How Data Action Golf Reads a Winner

Every tournament article on this site starts from one question: how did the winner actually win? Not who, not by how much, but where the strokes came from. The answer is built the same way every week, from the same data, with the same definitions, so that a reader can compare one week with another and know the comparison is fair. This page is the method. The tournament articles link here so they do not have to repeat it.

The data

Every strokes-gained figure on this site comes from Data Golf, a golf analytics service that computes strokes gained from shot-level PGA TOUR data. We take their round-by-round record for each PGA TOUR event: every player's score per round and his strokes gained in four categories, off the tee, approach, around the green and putting. Data Golf's numbers are close to, but not identical to, the official PGA TOUR figures. Their FAQ lists the reasons, including how they classify recovery shots, where they draw the line between approach and around-the-green shots (all shots within 50 yards of the hole count as around the green), and the fact that they centre each category on the field average per hole rather than per round. When a tournament article quotes a strokes-gained number, it is Data Golf's number, and it may differ by a few tenths from the figure on a broadcast graphic.

One more thing to know about coverage. Data Golf's own FAQ says that "only events that have the ShotLink system set up provide data on player performance in the strokes-gained categories," and that where round-level category data is missing it can be imputed from a regression model fitted on PGA TOUR data, with the caveat that imputed values "will show less variation than the true (unobserved to us) round-level data." The Winner Database takes whatever category figures Data Golf publishes for an event, including the majors, which are not run on the TOUR's ShotLink system. So a strokes-gained breakdown from the Masters, the U.S. Open or The Open Championship rests on Data Golf's own data collection or estimation for that week rather than on the TOUR's official shot tracking, and it deserves the same "a few tenths either way" caution as any other number here. Where Data Golf publishes no category data for an event, this site publishes no breakdown.

Some events have no strokes-gained record at all. Team events like the Zurich Classic of New Orleans and certain events played without the TOUR's shot-tracking system fall out of the database for that reason, and there is no tournament article for them. We do not estimate what the numbers would have been.

If strokes gained itself is new to you, start with what strokes gained measures.

The Winner's Five

For each winner we look at five things.

1. Off the tee. Strokes gained on tee shots at par 4s and par 5s, summed across the four rounds.
2. Approach. Strokes gained on shots into the green from outside Data Golf's around-the-green range.
3. Around the green. Strokes gained on shots inside that range, other than putts. The tournament articles also call this category short game.
4. Putting. Strokes gained on the green.
5. Scoring conversion. Not a strokes-gained category but a counting panel: birdies or better, bogeys against the field's average bogey count, double bogeys or worse, the final-round score and its strokes gained, and the winner's position after 54 holes.

The first four are summed over the winner's four rounds. Because Data Golf states each category against the field's average for that round, a winner's total in a category is his advantage over the field-average player in that part of the game for the week. A positive number means strokes better than the field. We never describe strokes as taken from the field, because that is not what the number measures.

Field rank

Each category also carries a rank: where the winner's four-round total sat among the players who completed all four rounds. That population is the field for every rank on this site, and the tournament articles state its size. It is not the full starting field, because players who missed the cut do not have four rounds to sum.

Share of advantage

This is the Data Action Golf reading, and it needs a precise definition because it does most of the work in the articles.

Take the winner's four category totals. Keep the ones that are positive. Add them up. Each positive category's share is its total divided by that sum. A category where the winner lost strokes to the field reads 0%, because you cannot claim a share of an advantage from a category that was a disadvantage.

Two consequences follow. First, when every category is positive, the shares describe how the winner's whole strokes-gained total was distributed. When one category is negative, the shares describe only the positive side, and his total is smaller than the sum the shares are drawn from. The articles say so whenever that happens. Second, shares are rounded to whole numbers, so the four can add to 99 or 101.

Every tournament article compares the winner's shares with the season winner average share: the mean share in each category across the earlier winners of that season who have a strokes-gained record. That average moves as the season goes on, and the articles quote it as it stood on the week of the event, not as it stood at the end of the year. It tells you what a typical winner's distribution looked like to that point, so you can see when a week was lopsided.

The winning-formula labels

Each win gets a label describing how the advantage was distributed. The labels are assigned by fixed rules, checked in this order, and the first rule that fits is the label.

The order matters. A comeback is checked first because the comeback is the story of that week regardless of where the strokes came from. A win with a clean scorecard is checked next. Only then do the share-based rules apply. The thresholds are ours, they are round numbers, and a win that sits one point either side of a threshold could reasonably carry the neighbouring label. Read the labels as a shorthand for the table, not as a verdict on the golfer.

The season comparison

Alongside the winner's own numbers, each article reports the average winning margin and average bogey count across the season's earlier winners with a strokes-gained record, and the running count of each label. These are simple means over a small and growing sample. In the 2025 season the database holds 40 winners with a strokes-gained record; in 2026 it holds 34 through the BMW Championship in August. A season average built from six winners in February is a rougher yardstick than one built from thirty in August, and the articles carry the count so you can weigh it.

What the numbers can and cannot say

They describe. They do not predict. A share of advantage says where one week's edge came from. It does not say the same distribution will win next week, or that a player who fits the pattern will contend. Course and field change, and a sample of one is a sample of one.

They attribute strokes, not causes. When a winner gains six strokes putting and wins by one, it is true that putting supplied the largest share of his advantage. It is not true that putting is why he won, because with a one-stroke margin every category was decisive. The articles keep to the first kind of sentence.

They are one measurement of many. Strokes gained has no opinion about wind, pin positions, pairings or nerve. The scoring-conversion panel adds the scorecard, but there is no shot-sequence data behind it, so an article cannot say that one mistake led to another, only how many mistakes there were.

Small samples are small. Three wins by one player in a season is a pattern worth noticing, and the articles notice it. It is not a rule about that player.

The research is repeatable. Every figure in a tournament article comes from the same database and the same arithmetic described here. Where an article adds a fact from outside the database, a result, a playoff opponent, a piece of history, it links the source in the same paragraph.

Where the articles come from

The tournament articles began as research for Data Monday, the weekly segment of the Data Action Golf podcast, where the question is the same one: how did he win, and what does it teach the rest of us? The written versions stand on their own. Each one ends with the same amateur translation, because the scoring principle behind a professional win, manageable next shots and fewer recovery shots, is the part that transfers to a Saturday morning, even when the swing does not.

Better data always means better golf. This page is the part that keeps the data honest.