A fifteen-page file with no tables: lessons from OWGR, the ball rollback and the golf analytics trade

In October 2026, the board of the Official World Golf Ranking met behind closed doors and rejected LIV Golf's application for world ranking points. The submission described a 54-hole format with no cut and an added team component. The board's answer was that there was no way to measure that tour with the same ruler used for the rest of the golf world. I copied the sentence into my notebook and read it four times that evening.

Three months later, a partner sent me a fifteen-page PDF. Full table of contents. Eleven recommendations in the conclusion. The appendix was blank: no tables, no comparison sample, not a single line of ShotLink data. The sender added a note: "Our data is complete, take a look." I replied that there was no data in the file for me to check. The exchange ended there.

Two events a quarter apart, one question: when a file contains no data, how should it be read?

Golf is the most densely measured sport in individual professional competition. From the early 2000s, the PGA Tour deployed ShotLink, installing measuring devices at every hole and recording the coordinates of every shot. A single round now generates tens of thousands of data points. In 2026, Mark Broadie, a professor at Columbia Business School, published the research that underpinned Strokes Gained, turning every shot into a difference against the tour average. Since then, a three-metre putt on the 17th and a three-metre putt on the 2nd sit on the same comparative axis.

The paradox is that the best data infrastructure in the sport sits alongside its least data-driven decisions.

On 6 December 2026, the USGA and the R&A announced changes to ball testing conditions, capping flight distance in elite competition from 2028 and in recreational play from 2030. The PGA Tour issued a statement opposing it. One side relied on modelled ball flight under standard conditions; the other on field observation of driving distance rising steadily decade after decade. Both had data. The two datasets did not speak the same language.

With LIV the story is similar but sits one layer deeper: a 54-hole format with no cut skews the sample. No cut means no eliminated group midway through the event, which means every comparison with a 72-hole tournament is missing a control variable. The OWGR board did not say LIV was weak. They said they could not measure it.

That is the point most golf commentary skips. The absence of data is itself data, and it is always data about the source, never about the subject. A blank file does not say the player performed badly. It says the person who wrote the file never opened a spreadsheet.

Rory McIlroy is the cleanest example of misreading a data gap. Between the 2026 PGA Championship and the 2026 Masters he went ten years without a major. Commentators called it a psychological collapse, and they told that story through the missed putts at Pinehurst in 2026, where he finished second to Bryson DeChambeau by a single stroke. But the numbers across those ten years were never blank. He stayed inside the world's top group and kept appearing in the final groups of majors time after time. The problem was sample size: four majors a year, four rounds each, and one putt on the closing hole can flip the entire narrative.

Three years of tracking domestic tournaments and PGA Tour data taught me a habit: before reading the conclusion, count the tables in the file. If the conclusion is longer than the data, I close the file and note the date to reopen it. Data is never in a hurry; it waits for someone who can read it.

In the domestic market the gap is wider. An expanded amateur event can finish with nobody recording hole-by-hole scores, only a final leaderboard. When I ask for data to build a model, the answer is usually a three-column spreadsheet. Writing about those events demands a different discipline: stating plainly what I do not know, instead of filling the void with a story that sounds reasonable.

In December 2026, Jon Rahm left the PGA Tour for LIV Golf. The market read the move through the contract figure, and most pieces that week contained exactly one number. Nobody could measure the technical consequence, because no sample existed yet. Two years later, one does.

The comfortable illusion of a full-looking but hollow file is the biggest trap in this trade. A PDF with a table of contents, recommendations and a signature passes every internal review, because nobody opens the appendix. Being pushed outside the game is the fastest way to see the whole board: precisely because it was excluded from the ranking list, LIV was forced to publish its format data, and the argument shifted from "is this tour strong or weak" to "which ruler is valid". That is progress, even if nobody in that meeting room wanted to admit it.

In meetings I meet plenty of people who open with "I have twenty years in this business". Twenty years of observation is an asset. It is not a table of numbers. When someone tells a golf story without naming a sample, I ask one question back: compared to what. If the answer is "compared to a feeling", the conversation ends there, the same way it ended with that fifteen-page PDF.

Correlation is not causation, and in golf the temptation appears in every table. A player changes his grip and wins the following week; a player takes three weeks off and returns a champion. Those are two data points, not a rule. Crowds applaud on emotion, but data hears a different rhythm.

The signal for the next cycle already has a date: from 2028, the ball distance limit takes effect in elite competition. That will be the cleanest natural experiment golf analytics has had in two decades: one measurement system, one group of players, one variable changed. I write the report, close the file, and the market reopens on its own. When that happens, who was right will no longer depend on the loudest voice in the room.