Eight Layers of Golf Analysis: The Discipline of the Empty Cell

EIGHT LAYERS OF GOLF ANALYSIS: THE DISCIPLINE OF THE EMPTY CELL

The analysis request arrived on a Tuesday morning. I opened the file, scrolled to the first line of the information table, and read the answer that not every analyst is willing to type: "Analysis subject: insufficient information." No shot-level data. No approach metrics, no green speeds, no wind direction, no field depth. Just a headline, a few summary lines, and an unspoken expectation that whoever sits behind the desk will turn that empty space into three thousand words.

Eleven years ago I did the exact opposite. In 2026, while studying international communication in Binh Duong, I spent three months building an xG model in Excel to analyse 26 rounds of the V.League. The output showed Quang Nam FC winning the title with an average possession share of 48 percent — the lowest of the leading group — but a shooting conversion rate of 17.5 percent, among the highest in the league. I published a piece titled "The Champion Doesn't Need the Ball," was mocked for it, and three months later Quang Nam lifted the trophy. The article was shared more than two thousand times.

The lesson was not that I had been right. The lesson was that a model only has value when it is built on real data, and the writer is accountable for every cell. When I moved into writing about golf for the Vietnamese market, I carried that discipline with me — and found a paradox: the more people write about golf, the fewer empty cells anyone is willing to admit to.

Data does not lie. But reputation whispers into the ear of anyone who never reads the table.

Why an empty cell is more trustworthy than a filled one

Vietnamese golf is in a phase where courses and tournaments are multiplying faster than measurement infrastructure. Song Be Golf Resort in Binh Duong opened in 2026 and is widely recorded as Vietnam's first international-standard golf course. Three decades later the country has dozens of courses from north to south; the figure commonly cited in industry summaries is somewhere between 70 and 80, depending on the source and on how "operational" is defined.

But a course count is not data. A golf course can be open for ten years without producing a single line of shot-level data usable for analysis. This is the fundamental difference between golf and football, the sport I came from. In football, every match generates thousands of data points automatically: ball position, pass counts, PPDA, circulation speed. In golf, data does not generate itself. It has to be actively collected, shot by shot, with equipment and with a person sitting there recording.

That gap creates a professional habit that is very hard to break: filling empty cells with narrative. When there is no approach metric, the writer uses a feeling about form. When there is no green speed, the writer uses the course's reputation. When there is no field depth, the writer uses the number of star names on the promotional board. Every time a cell is filled that way, a wrong conclusion is born and distributed with a highly professional appearance.

The eight layers below are the order I use when assessing any round, any player, any tournament. The order is not meant to make an article longer. It is meant to answer one question before every conclusion: does this cell have data, or am I talking to myself?

The technical layer: four doors of Strokes Gained

Strokes Gained was developed by Mark Broadie of Columbia Business School and adopted into the PGA Tour's published statistics from 2026. Its principle is simple enough to be easily underestimated: each shot is compared with the PGA Tour average from the same distance and lie, and the difference is the value of that shot. The metric splits into four categories: off the tee, approach, around the green, and putting.

The greatest value of the method is that it removes distance from the equation. A putt from 1.2 metres is not judged the same as a putt from nine metres. A chip from the fringe is not judged the same as a shot from 40 metres off the green. That is why rounds with extremely low putt counts are so often misread by media. A 22-putt round sounds like a masterpiece on the greens, but if that player hit only eight of 18 greens and six times left the ball inside a metre after a chip, the actual putting figure may be merely average. The scorecard praises the putt. The data praises the chip.

According to Broadie's research, across a full PGA Tour season, most of the scoring difference between players comes from the long game — off the tee and approach — rather than from putting. Putting still matters, but its weight rises sharply as the sample shrinks. This point is critically important for Vietnamese media, where a single round or a three-day event is routinely used to draw conclusions about an entire career.

