Mechanical Residue

The objection this answers

The standard objection to a Signature Move is that it has an expiry date. If a habit is repeatable enough to become a fingerprint, it is repeatable enough to end up in an opposition analyst's clip reel, and once it does the surprise is gone. I used to concede that point, and my own writing on Signature Move said as much — that it gets suppressed once an opponent has studied it properly. Working through the movement science, I think that concession was wrong, or at least far too generous. The advantage does not go to zero when the move is read. Part of it does. The other part is mechanical, and mechanics do not care what the defender knows.

The body sends two signals at once

Brault and colleagues (2012) put rugby side-steps through motion capture and virtual reality and separated what the body broadcasts into two kinds of signal. Exaggerated movements — upper trunk yaw, the placement of the outside foot — carry deceptive information, because they are not tightly bound to where the player is actually going. Minimised movements — the displacement of the centre of mass, lower trunk yaw — carry honest information, because they cannot be faked without also changing the outcome. This is the same split Deghaies and colleagues (2025) found inside a single tennis serve, where the loading phase varies between players by around 42 percent while the acceleration phase varies by only 15 percent. Put the two together and you get the sharpest statement I know of what a Signature Move actually is: a Signature Move lives in the degrees of freedom that mechanics leaves free. The centre of mass is not one of them. Which also means a player's budget for having a signature at all is set by their body — trunk range, foot placement width, deceleration capacity — not by how creative they are.

One half decays. The other does not.

That split hands you two different components of the delay a Signature Move imposes, and they behave completely differently against better opponents. The perceptual component is the defender reading the wrong signal. Brault's experts picked up centre-of-mass information 183 milliseconds before the attacker reoriented; novices got it 16 milliseconds before — a 167 millisecond gap, and the experts won precisely by moving later rather than earlier. Jackson and colleagues (2018) found the same thing for stepovers in football: high-skilled players discriminated genuine from deceptive actions at d-prime 1.46 against 0.63 for low-skilled. So the perceptual component shrinks as the opponent gets better, and against a truly elite defender it tends toward nothing. The mechanical component is different. Once a defender has committed his centre of mass, reversing it has a price in physics. Van den Tillaar and colleagues (2020) measured it: the plant-step ground contact for a 45-degree cut is about 0.15 seconds and for a 180-degree cut about 0.30 seconds, with deceleration steps rising from roughly 0.4 to 3.3. That number does not fall because the defender watched video. I call what is left the mechanical residue — the part of a Signature Move that survives being read — and its size is set by the speed, angle and footing of the moment, not by the opposition's analysis department.

Why that changes how you coach it

Three things follow. First, a Signature Move should be judged less on how convincing the feint looks and more on what angle it forces the defender to commit to — the convincing part depreciates against better opponents, the angle does not. Second, the value of a Signature Move changes composition as a player moves up levels: in youth and lower divisions the perceptual half dominates, higher up only the mechanical half is reliably left. A move that stops working after a promotion may never have had a mechanical half at all. Third, and this is the part that surprised me, physical conditioning raises the ceiling on a Signature Move directly. Better deceleration and better centre-of-mass control mean you commit less and the defender commits more. The habit and the body are not two separate development tracks; the body sets how large the residue can be.

How much is it actually worth

Brink and colleagues (2023) fitted 623 MLS dribbles to a pair of coupled equations of motion and recovered a parameter for how hesitant the defender is — how much he protects the goal versus how much he commits to the attacker. Defenders facing an elite dribbler were 27 percent more hesitant, at the median, than against a non-elite one. In the same study, raising a player's top speed by 25 percent improved their dribble score by only 2 to 8 percent. Those two numbers are not in the same units — one is the opponent's parameter and one is the player's own score, and I am not dividing them — but the shape is hard to miss. A 25 percent gain in top speed is close to unobtainable for a professional and buys single digits. The behavioural layer moves the opponent by double digits. That is the honest case for why a Signature Move is worth building: not everyone can get faster, and getting faster does not pay very much.

What I still cannot prove

One thing is missing and I would rather say so than paper over it. Nothing in this literature shows that a player with a sharper fingerprint performs better. Brink's elite label was assigned before the analysis, not measured from the movement, so it is not evidence that distinctiveness itself is what moves the defender. What the research did give me is a definition I can actually compute: combine Marineau and colleagues' 2024 review, where 48 of 59 studies found experts vary less within themselves, with Burdack and colleagues' 2020 result identifying individuals from their kinematics with 100 percent accuracy, and a signature becomes a separability ratio — small variation within a player, large distance between players. That is a number, and it can be regressed against output. Until someone runs it, the claim that every player needs a Signature Move stays an argument rather than a finding, and I would rather hold it that way than overstate it. A second gap I have since closed off elsewhere: this entry only counts what the defender pays — the attacker pays too, and at the top level that is what decides the duel. See Execution Cost.