Does the lead wrist have to be flat at the top?

Nothing measured here says so. The twelve PGA professionals in Carson's study averaged between 1.7 and 14.4 degrees of extension depending on grip, and still delivered a face within a few degrees of square. The handicap data shows a direction rather than a target: better players remove extension between address and the top, worse players add it.

Should I try to bow it like Dustin Johnson?

His figures are 45 degrees of flexion at the top on HackMotion's sensor and 21 still there at impact, about sixty degrees from their sub-10 handicap average. That is not a stronger version of the usual pattern but a different one, and the same page notes it needs the flexibility and timing to match.

Does a strong grip cup the wrist more at the top?

In Carson's twelve professionals it did, and by a lot: 14.4 degrees of extension at the top under a strong grip against 1.7 under a weak one. The clubface at impact went the other way, squarest with the strong grip. Wrist angle at the top only means something alongside where the hands sit on the handle.

Can a cupped top be squared on the way down?

The three published figures for how far lead-wrist extension changes between the top and impact are about one degree, eleven and nineteen, depending on the measuring system. None of them covers a top twenty degrees out on its own. Something else, forearm rotation or hand path or the grip, has to do the work.

Why does a swing app disagree with a wrist sensor?

They measure different quantities. A wrist sensor reports anatomical extension against a zero calibrated on your forearm. A camera reports a projected elbow-to-knuckle angle in the image plane, which mixes cupping with side-to-side deviation and loses whatever pointed at the lens. Compare each against its own history, never one against the other.

The swing, position by position

Cupped, flat or bowed: the lead wrist at the top

A flat lead wrist at the top is the most repeated checkpoint in golf instruction. The measurements behind it are stranger than the rule, and they disagree with each other by more than twenty degrees.

A golfer in a cap and patterned shirt at the top of the backswing, arms and gloved hand filling the frame. U.S. Air Force photo by Airman 1st Class Alexander Vasquez, Public domain

What the angle is, and where everybody starts

+5.5° against −4.8°Lead-wrist extension added between address and the top: 25-plus handicap against sub-10

One joint, one axis, one number. Extension is the back of your lead hand moving toward the top of your forearm, which golfers call cupping. Flexion is the same joint travelling the other way, toward the palm, which golfers call bowing. A wrist sensor writes extension positive, flexion negative, flat zero.

HackMotion, which sells a forearm sensor, has published a breakdown of more than a million recorded swings by handicap. At address the three groups are almost identical, and all cupped: 19.3 degrees of extension below 10 handicap, 19.6 between 10 and 25, 18.8 above 25. They separate going back. The sub-10 group reaches the top at 14.5 degrees, having lost 4.8. The 25-plus group reaches it at 24.3, having added 5.5.

Nobody in that sample sets up flat. The ten-degree gap at the top is manufactured on the way back, between address and the top, and manufactured in opposite directions by the two ends of the handicap range. That is a more useful thing to know than any target position.

One degree of wrist for one degree of face

12.7° of wrist, 4.9° of faceSpread across weak, neutral and strong grips, twelve PGA professionals

The rule attached to all this is that a degree of extension at the top buys a degree of open clubface. No published measurement of that ratio appears anywhere, only the assertion. One study has recorded both ends of it in the same swings.

Carson, Richards and Mazuquin filmed twelve PGA professional coaches with ten Qualisys cameras at 300 Hz, driver in hand, under a weak, a neutral and a strong grip. Lead wrist at the top: 1.7 degrees of extension weak, 6.6 neutral, 14.4 strong. Clubface at impact: 6.36 degrees right of the target line weak, 2.57 neutral, 1.51 strong.

Put the two rows together. Twelve and a half degrees of wrist moved the face under five, and moved it the wrong way for the rule: the most cupped top delivered the squarest face. The authors call that somewhat contrary to the coaching explanation. Grip shifted both numbers at once, so the ratio only means anything with the hands fixed on the handle.

