Strokes Gained: How Data Rewrites the Way We Read a Round of Golf

Two golfers card the same 68 in a round. The leaderboard places them side by side, one line, one number. But when ShotLink data is fed into a model, the real gap between them stretches to nearly three strokes, roughly a quarter of a round. One reached 68 on hot putting and a couple of lucky chips; the other reached 68 by repeatedly pinning approach shots from about 150 metres. The same number, two entirely different fates.

That is why I always open with Strokes Gained instead of the raw score. Numbers do not lie. But reputation whispers into the ear of anyone who does not read the table.

How Strokes Gained Came to Be

Strokes Gained (SG) grew out of research by Mark Broadie, a professor at Columbia University, published around 2026. The core idea is simple but transformed the sport: each shot is not measured by its raw outcome, but by the change in the expected number of strokes remaining before and after that shot, relative to the tour average. The PGA Tour began publishing official SG figures from the 2026 season, powered by ShotLink, a network of measuring devices and volunteers that records every shot at every event.

SG splits into four main categories: Off the Tee, Approach, Around the Green and Putting. Add the four together and you get a picture of where a golfer wins or loses. This is the foundation of nearly all modern professional golf analysis, and it is also the line separating those who read data from those who read the leaderboard.

I have followed professional golf since my student days in communications in Binh Duong, writing down every round of the majors by hand. That experience taught me something small but foundational: the leaderboard answers the question "how many", while Strokes Gained answers "how".

Contextualising Every Metric

When I read golf data, the first thing I do is contextualise. An SG: Approach figure of plus 1.8 strokes per round sounds very strong, but you have to ask immediately: on what surface, in what weather, and most importantly, where were the pins set? The same golfer, the same shot, posting that figure on a windy links course is entirely different from posting it on soft Bermuda grass with slow greens.

There is more. An impressive figure only means something next to its sample size. Putting is the most deceptive category in golf. A golfer can post plus 2.5 SG: Putting over two rounds and then return to average over the next two. If I cherry-pick the first two rounds to draw a conclusion, I am selling readers a pretty but false story. I always state the sample size, the assumptions and the variables I cannot control before I let a number speak.

At the level of a full season, the picture sharpens. Studies of the age curve in golf show that most professionals peak between 30 and 35, with driving skill peaking earlier and approach and putting stabilising later. That means a 24-year-old leading in driving distance is not necessarily a major contender, while a 33-year-old with a stable SG: Approach across multiple seasons is a far more reliable candidate profile.

The Blind Spot of Reading a Single Metric

This is where I separate myself from shallow analysis. Correlation does not mean causation. A golfer with a high SG: Putting is not necessarily the best putter; he may simply keep leaving the ball close to the pin through outstanding approach play, making every putt easier and letting the putting figure benefit indirectly. If you only look at the putting table to praise him, you have missed the real cause in the approach game.

I have seen this while working with data. A golfer the media calls a putting master was in fact only holding that figure thanks to the quality of chances created before. Split the categories apart and the picture flips entirely. Reputation says one thing, data says another, and every time, I choose the data.

At the macro level, the story gets even more complex. The split between the PGA Tour and LIV Golf since 2026 created a variable never seen before in golf history: two parallel competition systems, with two scoring structures and two different consequences for a golfer's career. LIV Golf launched with financial backing from Saudi Arabia's public investment fund. The Official World Golf Ranking (OWGR) initially awarded no points to LIV events, leaving many golfers who left the tour facing the risk of falling in the rankings and losing major exemptions. In June 2026, the PGA Tour and the Saudi fund announced a framework agreement to merge commercial interests, but negotiations dragged on and remain unfinished.

The Counter-Intuitive Angle

The counter-intuitive part is that in this uncertain environment, data matters more, not less. When a golfer moves to a new competition system, all of his old data loses comparative value: different courses, different fields, different schedule density. Anyone using old numbers to predict new form is reading an outdated map.

I hate uncertainty. But 2026 taught me that an unforeseen variable can be stronger than any algorithm. When courses closed because of the pandemic, home advantage vanished, and models built on crowd data became meaningless overnight. That lesson applies directly to golf: every model has a breaking point, and a good analyst prepares a Plan B before the breaking point appears.

For Vietnamese golf, the data gap is far wider. Domestic tournaments still lack a shot-by-shot system like ShotLink, so most analysis stops at the leaderboard. This is both a limitation and an opportunity: whoever builds the data infrastructure first will hold the edge in spotting young talent.

Takeaway

I do not predict. I read the data and accept the consequences. Next season, the signal I track is not who posts the most under-par strokes in a single round, but who holds a stable SG: Approach across different courses. A single metric can fool you for a week. A large enough sample, properly contextualised, cannot.