Strokes Gained and the Three Blind Spots of Golf Data

One April afternoon, I sat behind the 16th green, notebook open, recording a golfer who misread the slope three times. The tablet beside me showed his Strokes Gained: Putting running nearly 1.8 strokes better than the tour average per round. He still holed the putts. The green still fed the ball toward the cup. And the number still looked immaculate, as if nothing had happened. I wrote in the margin of my notebook: the ball rolls across the course, but what people are reading is the flow of money moving behind it. Fifteen years ago, Mark Broadie, a professor at Columbia Business School, published the Strokes Gained method, and golf entered a new era. Before that, a golfer was judged by GIR, the rate of greens hit in regulation, by par-saving percentage, by fairways found. Those metrics shared one feature: they counted events, not value. Broadie showed that a putt from 1.2 metres and a putt from 9 metres, if both drop, are recorded identically in traditional statistics. That is like scoring a three-pointer and a layup in basketball with the same ruler. Strokes Gained changed that. It splits the game into four areas — off the tee, approach, around the green and putting — then compares each of a golfer's shots against the tour average under the same conditions. The primary data source is ShotLink, the shot-tracking system the PGA Tour operates at most of its venues. Since then, every contract, every sponsor exemption, every television breakdown has run, to some degree, through those four numbers. But this is where I began to doubt. Across the last three seasons covering weekend pairings, I noticed a repeating pattern. Golfers with a sudden spike in Strokes Gained: Putting rarely sustain it beyond six weeks. The reason is simple and deeply inconvenient: putting is the category most exposed to luck. A five-metre putt that rolls to the edge of the cup and drops, and a five-metre putt that rolls to the edge of the cup and stops, produce two entirely different figures on the Strokes Gained table, even though the strokes were nearly identical. At a one-season sample size, that gap is enough to turn an average golfer into a putting specialist in the public eye, and enough to underrate a genuinely fine putter. I once tracked such a golfer through an entire season. He ranked 12th on tour in Strokes Gained: Approach, meaning his approach play was the twelfth best among more than two hundred players. But he putted poorly, and the end-of-season standings never mentioned him. Nobody wrote about him. Because most writers read the total column, not the component columns. That is the first blind spot. The second blind spot sits inside the name itself: Strokes Gained measures a shot against the tour average, but it knows nothing about the player's intent. A golfer who deliberately plays to the middle of the green to avoid a pin tucked near a hazard earns a very low approach score, while a reckless golfer firing straight at the flag earns a very high one — even when the second decision was right and the first was wrong. Strokes Gained cannot tell those two apart. It only knows where the ball stopped, not what the player was thinking. The third blind spot concerns course context. The same Strokes Gained figure means something entirely different on a coastal links course in strong wind and on a parkland course with soft greens. Yet media breakdowns routinely place two numbers side by side as if they shared a unit of measurement. I have seen articles compare one golfer's Strokes Gained: Approach against another's from a different event, on a different course, in different weather, and then draw conclusions about form. The only way to read Strokes Gained correctly is to read it event by event, course by course, always accompanied by one question: how many shots are in this sample? Below forty shots, the number is nearly meaningless. A round contains roughly seventy shots, of which fewer than thirty fall into any single category. Yet people still talk about a golfer's form after two rounds. A season is only one sentence in a book a decade thick. Now let us step off the course, because the golf data story does not end at Strokes Gained. In October 2026, the Official World Golf Ranking board rejected LIV Golf's application for ranking points. That decision did not come from a technical failure but from a difference in model: OWGR members argued that LIV operates a team format, has no cut, and — most importantly — lacks enough turnover to guarantee a representative sample. This is an entirely valid statistical argument. But it is also a political one, because if LIV held OWGR points, the majors would have to open their doors to golfers who had signed with Saudi Arabia's Public Investment Fund. The result is a paradox: Jon Rahm, who left the PGA Tour for LIV in December 2026 on a deal international media valued at roughly half a billion dollars, retained his place among the world's top-ranked golfers for months afterwards — not through points from his new tour, but through frozen old ones. Data is never neutral. It is only neutral to those who do not know who set the measurement. Around the same period, the USGA and the R&A announced a ball specification change in December 2026, under which balls will fly shorter at high swing speeds, applying to professional play from 2028 and to amateurs from 2030. This was a data-driven decision: average driving distance among professionals had risen continuously for two decades, and traditional courses had been stretched to the point of running out of land. But it was also a fiercely contested one, because it changes the sport itself on numbers most amateur players cannot verify. Here the writer's role becomes clearer than ever. They doubt the voice before hearing the argument. I learned to gather evidence first and expect afterwards. For every figure I cite, I must be able to answer three questions: where it came from, how many shots it contains, and who decided how it would be measured. In fifteen years of writing about golf, I have never seen a season in which data reduced the argument. It only moved the argument from the green to the spreadsheet. What I take away is not that golf data is useless. On the contrary, it is the best tool this industry has, and I use it every week. What I take away is that data has no meaning when detached from the context that produced it. A number without provenance, without sample size, without definition, is just an unverifiable sentence. The season is still running. The standings still update every Monday. And somewhere next week, a golfer will misread the slope three times in a row, hole all three putts, and the press will call it form.