In Vietnam, shot-level data at domestic tournament level essentially does not exist. Launch monitors such as TrackMan or GCQuad are present at many practice facilities, but they mostly serve club fitting and range measurement. The same device, two entirely different purposes: measuring a shot under ideal conditions tells you nothing about how a player handles the 14th shot when tired, when the wind turns, when holding a fragile one-shot lead.

The consequence is that every technical analysis of Vietnamese golf today must state its limits. I can analyse swing mechanics from video. I can compare launch monitor numbers between two fitting sessions. I cannot say player A has a better approach profile than player B if neither has shot-level tournament data. The correct sentence in that situation is: not enough data to conclude.

The human layer: the age curve and the form curve

Golf has an unusually long age curve. If peak speed for a footballer sits between 24 and 28, golf allows a player to remain competitive at major level far longer. Phil Mickelson won the 2026 PGA Championship at 50, becoming the oldest major champion in men's golf history. Tiger Woods won the 2026 Masters at 43, after years of back injuries and multiple surgeries.

Those two facts do not mean age is irrelevant. They mean the effect of age on golf performance must be decomposed into components. Clubhead speed may decline with age, but course management, target selection and decision-making under pressure often improve with experience. A 40-year-old who has lost 12 metres off the tee but gained a better decision sequence can still win.

To assess form seriously I need at minimum four data groups: world ranking and its 12-month trend; current tour tier; recent results with an explicit sample size; and major record — wins, top-10s, cut rate, and above all the conversion rate from contention to victory.

Conversion rate is the most neglected number. One player may have 15 major top-10s without a win. Another may have four top-10s and three wins. By top-10 count the first looks more consistent. By conversion rate they are in different categories. Media usually picks the first metric because it narrates easily as "persistence will eventually pay."

Based on my experience following matches and rounds for which data exists, a player's form only genuinely shifts when at least one of those four groups changes with statistical significance — not when he shoots a 64. A 64 is an observation. It is not a trend. Confusing the two is the most common error in all sports writing, not just golf.

The tournament layer: weight lives in field depth

A title does not have a fixed value. It has a weight, and that weight depends on the field. The world ranking system awards points based on the rankings of the participants and their positions at the time of the event. The same trophy therefore carries a completely different point value depending on whether the winner beat a 30-player field with 25 players outside the world's top 300, or a 156-player field with 60 players inside the top 100.

This is why reading golf results while ignoring field depth is a structural error. A player can win four events in a year yet accumulate fewer ranking points than someone who won two events at a higher tier. Meanwhile a young player winning a thin-field event will be praised domestically as a career breakthrough, while the ranking data shows he remains a long way from major eligibility.

Beyond field depth, tournament structure affects a player's career through four channels: world ranking points, prize money and commercial value, tour card retention, and season rhythm. Rhythm is the most underrated channel. A player moving from Asia to Europe to the United States across three consecutive weeks accumulates not only physical fatigue but also error in swing timing and green-speed calibration. Events in different regions run different green speeds, and green speed directly affects putting in ways the scorecard cannot express.

Within Vietnam, most domestic events sit outside the world ranking system, or only a small subset carries points. The consequence is that no objective yardstick exists to compare Vietnamese players with each other, let alone with the region. When there is no yardstick, people use reputation. Reputation is a metric with no sample size, no confidence interval, and no mechanism that ever corrects it when it is wrong.

The governance layer: when money and rules stand on opposite sides

Modern professional golf cannot be analysed while ignoring governance. LIV Golf launched in 2026 with funding from Saudi Arabia's Public Investment Fund, running 54-hole, no-cut events with team formats. In October 2026, the world ranking body announced that LIV Golf did not meet the criteria to receive ranking points.

That decision had very concrete technical consequences, not merely symbolic ones. Without ranking points, the pathway to majors narrows. A player moving to LIV does not lose major access immediately, because many majors have their own criteria such as recent victory or a ranking threshold. But over time, as old exemptions expire, major access depends increasingly on world ranking — precisely the pathway that was closed.

