The Evolution of Golf Statistics: A Short History
For most of its existence golf measured itself with one number per round. It took until 2003 for anyone to record where the ball actually went, and until the 2010s for that idea to reach club golfers.
Golf spent roughly four centuries measuring itself with a single number. You played eighteen holes, you added up the strokes, and that total was the entire statistical record of your afternoon. The golf statistics history that followed is really the story of the sport slowly admitting that one number cannot tell you why you shot it.
The interesting part is how recent most of it is. Fairways hit did not become an official statistic until 1980. Nobody recorded where a tour player’s ball actually landed until 2003. Strokes gained, the stat that reorganised how the sport thinks about performance, has been official on the PGA Tour since 2011.
The scorecard era: one number for eighteen holes
The scorecard is old. Stroke play as we know it goes back to the eighteenth century, and by the late 1800s clubs kept records of medal rounds. What they kept was the score, the date and the name. Nothing else.
That was not carelessness. Recording anything more required someone to stand there and write it down, and the categories to write down had not been invented yet. Golf instruction ran on eyesight and anecdote, which worked reasonably well for the swing and disastrously badly for strategy. If the best players said the short game was where scoring happened, that was the received wisdom, and there was no dataset to argue with.
The scorecard did leave one lasting inheritance: par. Standardising an expected score per hole gave every round a reference point, and it is the ancestor of every benchmark that came after. Expected strokes from 150 yards is the same idea with more decimal places.
1968: the first people to actually measure
The first serious attempt to study golf scientifically came out of Britain. The Search for the Perfect Swing, by Alastair Cochran and John Stobbs, was published in 1968 after several years of work funded by the Golf Society of Great Britain. Physicists and engineers put golfers in front of cameras and force plates and produced the first real numbers on clubhead speed, launch conditions and ball flight.
It was serious work, and it was aimed entirely at the swing. The book could tell you a great deal about what happens in the 0.0005 seconds of impact and almost nothing about whether you should lay up. Golf’s measurement problem was split in two from the start: biomechanics got the science, scoring got the folklore.
1980: the tour starts counting
The PGA Tour began keeping official statistics in 1980. The categories will be familiar because they still appear on every amateur scorecard app: driving distance, driving accuracy, greens in regulation, putts per round, sand saves, scoring average.
These are counting stats. They record whether an event happened, not how well it went, and that is where they fall apart:
- Putts per round rewards missing greens. Chip it to two feet and you have one putt; hit the green to 40 feet and you probably have two. The better shot produces the worse statistic.
- Fairways hit treats a drive that finishes 320 yards in the first cut the same as one that finishes 220 yards in the trees. Both are misses.
- Greens in regulation treats a 45-foot birdie putt and a three-footer as identical successes.
Golfers noticed the contradictions and mostly shrugged. The stats were the stats. Why fairways hit can mislead you goes through the worst of it, and the pattern is always the same: a yes-or-no question asked about a game played in yards and feet.
2003: golf gets coordinates
ShotLink changed the raw material. When the PGA Tour rolled the laser-based system out across its events in 2003, every shot at a tour event became a position rather than an outcome: this ball, this lie, this distance from the hole, and where it finished.
Millions of shots a year, with the context attached. That is the point at which golf stopped being a sport with statistics and became a sport with data, and the consequences took another decade to work through. ShotLink and how PGA Tour data changed golf covers the system itself in more detail.
2008 to 2016: from counting events to valuing shots
Mark Broadie, a Columbia Business School professor, had been building an amateur shot database called GolfMetrics since the early 2000s, mostly because he wanted to know why his own scores were what they were. His 2008 paper, Assessing Golfer Performance Using Golfmetrics, introduced the idea he then called shot value: compare every shot to the average number of strokes a reference golfer needs to hole out from the same spot, and the difference is what the shot was worth.
A 20-foot putt is not a success or a failure. It is worth about 1.87 expected strokes before you hit it. Hole it and you gained 0.87 strokes; two-putt and you lost 0.13. Every shot in the round gets a number, and the numbers add up to your score.
The timeline from there:
| Year | What happened |
|---|---|
| 2008 | Broadie publishes the GolfMetrics paper defining shot value |
| 2011 | The PGA Tour adopts strokes gained putting as an official statistic |
| 2014 | Every Shot Counts takes the method to a general audience |
| 2016 | The tour adds strokes gained off the tee, approach and around the green |
Every Shot Counts is where the conclusions became public property, and several of them contradicted a century of instruction. Driving and approach play accounted for roughly two thirds of the scoring difference between skill levels. Putting mattered far less than the “drive for show, putt for dough” era had assumed. The gap between a 90s golfer and a 70s golfer was mostly in the long game, not around the green.
The 2010s: statistics leave the tour
Everything above applied to about 200 professionals. Sensor systems arrived mid-decade to close the gap, with Game Golf in 2013 and Arccos the following year putting a chip in the grip of every club so that a round produced a shot-by-shot record automatically.
Phones did the rest. A GPS-equipped device in every pocket meant the position data ShotLink needed volunteers and laser rangefinders to collect could be captured, roughly, by the golfer. The precision is lower. The usefulness is not, because the questions an amateur needs answered are coarse ones: is my approach play or my putting costing me more, and from what distances.
Handicapping modernised on a similar arc. The slope system spread through the 1980s and 90s so that a score could be adjusted for course difficulty, and the World Handicap System unified the competing national systems in 2020. That solved portability, not diagnosis. Your handicap index still tells you how well you score, not where the strokes go.
What each era could actually answer
| Era | What was measured | Question it could answer | Blind spot |
|---|---|---|---|
| Pre-1980 | Total score | Did I play well? | Everything else |
| 1980–2003 | Counting stats | Did the event happen? | Quality of each shot |
| 2003–2011 | Shot positions | Where does the ball go? | What a position is worth |
| 2011 onward | Strokes gained | What did each shot cost or earn? | Nothing structural, but needs volume |
The direction never changes: each step moved from asking whether something happened to asking what it was worth.
Where this leaves your game
The practical consequence of 140 years of measurement is that a 15-handicap with a phone now has better diagnostic information than a tour player had in 1995. Not more data, better organised data, which is the difference that matters.
Two things follow from the history. First, benchmarks are the entire mechanism, so choosing a sensible one matters more than most golfers realise. Comparing yourself to a tour baseline when you play off 18 produces a column of negative numbers and no information, which is why comparing to a 10-handicap is often more useful.
Second, statistics have always needed volume before they mean anything. Broadie’s conclusions came from hundreds of thousands of shots. Yours will need 10 to 20 rounds before the noise settles, and no amount of measurement precision shortens that.
The sport took a century to stop counting outcomes and start valuing shots. Running your own rounds through the strokes gained calculator takes about two minutes, which is a reasonable return on the wait.