Dustin Johnson, and a sign that will not sit still

HackMotion's figures for Dustin Johnson run 25 degrees of extension at address, 45 degrees of flexion at the top, a 70-degree change they call one of the largest seen on tour, and 21 degrees of flexion still present at impact. Against their own sub-10 figure of 14.5 degrees of extension, he sits about sixty degrees away.

Their two pages then contradict each other. The Johnson article gives low handicappers about 14 degrees of flexion at the top; the data article's table gives the same group 14.5 degrees of extension. Same magnitude, opposite sign, one company. Use the size, not the sign. Dana Dahlquist, writing for Golf.com, describes three separate working patterns among flexed-wrist tour players, which argues against a single target.

Hear it

Three clicks: takeaway, top, impact. Start at the tour beat and move to whichever one you can actually swing to.

How much of it survives the way down

About 1°, 11° or 19°Change in lead-wrist extension from the top to impact, three published sources

Three sources have measured how far the lead wrist travels from the top to impact. Carson's professionals on a neutral grip went from 6.6 degrees of extension to 5.5, about one degree. HackMotion's sub-10 group went from 14.5 to 3.6, about eleven. Zhang, Liu and Liu, recording twelve female professionals on Qualisys at 250 Hz, put the driver lead wrist 1.2 degrees flexed at the top and 20.2 at impact, about nineteen.

One degree against nineteen is not a disagreement about golfers. It is three instruments with three zeroes: different marker sets, calibrations and definitions of neutral. What survives all three is the direction, toward less extension, and a scale smaller than the folklore implies. In Carson's professionals the large move on the way down was on another axis, from 26.3 degrees of radial deviation at the top to 24.2 of ulnar at impact.

What a camera can see of it

HackMotion makes one point worth holding on to: five to ten degrees of extension is difficult to see, and that much is enough to noticeably open the face. The lead hand at the top is small in the frame, moving, and often hidden behind your head or the shaft.

Pebble Swing reads joint angles out of the frames and reports a lead-wrist angle at the top. Be exact about what that is: the elbow-to-wrist-to-knuckle angle as the lens projected it, where 180 degrees is a straight line from forearm to knuckle. It is not the cupping angle a sensor measures. It blends extension with side-to-side deviation and flattens whatever pointed at the camera.

Use it against your own earlier swings from the same camera position and against nothing else. Two degrees of difference between clips filmed a week apart, from a tripod you moved, is noise. If the answer matters at five degrees, the instrument for it straps to your forearm.

Questions

Does the lead wrist have to be flat at the top?: Nothing measured here says so. The twelve PGA professionals in Carson's study averaged between 1.7 and 14.4 degrees of extension depending on grip, and still delivered a face within a few degrees of square. The handicap data shows a direction rather than a target: better players remove extension between address and the top, worse players add it.
Should I try to bow it like Dustin Johnson?: His figures are 45 degrees of flexion at the top on HackMotion's sensor and 21 still there at impact, about sixty degrees from their sub-10 handicap average. That is not a stronger version of the usual pattern but a different one, and the same page notes it needs the flexibility and timing to match.
Does a strong grip cup the wrist more at the top?: In Carson's twelve professionals it did, and by a lot: 14.4 degrees of extension at the top under a strong grip against 1.7 under a weak one. The clubface at impact went the other way, squarest with the strong grip. Wrist angle at the top only means something alongside where the hands sit on the handle.
Can a cupped top be squared on the way down?: The three published figures for how far lead-wrist extension changes between the top and impact are about one degree, eleven and nineteen, depending on the measuring system. None of them covers a top twenty degrees out on its own. Something else, forearm rotation or hand path or the grip, has to do the work.
Why does a swing app disagree with a wrist sensor?: They measure different quantities. A wrist sensor reports anatomical extension against a zero calibrated on your forearm. A camera reports a projected elbow-to-knuckle angle in the image plane, which mixes cupping with side-to-side deviation and loses whatever pointed at the lens. Compare each against its own history, never one against the other.

Sources

Every number on this page comes from one of these.