ShotLink Stops at Ponte Vedra: The Data Gap Reshaping Asian Golf

A 168-yard approach from the rough, hole 14, final round of a JGTO event in Aichi. The ball stops four metres from the pin. The scorecard reads "par". On the PGA Tour's ShotLink system, that same shot would be logged as a chain of fields: distance to the pin before the shot, lie type, distance remaining, lateral deviation, and the strokes-gained value of the shot against the tour baseline. In Japan, that shot exists as exactly one line: a number of strokes.

The distance between those two records is not a matter of golfer quality. It sits in measurement infrastructure — the thing that decides who gets to tell a golfer's story, and in what language. After years in sports data analysis, I have concluded that most "who is better" arguments in golf are arguments about which side owns more data fields.

Strokes Gained came out of academia, not a tour's marketing department. In 2026, Mark Broadie — a Columbia Business School professor and amateur golfer — published research in the journal Interfaces that reconstructed the value of every shot by comparing it against the tour-wide distribution of outcomes from the same distance and the same lie. Three years later he systematised the method in Every Shot Counts.

The core idea fits in one sentence: a three-metre putt is not worth the same in every situation; it is worth the difference in expected strokes remaining before and after the stroke. Every shot collapses into one common unit, and four skills — driving, approach, short game, putting — become directly comparable. Before Strokes Gained, golf measured skill with counting stats: fairways hit, GIR, putts per round. Those count outcomes, not difficulty.

The method needs raw material. To compute strokes gained you must know where each shot started, what lie it sat in, and where it finished. The PGA Tour has ShotLink: a laser-based system installed at every hole, paired with volunteer spotters recording ball positions, operating since 2026. ShotLink turns each round into a shot-level dataset. No ShotLink, no strokes gained.

The boundary of ShotLink becomes the boundary of golf's knowledge market.

From the official statistical portals I have been able to collect, shot-level coverage currently splits into three tiers. The full tier is the PGA Tour, where nearly every event carries shot-by-shot data. The middle tier is the DP World Tour and the LPGA, where data exists but unevenly across events and seasons. The near-empty tier is the JGTO, JLPGA, Korean Tour, Asian Tour and most regional circuits: public leaderboards, event-level aggregate statistics, and almost no shot-level dataset from which to compute strokes gained.

That stratification produces three knock-on effects.

The first is linguistic. Strokes gained leaderboards have become the default vocabulary for describing golfer quality: team selectors use it, sponsorship contracts use it, academies use it to design training programmes. A golfer without strokes gained has almost no voice in that argument, however well he strikes the ball.

The second is about valuation. Golfers outside the coverage zone get assessed with crude proxies: scoring average, win count, driving distance. All three carry systematic error. Scoring average depends on course quality and wind conditions week to week — a season played on tight, windy layouts will wreck the number of a strong approach player. Win count depends on how many invitations a player receives, which depends on relationships and budget. Driving distance is measurable but barely correlates with scoring at the elite level.

What did NOT happen usually tells the truth more clearly than what did. No shot-level dataset means we do not know how many strokes a JGTO player loses on approach against the baseline, and we do not know how much he recovers around the green. The market is forced to fill that gap with intuition. Intuition tends to reward a hot putting season and punish a good approach season with a cold putter — two errors pointing in opposite directions, both of them systematic.

I have made exactly this mistake. In 2026, building a manual xG model from video for a J.League 2 club, I omitted the home-venue factor and got 6 of the final 10 matchdays wrong. In 2026 I read the PPDA from a World Cup round-of-16 tie and drew conclusions about pressing without any data on distance covered in 15-minute blocks. Neither model failed for lack of ambition. They failed because of a missing variable I did not know I was missing.

The third effect is about policy. On 6 December 2026, the USGA and the R&A announced a Model Local Rule limiting ball flight distance, effective for elite golf from January 2028 and for recreational golf from January 2030. It is the biggest equipment change in decades, and the only way to know what it actually does is to compare shot-level data before and after, across tours, course conditions and age groups. On tours without shot-level data, the reform's impact will be recorded as impression. A Japanese golfer loses eight yards off the driver after 2028 — nobody can prove it, and nobody can disprove it.

The same logic repeats at governance level. On 11 October 2026, the Official World Golf Ranking board rejected LIV Golf's application for ranking points, citing technical grounds around the 54-hole format, the absence of a cut, and pathways into the league. Whether that decision was right is a separate question. What gets less attention is this: OWGR is a model built on results data, and the quality of that model's inputs is uneven across tours. A system used to classify all of professional golf operates on datasets with different levels of coverage.

Data is never wrong; I simply asked the wrong question. The question "who is better" cannot be answered with an incomplete dataset. The answerable question is: given the data that exists, which conclusions hold and which are merely inference.

I have to state my own limits here. I have no access to internal JGTO or JLPGA data, and I cannot rule out that these tours collect shot-level information privately without publishing it. My conclusion applies only to public data — the portion an outside analyst, a journalist, or a scout can actually reach. That limit matters, because most valuation decisions in the market are made by people holding only public data.

Vietnamese golf sits exactly at that junction. Course counts and tournament counts have grown for years, but shot-level data infrastructure is close to zero. Without building it from the start, the path will resemble the JGTO's: exporting golfers, importing ways to evaluate golfers. A Vietnamese player may have a strokes gained approach figure above the average of a major tour, and nobody will know — including him.

Hideki Matsuyama is the clearest counter-example. He is the most thoroughly measured Japanese golfer in the country's history, simply because most of his career has been played on the PGA Tour, where ShotLink operates at every event. Same golf nation, same coaching system, two entirely different levels of data visibility. The difference is not in the swing. It is in who pays for the measuring system.

There is a counter-argument worth taking seriously, and I have put it to myself many times: more data does not automatically produce better decisions. The PGA Tour owns the richest dataset in golf and still misprices players with regularity. If data were the answer, that error would not exist.

The second counter-argument is more conservative: in Japan, a coaching culture built on the coach's eye and the caddie's memory produced world-class golfers for decades. That is real knowledge, not noise. I do not dismiss it. But I cannot forget the 2026 lesson either: an eye without a comparison baseline cannot separate luck from skill, and it will repeat its mistakes with confidence.

The two need separating. The correlation between good approach play and low scoring was established on one tour's data, against one baseline. Extrapolating it to another tour, in different course conditions, is an unverified leap — and I have seen enough of those leaps fail not to assume it holds. When data hides its face, error becomes the guide, and that guide needs to be named in every conclusion.

The signal to watch in the next cycle is not a trophy. It is a line in a contract: whether any regional tour shares shot-level infrastructure, or at minimum publishes raw data in an open format. If that happens, the conversation about Asian golfers changes language within two seasons. If it does not, we will keep describing, very elegantly, the things we cannot measure.