Alongside that, on 6 June 2026 the PGA Tour, the DP World Tour and Saudi Arabia's Public Investment Fund unexpectedly announced a framework agreement to consolidate commercial interests. In January 2026, PGA Tour Enterprises announced an investment from the Strategic Sports Group with a potential value of up to three billion US dollars. The governance picture is therefore in constant motion, and any analysis that freezes it at a single moment will be out of date.

For Vietnamese viewers this layer looks distant. It is not. The schedules of major tours determine when and where leading players can appear in Asia. Equipment brands' sponsorship policy determines how much budget flows into emerging markets. And tour membership rules determine whether a Vietnamese player has any chance of receiving an exemption.

The rules and equipment layer: changes with exact effective dates

Golf rules and equipment regulation is the layer media rarely covers, yet it has the clearest dates and is therefore the most forecastable. The ban on anchoring the club against the body, under Rule 14-1b, took effect on 1 January 2026. It was announced years in advance and directly affected the group of players using long putters.

The next major change concerns the ball. On 6 December 2026, the USGA and the R&A announced revised ball testing conditions, raising the test clubhead speed and adding a spin criterion. The rule applies to elite competition from January 2028 and to recreational play from January 2030.

For Vietnamese golf, the milestone to watch is 2030 for most players, plus the possibility that domestic professional events adopt the elite standard earlier. When that happens, a practical question appears: how many Vietnamese players are using balls in the affected group, and what is the switching cost measured per practice round? That is a question data can answer. Emotion cannot.

What is worth noting is that rules and equipment changes always come with long transition windows, and the transition window is precisely when competitive advantage shifts. Early movers hold an edge for roughly two to four seasons before the field catches up. I have seen this in football with changes to offside law and substitute numbers. The mechanism in golf is no different.

The risk layer: probability instead of adjectives

Risk in sports analysis must be described with probability, not with adjectives. "This player tends to fold on Sunday" has no analytical value. "Based on 40 rounds with data, the probability of losing at least two shots over the final nine is 31 percent, within an acceptable risk band if the objective is a top-10 finish" does.

The risk matrix I use has six groups. Competitive risk: a direct rival stronger on the decisive stretch. Psychological risk: handling the decisive shot at 17 or 18. Injury risk: history and workload. Career and commercial risk: losing playing status or sponsors. Governance risk: changes to entry conditions. Systemic risk: equipment or rule changes affecting an entire cohort.

A concrete example of misreading risk. When a player has three events with excellent putting, media concludes he has "found his feel again." But putting has the highest variance and the clearest mean reversion of the four Strokes Gained categories. Three events is a small sample. The probability that the streak continues in a fourth event, based on the streak alone, is barely above the base rate.

There is one more risk group Vietnamese golf is accumulating: the risk carried by young players pushed into excessive competition and training volume before their bodies mature. In football I repeatedly presented data on early-developing youth players being overused. In golf the same mechanism appears as 36-hole practice days and dense amateur calendars. Back and wrist injuries among under-18s are warning signals, not random accidents.

The narrative layer: the heat cycle of a story

Every famous player exists in two places at once: on the ranking table and in the story the public tells about him. The two do not always match, and the gap between them is the raw material of most bad analysis.

Rory McIlroy is the clearest example of the past decade. His story was the story of a man who could not complete the Grand Slam. His data, across most of that period, showed final-round scoring still among the best in the world, with the error concentrated in specific rounds and in putting on fast greens. In April 2026 he won the Masters and completed the Grand Slam. The collapse narrative died in a single day. The data did not change that abruptly.

The mechanism is the heat cycle of narrative. A story retold often enough generates its own evidence: every failure is cited, every success is called an exception. This structure cannot be refuted with data, because it never ran on data in the first place.

In Vietnam this cycle often attaches to young players. A 15-year-old winning an amateur event is called a phenomenon. Three months later, if he fails to hold form, the story becomes "the talent has stalled." Both descriptions skip one simple data question: how many rounds measured that form, on which course, in what weather, and at what level were his opponents?

