Four Swing Thoughts on the First Tee and the Dispersion Gap Nobody Measures

Eight in the morning at a course in Aichi, six degrees Celsius. A 12-handicap player sets the ball on the first tee, steps back two paces, and recites four things: turn the right shoulder fully, do not roll the wrists, keep the face slightly open, and do not push it right. The ball goes right, fourteen metres from the target zone.

Forty minutes earlier, on the range, the same man with the same driver put eight of ten balls inside a zone roughly ten metres wide. Same swing. Exactly one difference: nobody heard him recite those four things.

Jamie Mulligan, a PGA Professional and one of the GOLF Top 100 Teachers, calls the range state "in it." He calls the first-tee state standing outside the party. His fix fits in three steps: see the ball flight, feel the shot, trust what was built in practice. He coaches Patrick Cantlay and Nelly Korda, and that detail is placed immediately before the method is explained, functioning as a credibility bridge.

Across the entire argument there is not one line of dispersion data. No strokes gained, no ShotLink, no measurement of pattern width under two different attentional states. I read the instruction piece twice: once to understand what it says, once to find what it measures. The second pass took far less time.

Why an instruction piece deserves this much analysis

Golf is the sport where amateurs consume more technical information per week than most professionals in other sports. The phone holds slow-motion video of tour players, thousands of instruction clips, and biomechanical models labelled correct. The 12-handicapper in Aichi can describe the exact wrist position at the top of Rory McIlroy's swing and cannot reproduce his own.

Mulligan offers a hypothesis worth testing: that this abundance may be a net burden. The more a golfer understands about the swing, he argues, the more opportunities exist to interfere with it.

I do not argue with the conclusion. I argue with how it is presented, and with the fact that it never states its own boundary conditions. Those are two different faults. The first belongs to editing. The second can hurt a reader.

The source sits in GOLF.com's Instruction section, which is a persuasive genre rather than a reporting genre. It carries no obligation to cite. But it asserts a causal relationship, and a causal relationship — even a cognitive one — still has to survive measurement.

The mechanism is right, but not because of this source

The first thing I did after finishing was reopen the motor-learning literature. Not because I doubted Mulligan, but because I wanted to know where he stands in a debate that has run for thirty years.

The central principle — conscious control degrades execution, reacting to a target preserves it — maps almost perfectly onto the attentional-focus literature. Gabriele Wulf's 2026 review of fifteen years of data concluded clearly that in skilled performers an external focus (on the target, on the effect of the movement) reliably outperforms an internal focus (on body parts). The mechanism is the constrained action hypothesis: when consciousness interferes with an already automated motor program, it re-imposes constraints on degrees of freedom that had been released.

In 2026 Richard Masters published implicit-learning results in putting: a group taught without explicit technical rules held up better under pressure than a group taught explicitly. In 2026 Wulf and Lewthwaite extended this into the OPTIMAL theory, where an external focus is one of three pillars.

The mechanism Mulligan describes has a far stronger empirical foundation than his own article supplies: the principle is correct, but it is correct because of three decades of literature the piece never cites.

This is where I have to be blunt about my own trade. I do not believe in luck; I believe in cultivated probability. A coach saying something the literature has said for decades is not wrong. But when an instruction piece presents an established principle as though it were the discovery of a proprietary method, the reader loses the most valuable thing: the ability to know where they stand on the map.

Two different tasks, and a boundary never drawn

The strongest technical content in the piece is the distinction between developing a swing and playing a shot. Great players, Mulligan says, know when to think and when to trust.

In motor-learning language this is the gap between skill acquisition and skill expression. Acquisition is repetition with feedback, conscious, often blocked — conditions nothing like expression, which is a single trial with consequences and must run implicitly. The two tasks require two attentional states. It is no surprise that a player who is excellent on the range is poor on the first tee: he is applying the attentional state of the first task to the second.

Mulligan also names the classic amateur error: confusing the practice environment with the competition environment. That is a point I accept without needing more data.

But he stops exactly where the next step was needed. Technique belongs on the range and the practice ground — and then nothing about a threshold. How much technique is enough to earn the right to trust? The article does not say. That is where a genuinely harmful gap opens.

For a player with a real mechanical fault — a systematic path-and-face mismatch, a two-way miss, a repeating block — committing harder to the target does not correct the fault. It produces a more consistent version of it. The piece implicitly acknowledges this by locating technique work on the practice tee, but never states the boundary condition. For an article read by hundreds of thousands of amateurs, the boundary condition is the most important part.

The practice–play firewall

I have spent much of seventeen years watching tour rounds in Japan to record something very hard to measure: timing. Not timing within a day, but timing within a week.

In off weeks, professionals do technical work. In tournament weeks, they largely do not. This is an operating convention rather than a biological law, passed from coach to player as a trade habit. That firewall is exactly what Mulligan's article is trying to teach amateurs — except amateurs have nobody guarding the wall for them.

An amateur's week has seven days, six of them practice and one of them competition. There is no off week. Every round is a scored round and every range session is a repair session. With no mode switch, the attentional state has no reason to switch either.

Here I borrow a word from another sport, and with reason. Gegenpressing does not break the data; it breaks my assumptions. In 2026 I built a pressing model and omitted a time-varying variable. The model was right in the first half and useless after the seventieth minute. The lesson was not that pressing is useless, but that any variable that changes over time must be modelled as a variable that changes over time.

A swing thought is exactly that kind of variable. The question is not how many thoughts, but where the last thought falls on the timeline of the shot.

An unfalsifiable diagnosis

One sentence in the piece stayed with me: the problem is not necessarily the golfer's mechanics.

