Hip Turn, X-Factor and Shoulder–Hip Separation

Hip turn is how far your pelvis rotates away from the target. Shoulder turn is how far your upper torso rotates. X-factor is the difference between the two at the top of the backswing — the separation, not the shoulder number on its own. This guide describes and measures both.

This article is descriptive. It explains what these angles are and how they are measured from video. It is not an instruction programme, and it deliberately contains no physical training of any kind. For individual instruction — including whether any change to your motion is appropriate for you — work with a qualified coach.

Key takeaways

• Hip turn and shoulder turn are two separate rotation measurements, taken from the same frame of video.
• X-factor is a derived number: shoulder turn minus hip turn at the top of the backswing.
• The separation is what people mean when they talk about "coil". A big shoulder turn with an equally big hip turn produces very little separation.
• X-factor stretch describes separation continuing to grow early in the downswing, because the hips start forward while the upper body is still finishing back.
• The concept is widely cited, but "more is better" is not what the research literature supports. It is a description of a pattern, not a target.
• A 2D camera measures these angles by projection, which introduces real, predictable error. Know the limits before you trust a number to the degree.

What is hip turn in the golf swing?

Hip turn is the rotation of your pelvis around your spine, away from the target, measured in degrees from its address position. Zero degrees means the pelvis is still square to where it started. Fifty degrees means the belt line has rotated fifty degrees away from the target. It is a rotation measurement, not a measurement of sway or lift.

The reference for the measurement matters. The usual convention is to take address as zero, then measure the pelvis line — a line drawn through the two hip joints, viewed from above — against that starting line. From a face-on camera you cannot see that line directly, so software infers it from how the two hip landmarks move horizontally relative to each other. More on that limitation below.

Hip turn is not the same as weight movement. A golfer can rotate the pelvis a long way while pressure stays badly distributed, and a golfer can shift pressure correctly with a modest turn. When pressure moves the wrong way during the backswing, you get a reverse pivot, which is a separate fault with its own signature on camera.

What is shoulder turn, and how is it different from hip turn?

Shoulder turn is the rotation of the upper torso, measured from the line through both shoulder joints against its address position. It is a different segment from the pelvis, so it gets its own number. In almost every golfer the shoulders rotate further than the hips, because the torso can rotate on top of a pelvis that has rotated less.

The two numbers are usually reported together and at the same moment — the top of the backswing, or P4 in the P1–P10 sequence. Reporting them at different moments makes the pair meaningless, because both are changing continuously through the swing.

A common confusion is treating "shoulder turn" as a proxy for how good the backswing is. It is not. A shoulder turn number tells you how far the torso rotated; on its own it says nothing about what the pelvis did underneath it, which is the half of the picture that actually generates separation.

What is X-factor?

X-factor is the difference between shoulder turn and hip turn at the top of the backswing. If the shoulders have turned 90° and the hips 50°, the X-factor is 40°. It is arithmetic on two measurements, not a third thing measured directly. The term was popularised in golf instruction media in the early 1990s.

Because it is a subtraction, X-factor can be the same number for two very different swings. Shoulders 100° over hips 60° gives the same 40° as shoulders 80° over hips 40°. Those are not the same backswing, and a single X-factor figure will not distinguish them. This is why any honest report gives you all three numbers — shoulders, hips, and the difference — rather than the difference alone.

Some analysis systems define the angles against slightly different references (address-relative versus target-line-relative, pelvis-relative versus global). The numbers shift accordingly. Two apps can both be right and still disagree by several degrees, which is a good reason to compare a number to your own previous number rather than to a figure quoted somewhere else.

Why does the separation matter more than the shoulder number?

Because separation is what stores the relationship between two segments, and the shoulder number alone cannot describe a relationship. A 95° shoulder turn over a 90° hip turn is 5° of separation — the whole body has simply rotated together. The same 95° over a 45° hip turn is a completely different backswing.

Think of the two segments as the two ends of a torsion. What matters mechanically is how much one has moved relative to the other, and in which order they then unwind. A golfer who rotates everything together as one block reaches the top with a big-looking turn and almost no differential between segments.

This is also why "turn your shoulders more" is such an unreliable instruction. If the pelvis follows the torso by the same amount, the shoulder number goes up and the separation does not move. The visualiser below lets you see that directly: hold the shoulder dial still, move the hip dial, and watch the separation readout change without the shoulder number changing at all.

What is X-factor stretch?

X-factor stretch describes separation that continues to increase after the top of the backswing. The pelvis begins rotating toward the target while the upper torso is still completing its motion away from it. For a brief window the two segments are moving in opposite directions, so the difference between them grows past whatever it was at the top.

This is the sequencing story that sits underneath the whole concept. The transition is not a single instant where everything reverses at once. Viewed frame by frame, the lower body changes direction first, the torso follows, then the arms and finally the club. That order is what people are describing when they say a swing "sequences from the ground up".

In the SwingOK sample swing, the note at the top says the hips fire first at transition, and the club shallows and drops as a consequence. That ordering is the mechanism. Reverse the order — upper body first — and the club typically moves out and over, which is the pattern behind an over-the-top move and the resulting slice.

Measuring stretch well is harder than measuring the top, because it needs frames close enough together to catch a fast-moving transition. At 30 or 60 fps the transition can pass in one or two frames. At 240 fps you get enough frames to see the order of events, which is what a higher frame rate is actually for.

What do the numbers look like in a real example?

Here is one labelled swing — the sample swing used throughout SwingOK — measured at P4, the top of the backswing. Shoulders 92°, hips 51°, giving 41° of separation. The accompanying note flags the hips as roughly 7° more open than this particular swing wants. These are numbers from one swing, not a target for yours.

