The Three-Week Variable: What a Golf Scorecard Never Records

At the 15th hole of the final round, with the lead down to a single shot, his approach from 168 meters came up four meters short. The ball rolled down the slope behind the green, and he needed two more strokes to get back onto the putting surface. Nobody in the gallery remembers that shot. They remember the birdie putt on 16, and they remember that he lost. When I reopened the Strokes Gained table for that round, the miss on 15 was only the surface. Beneath it, a signal that had appeared on Friday was closing — nobody had read it.

I follow golf by one dry principle: if something leaves no trace in the data, it can still exist — we simply have not set the right axis to measure it. The bad shot on 15 did not begin on 15. It began with the three-week break before the tournament.

That is the hypothesis I have carried for three years, and the variable that makes most scorecards meaningless if you only read them top to bottom.

Strokes Gained does not measure a beautiful shot. It measures whether a shot was better or worse than the field average, converted into expected strokes. The metric was founded by Mark Broadie, a Columbia University professor, and it is now the backbone of all professional golf analysis. The PGA Tour's ShotLink system, running since 2026, records every shot to feed that metric. A 320-meter drive is not automatically good if the fairway is sixty meters wide. A putt from 1.2 meters is not automatically easy if the green slopes toward water. Strokes Gained separates the shot from the feeling, and that is why I trust it more than any commentary.

The standard table splits into four buckets: Off the Tee, Approach, Around the Green, Putting. Add the four together and you have a round. The trouble lies in that addition.

When you add the four, you lose the order of time. A round is a sequence, not a sum. The approach on 15 does not exist independently of the putt on 14, and it does not exist independently of the three weeks before.

Since I was 19, I have logged every dangerous situation by hand. That work taught me something software does not: the most important data column is usually the one absent from the table. In golf, that is the number of days since the last competitive round. I added it to my sheet as a fifth column, and it quickly became the first one I look at.

In my dataset — more than four hundred rounds across professional and domestic amateur events — there is a pattern that is anything but subtle. Rounds played fourteen to twenty-two days after the previous competitive appearance show a lower SG: Approach over the final nine holes than the same player's own average. Not over the first nine. Not in Putting. Only in Approach, and only in the back half of the round.

The specific figure in my sample: an average gap of about 0.4 strokes per round in back-nine SG: Approach, compared with when the same player competed within seven days. At the elite level, 0.4 strokes per round is the difference between a top-10 and a missed cut.

I do not rush to conclude. Data is never in a hurry; it only waits for someone who knows how to read it.

What makes me believe this pattern is not random is the way it appears. In the early stages of the round, the player coming off three weeks still hits good approaches. The decline only arrives after the 12th hole, when the body has accumulated fatigue and focus begins to depend on muscle memory.

That is the crux. Muscle memory operates as a data sequence reloaded through real competition, and three weeks is the point at which it starts to blur. A player can hit three hundred balls a day on the range, but the range has no pressure. The range has no gallery. The range has no 15th hole with a one-shot lead.

The approach on 15 is where the three-week break pays its bill.

I tested the hypothesis another way. If the problem were physical, the decline would spread to Off the Tee and Around the Green. It does not spread. If the problem were purely psychological, the decline would also appear in Putting under pressure. It does appear, but it does not correlate with days of rest — it correlates with crowd size. These are two different variables, and merging them is the most common mistake of scorecard readers.

On the crowd variable, I keep separate notes. In rounds with large galleries, the success rate of putts from 1.5 to 3 meters among the players I track falls by about four percentage points. The figure is not large. But it is stable across many events, and stability is what deserves trust, not magnitude.

The gallery applauds by emotion, but the data hears a different rhythm.

There is one more variable a scorecard never records: temperature. In my data, rounds played above 34 degrees Celsius show a marked rise in grip-control errors from the 14th hole onward, especially among players who tend to squeeze the club. This is why I always cross-check SG: Approach against temperature before concluding anything about technique. An approach missed in 36-degree heat tells a very different story from the same shot on a 22-degree morning.

I always cross-check in a simple way. If the three-week pattern is real, it must appear among both young and veteran players. It must appear among both fast-swing and slow-swing players. If it appears in only one group, I do not hold a variable about rest schedules — I hold a variable about a type of player, and those two things are entirely different.

In my dataset, the three-week pattern appears in both age groups, but more strongly above age 32. In the under-25 group, the gap is only about 0.15 strokes per round, close to the noise band. This detail weakens my own hypothesis, and I record it honestly.

At this point, I have to stop myself. A stable pattern is not necessarily causal. The three-week break may be only a proxy for something else: flight schedules, training volume, a swing change, or simply a personal stretch. I have no data on those. In my sheet, three weeks is just a number.

The danger of this profession lies there. The deeper you dig, the more easily you see patterns in places that hold only noise. I set myself a rule: a hidden variable earns the name only when it appears in at least three independent tournaments, with at least thirty observations each. If it falls short, it stays in the drawer.

A report sitting in a drawer is not a conclusion, but a graph waiting for its time axis.

There is a second temptation, more dangerous than the first. It is stretching a short phenomenon into a long trend. When a player has three hot putting rounds, the market instantly prices them as a putting specialist. Three rounds is about one hundred and twenty putts. Given putting's variance, one hundred and twenty putts cannot separate skill from luck. But the media does not wait for an adequate sample. They sell the story before the data has grown up.

I know the feeling of being pushed aside. At 19, when I presented an analysis and was told the young do not understand golf, I did not argue. I wrote the report, with charts, and let the numbers do the rest. Being pushed out of the game is the fastest way to see the whole board.

So what is the signal for the next round?

If the three-week pattern holds, tournaments following a long break should be read through back-nine SG: Approach rather than total score. A player leading after two rounds at a post-break event is not necessarily safe. And a player who finishes slowly at the 15th should be watched in the next event, rather than marked down in the current one.

I set a time limit on this judgment: six months. If, by the end of that cycle, the pattern has not repeated in at least two new events, I will close the file and wait for the market to reopen. I write the report, close the file, and then the market reopens on its own. That is my working rhythm, and it does not change with the result of a single round.

What I want readers to carry away does not lie in the 0.4-stroke figure. It is a question: which column on your scorecard is missing? Every dataset has a blank column, and that column is usually the decisive one. People watch the deciding putt. I watch the ball flight before that putt, and the rest schedule before the entire round.

Numbers know their own way of telling a story. My job is to read their time axis correctly.