works only because the partner goes along is not tested. It is demonstrated.

And it gets worse: the partner knows what is supposed to happen. They step because they have learned that a step comes here — not because their centre of mass was outside. So the best test in this programme is also its most fragile.

Three countermeasures, taken from the Kata-Bunkai Programme:

  1. The partner gets an instruction, not an expectation: stay standing as long as you can. Not let yourself be thrown and not resist.
  2. The principle is announced beforehand, not afterwards. Whoever decides after the entry what to call it is always right.
  3. The partner is recorded. Kuzushi that works only on one person hangs on that person’s build.

Where the prescriptions come from

PrescriptionOrigin
Two principles, no thirdSacripanti 1987 — mechanics, not training doctrine
Kuzushi as centre-of-mass projectionthe same; in the stock under throwing-principles-comparison
Limitation to the grip-independent partthe comparison finding: kuzushi transfers, the entry does not
Breakfalls as an entry conditionthe comparison: where the thrower stays standing, the thrown lands alone
Osoto-gari last, and only with the thrower holdingFinding 4 of the breakfall programme — the throw with the highest head accelerations
The step as proof of kuzushimy construction — it follows necessarily from the geometry, but appears nowhere as a test procedure
Announcing the principle before the entrymy construction, taken from the kata-bunkai programme
The grip-change testmy construction — follows from the finding about changing styles

Stage 0 · The entry condition

Whoever is thrown must be able to fall. Not the thrower — the one falling.

In the other programmes of this series safety is one rule among several. Here it is the condition of entry, because the mistake costs the partner, not the practitioner.

Proof: Phase 2 of the Breakfall Programme completed — ushiro, yoko and mae ukemi from standing, both sides, without head contact. Whoever lacks that still trains Stage 1: it contains no throw.

The four stages

Stage 1 · Kuzushi without a throw

Prescription: 3 × 10 attempts per direction (forward, back, sideways), standing, with a partner. The goal is not a throw but the step.

Why: kuzushi is almost everywhere practised as part of the throw and therefore never tested on its own. If the throw works, the kuzushi counts as successful — even when strength and entry speed in fact replaced the missing break of balance. Separating them shows for the first time how often you throw without kuzushi.

Three directions, because the base of support of a standing person is a different size forward and back (foot length) than sideways (stance width). Geometry, not opinion.

Exit: 7 of 10 forced steps per direction, both sides, three consecutive sessions — and with two different partners.

Stage 2 · Naming the principle

Prescription: 3 × 6 entries per side. Before each one, announce: lever or couple; for the lever also the fixed point.

Why: the two principles demand two entirely different fault-searches. For the lever you ask whether the fixed point held; for the couple, whether the forces really acted at different heights. Whoever does not know which search they are running searches wrongly.

Exit: in 5 of 6 entries announcement and execution match, tested separately for both principles, both sides.

Stage 3 · Resistance

Prescription: the same entries across four gradations. Whoever fails one does not go further.

LevelWhat the partner does
astands normally and stays up as long as they can
bstands more stably: wider stance, lower centre of mass
cmoves freely, without attacking
dgrips back and attempts their own kuzushi

Level b is the real test and is skipped most often because it is unspectacular. A wider base of support demands a larger displacement — and that is where every kuzushi fails that in truth rested on surprise.

Expected, and not a failure: a considerable share of what worked at a drops out at b. That is the level’s yield.

Exit: a result up to level c in the log for both principle families; level d attempted and recorded, including a negative result.

Stage 4 · Maintaining, and the grip change

Prescription: the volume reached, permanently. Plus the grip-change test at longer intervals.

Why it stops here: kuzushi does not get bigger, it gets more reliable — and reliability shows in the range of conditions, not in the number of repetitions. The continuation is therefore not more of it but the same under a different grip.

There is deliberately no Stage 5.

Counter-positions

Kuzushi cannot be separated from the throw. The strongest objection, from the practice of almost every throwing system: the break of balance arises within the entry, and whoever practises it in isolation practises something else. The counter-argument is that the separation enables a test that otherwise does not exist — not that it is the better way to learn. Whoever accepts the objection uses Stage 1 as a diagnosis rather than as training.

Sacripanti’s classification is an ordering, not a curriculum. Entirely true. That one should learn the distinction is my claim — the great majority of good throwers never learned it.

Without a system this is useless. Justified. This programme has no catalogue of throws and cannot have one, because that hangs on the grip. It supplements training in a system; it does not replace it.

The partner distorts everything. So strong that a section above is given to it. The countermeasures reduce the problem; they do not solve it.

The leg-grab ban makes the principle theory pointless. It shifted the rule set towards the lever — but the physics does not shift with it. The couple of forces stays what it is, even when a rule set blocks its most convenient entrance. The IJF began loosening the ban in 2025.

Safety — without exception

This is the only programme in the series with a second person at risk. Every rule below protects the partner, not the practitioner.

  • Stage 0 is not a recommendation. Without ukemi from standing on both sides, nothing is thrown.
  • Osoto-gari last, and only with the thrower holding. It has the highest head accelerations among the throws studied — the only point here where a measured figure grounds a prescription.
  • The thrower holds on. Letting the partner land alone means the throw was abandoned, not executed.
  • No throw against unannounced resistance. Levels c and d are agreed beforehand. Surprise resistance produces exactly the landings nobody is prepared for.
  • Neck complaints in the partner the next day: reduce volume, do not tell them off. Contracting on impact is a protective reflex and cannot be corrected by instruction.
  • Whoever throws is responsible for the landing — even when the partner “can already fall”.

What this programme cannot do

  • It teaches not a single throw. It tests the condition under which throws work.
  • It does not replace training in a system. Grip fighting, entries, the transition to the ground and tactics do not appear here and cannot.
  • Its training doctrine is untested. That separating kuzushi from the throw produces better throws is plausible and unproven. The physics behind it is not — that is the distinction that matters here.
  • It cannot abolish the polite partner. Only make them less comfortable.

A longer version with plan sheets, the log and the grip-change record is in preparation.

Author: Redaktion ·August 2026
← All Training & Techniken