Measurement at P4	Sample swing
Shoulder turn	92°
Hip turn	51°
Separation (X-factor)	41°
Note on the swing	Hips approximately 7° more rotated than this swing wants

Read that table the right way round. The interesting part is not "92" or "51" — it is that the note is about the hips, even though the shoulders are the bigger number. If the hip turn came down, the separation would go up without the shoulders doing anything different. That is the whole point of treating the pair as a relationship.

There is no universal correct value in that table. Golfers differ in build, in how they set up, and in what their motion is organised around. A number that describes one player accurately can describe another player's swing badly, which is why the sensible use of these measurements is tracking change in one golfer over time.

Is more X-factor always better?

No. This is the part of the popular story that outruns the evidence. The idea that separation relates to how a swing generates speed is well established in general terms, but "more separation produces more distance" is a much stronger claim than the biomechanics literature supports, and it is not how coaches use the number in practice.

Several things get lost in the simplified version:

1. Separation is a byproduct, not a control. It results from how the pelvis and torso move and in what order. Chasing the number directly usually changes something else — often the shape of the backswing — rather than the underlying pattern.
2. The pairing matters. Two golfers with identical X-factor can have completely different hip and shoulder numbers, and their swings will not behave the same way.
3. Sequencing outranks magnitude. A modest separation that unwinds in the right order can deliver the club better than a large separation that unwinds out of order.
4. Individual variation is large. Published biomechanics work on golf, indexed through PubMed, consistently finds meaningful spread between players. Averages across a group do not describe any single golfer in it.

The honest framing: X-factor is a useful descriptive number. It tells you something real about the shape of a backswing, it is easy to track over time, and a large change in it usually means something in the motion changed. It is not a score, and it is not a lever you pull.

How does a 2D camera measure shoulder and hip turn?

A phone camera does not see rotation directly. It sees where body landmarks appear in a flat image, and rotation is inferred from how those landmark positions change. As the torso rotates away from a face-on camera, the two shoulder landmarks appear to move closer together horizontally. That foreshortening is the signal the software converts into an angle.

The chain runs like this:

1. A pose model locates body landmarks in each frame — the joint positions a pose estimator tracks.
2. The two shoulder landmarks define a line; so do the two hip landmarks.
3. The apparent horizontal width of each line, compared with its width at address, gives the projected rotation.
4. That projection is converted to degrees against the address reference.
5. Shoulder minus hip at the top gives the separation.

Camera position feeds directly into every step. A face-on view resolves rotation well because it sees the foreshortening cleanly; a down-the-line view resolves plane and path better but sees rotation edge-on. Which is why the two standard angles answer different questions, and why a proper camera setup is not a formality.

What are the limits of measuring turn from a phone video?

Real ones, and worth knowing before you trust a degree. A single 2D camera reconstructs a three-dimensional rotation from a flat projection, so error creeps in from camera angle, lens distortion, landmark noise, clothing, and the fact that pelvis rotation and pelvis tilt look similar in projection. Treat the numbers as consistent, not absolute.

The specific limitations:

• Camera angle. If the phone is not square to the golfer, the address reference is already skewed and every subsequent angle inherits that skew. Small setup differences change the numbers.
• Depth ambiguity. One camera cannot separate rotation from tilt perfectly. A pelvis that tilts can read as a pelvis that turned.
• Landmark noise. Pose models estimate joint centres from surface appearance. Loose clothing, low light and motion blur all degrade that estimate.
• The reference frame. "Zero" is your address position. If your address changes between sessions, so does every angle measured against it.
• Frame rate. Angles at the top are relatively forgiving. Anything about the transition needs a high frame rate to be meaningful.

None of this makes the measurement useless. It makes it relative. Filmed the same way each time, from the same spot, the numbers are a reliable record of how your own motion changes. Compared across apps, camera positions or golfers, they are much softer than the decimal place suggests. Marker-based laboratory systems of the kind used in published biomechanics research exist precisely because this is hard to do accurately.

How should you actually use these numbers?

As a record of change in your own swing, filmed consistently, over time. Log the shoulder number, the hip number and the difference together. Watch which of the three moves when your ball flight changes. Do not compare your figures with a number quoted from a tour player, and do not treat any published range as a target to reach.

Three practical habits make the numbers worth keeping:

• Film the same way every time. Same distance, same height, same angle. The measurement is only as stable as the setup.
• Record all three numbers. Separation alone hides which segment moved.
• Pair them with ball flight. A change in the angles that comes with no change in strike or shape is not necessarily a change worth having.

SwingOK is built to take this measurement off you — film one swing on your phone and the shoulder and hip angles come back from the frames themselves, so the comparison over time is like-for-like. What it does not do, and no measurement tool does, is tell you whether a change is right for your body. That is a conversation for a qualified coach who can see you in person.

Where should you go next?

If you want to work through the rest of the mechanics, these connect directly to what is on this page:

• The Golf Swing Sequence: P1–P10 Explained — where P4, the moment these angles are measured, sits in the full swing.
• The Reverse Pivot — what happens when pressure moves the wrong way during the backswing, and why it is not the same thing as a small hip turn.
• Golf Swing Plane and Club Path Explained — the other half of the transition story, on the club side.
• Casting vs Lag — what an out-of-order downswing costs at the delivery position.

You can also try the Shoulder & Hip Turn Visualiser on its own, or the P1–P10 Swing Position Explorer to see the sample swing frame by frame.