The transmission layer: from the green to the balance sheet

Golf analysis does not stop at the course. Every result transmits through a long chain before it touches the real economy.

First, course economics. A well-performing course raises surrounding property values, and conversely, property demand determines the ability to invest in course upgrades. In Binh Duong and neighbouring provinces, the golf-course-plus-residential model has shaped how the industry developed over two decades. Analysing an event at such a venue requires reading two layers: the sporting layer and the infrastructure layer.

Second, equipment brands. Club and ball sales are directly influenced by elite results, usually with a one to two season lag. A player winning a major with a specific model creates measurable demand in new markets, Vietnam included.

Third, sponsorship and broadcast. Individual endorsement values in golf are tied closely to broadcast hours and to how often a player appears in the final-round lead group, not merely to title counts.

Fourth, data and analytics products. This is a fast-growing segment and one where the Vietnamese market has a large deficit. A round-by-round form index system — for example a player depth index of the kind regional data platforms are building — can partially fill the gap while shot-level data does not yet exist.

Fifth, the talent pipeline. The number of properly coached young players depends on how many events carry ranking points, on training costs, and on access to courses. It is a slow variable, moving on a five-year cycle rather than a tournament cycle.

Sixth, the capital network. Investment flowing into Vietnamese golf courses over many years has come from multiple sources, including Asian investors. When that flow changes direction, the domestic calendar changes with it, and when the calendar changes, Vietnamese players' opportunities to accumulate ranking points change too. The chain is long but traceable.

The contrarian angle: the trap of the empty cell

The point where I disagree with the majority sits here. In data analysis, people assume value lies in the cells that contain numbers. In golf, and especially in Vietnamese golf right now, I argue the greatest value lies in the empty cells — provided the emptiness is correctly recorded.

A recorded empty cell identifies exactly where the measurement system is broken. No green speeds means the problem sits in course data collection. No field depth means the problem sits in tournament organisation. No conversion rate means the problem is sample size — Vietnamese players simply have not played enough rounds at a comparable level. Every empty cell is a specification.

A cell filled with narrative specifies nothing. Worse, it blocks questioning, because it already contains an answer.

There is a second error class I call benchmark misapplication. A writer takes a PGA Tour standard — say, the relative weight of Strokes Gained categories — and applies it directly to a domestic amateur event. This is a category error, not an arithmetic one. Course conditions, green speeds, field depth and even access to measurement equipment all differ. A metric that is correct in environment A is not automatically correct in environment B, and an honest analyst states where the benchmark was borrowed from.

A third error is turning correlation into causation. A player leads putting and wins an event. The conclusion drawn: putting decides victory. But his approach number in that event may also have been among the leaders, and approach is what brought the ball close to the hole and made the putting look good. In golf the categories are so tightly linked that isolating one to celebrate is almost always methodologically wrong.

And the fourth, most common error: judging with a single yardstick. One metric is never enough to convict or to crown a player. That holds for xG in football and for Strokes Gained in golf. If I have only one number, I have nothing.

Data does not lie. But a number standing alone stays silent, and in that silence, reputation speaks instead.

Takeaway

If I have to pick one signal to track for the next cycle, I pick data infrastructure. The 80th golf course will not make Vietnamese golf analysis better. A system that records every shot at one domestic event will. When we can read approach numbers from a Vietnamese round, when we know how a course's green speed shifts between morning and afternoon, when we have enough rounds to compute conversion rates — that is when domestic golf writing starts to carry genuine information gain.

Until then, the most honest thing an analyst can do is leave the empty cells empty and label them clearly. Not out of ignorance. Because understanding begins with knowing what you do not yet know.

I hate uncertainty. But I have learned that one unforeseen variable can be stronger than any algorithm, and the only way to live with it is to write it into the table from the first row, instead of letting it appear in the final minute as an excuse.

I do not predict. I read the data and accept the consequences. Even when the data is empty.