The hedge does the damage. A falsifiable hypothesis must state when it fails. This one exempts itself in advance. If mechanics are the problem, the thesis holds in negative form. If they are not, it holds in positive form. A proposition true in both cases carries no information.

The gap in the table speaks, if we are willing to listen. Here it says: nobody has measured. And I can state exactly what to measure.

A minimum test has four steps. One, select a player with a stable handicap between 10 and 18. Two, measure the dispersion ellipse of 30 drivers under target focus, recording standard deviation on the left–right and long–short axes. Three, repeat under a single designated technical swing thought, randomising condition order. Four, compare pattern width and centre offset.

If Mulligan's hypothesis holds, the swing-thought condition produces a wider but not offset pattern; if the player has a real mechanical fault, it produces a narrower pattern offset to one side. Those two outcomes point to opposite conclusions, and only measurement separates them.

I have not run this at adequate scale. I say so because I once published a model that missed six of ten rounds, and I do not want to repeat the habit of hiding my limits behind a fluent sentence. But the design is clear, and anyone with a launch monitor and thirty balls can run it this weekend.

A contradiction inside the author's own toolkit

There is a detail most readers will skip. The author describes himself as a PGA Professional working with 3D motion analysis, force and pressure plates, high-speed video and launch monitors. That is the most expensive toolkit coaching can buy. And he uses it to argue that knowing more about the swing makes golfers worse.

At a glance this is a contradiction. On inspection it is not. The two serve different audiences. The instruments serve the coach, who needs diagnosis. The words serve the player, who needs execution. A surgeon using a CT scan to plan an operation does not mean the patient should carry the scan into theatre.

But the article never draws that line. A sceptical reader sees the contradiction, discards everything after it, including the parts that are right. That is the price of leaving the audience undefined.

The missing threshold and what it costs

I return to the largest question left open. Trust what was built in practice — but what if what was built in practice is poor?

Mulligan does not answer. He has no obligation to answer in an instruction piece, but the reader needs an answer. And the answer depends on a variable the article never mentions: access volume.

A player who hits forty balls a week and a player with an indoor bay, daily practice and a coach receive the same advice and it means opposite things. The first has nothing yet to trust. The second has too much to trust and needs permission to trust it.

This is where the Vietnam–Japan axis earns its place, and I only use it when the gap is wide enough to matter. In Japan, range density and a repetition-based training culture mean the average amateur touches far more balls per month than the baseline in Vietnam, where practice facilities are thinner and most learning happens on the course, inside a scored round, under consequence. Those two environments produce two different players facing the same advice. The Japanese amateur drifts into information overload without mode switching. The Vietnamese amateur drifts into insufficient base volume while being told to trust something not yet built.

One sentence, two opposite failure modes. That is why I distrust golf advice without conditions attached.

Pace of play: the real contradiction inside the framework

The most analytically interesting tension in the piece has nothing to do with technique.

The state Mulligan describes — select target, select shot, commit to intention, let the body do the rest — is an internal commitment state. It is not automatically a fast one. And Cantlay, named right beside the method, is a golfer who has been publicly criticised many times over pre-shot pace. That is external to the article, but it is public and widely reported.

The framework says nothing about the time cost of deep commitment. There is no number for pre-shot routine duration. And Rule 5.6a of the Rules of Golf covers unreasonable delay — a rule modernised in the 2026 code and still among the hardest to enforce on tour.

I am not claiming the method causes slow play. I am claiming the framework ships without a clock, and any instruction that adds cognitive steps before the shot without adding a time budget contributes at the margin to the problem.

The champion-inference problem

One detail needs stating, because it is the easiest trap in this entire genre.

Cantlay won the 2026 FedExCup. Korda won five consecutive starts in 2026 and holds an Olympic gold from Tokyo. Those are verifiable facts. But nothing in the article shows this method produced those results. The causal arrow is unestablished.

And there is a very old sampling problem. A coach is credited with a player's winning run precisely while the player is winning. The same method is rarely invoked to explain a slump. That is textbook outcome bias, and it is a fault of the genre rather than of this piece alone.

What does not happen often tells the truth better than what does: no article about a struggling student is ever used as evidence for the power of a psychological method.

The contrarian angle: not how many, but when

I began this piece with a silent assumption: fewer thoughts are better. That assumption was mine, not Mulligan's, and I carried it in before reading.

It does not hold.

Two players can hold the same four thoughts. The first recites them at home, in the car, while choosing a club — then steps in with a single target. The second recites them at the exact moment the club starts back. The first has a plan. The second has an intervention. Identical thought count, opposite outcomes.

If that is right, stop thinking is advice aimed at the wrong address. It targets quantity while the mechanism lives in timing. And a player taught to think nothing loses the one tool that genuinely requires consciousness: shot selection.

Data never lies; I simply asked the wrong question. I asked how many thoughts. The right question is where the last thought falls relative to the trigger.

Signals for the next cycle

Three things I will track in the coming rounds, and three things I suggest anyone reading this measures on themselves.

One, driver dispersion width under two attentional conditions, on your own body, without waiting for a publication. Two, pre-shot routine duration, timed with a watch, before and after dropping swing thoughts — because if commitment shortens the routine, the pace-of-play argument reverses, and I want to know whether it reverses. Three, the timing of the final thought, logged with a single marker: before or after the club starts moving.

If two of those three measurements contradict what I have just written, I will rewrite it. I have done that at least twice in my career, and each time it hurt more than the last. But an analyst has no right to keep an assumption merely because it has already been printed.