The six percent rule

R3 bet5 Sept 2026

The six percent rule

Jeremy Paul is tied second at 11 under and rated 65th of the 70 players left. Across 14,575 player-events, almost none of a hot 36 holes carries into Saturday — which makes his round-3 line the biggest edge on the board.

14,575 player-events in the reversion sample150 rounds in his baseline221 tournaments

65th

Where he is rated

Of the 70 players who made the cut. He has spent two days as the second-best player in the field.

How much of it carries

The share of a hot 36 holes that shows up again in round 3, across every tournament we hold.

25/36

Greens hit

69.4% against a field average of 69.6%. He has been, precisely, an average ball-striker.

01 — THE SETUPTwo rounds that don’t belong to him

65–64 is a serious 36 holes. At Crans it has Jeremy Paul tied second, one behind Thriston Lawrence, and in the final group on Saturday afternoon.

It is also, by our reckoning, about eleven strokes better than he should have played. Paul comes into this week rated 65th of the 70 players who survived the cut — a number built on 150 rounds, so it is not a small-sample artefact. He has been the second-best player in the field for two days and roughly the sixth-worst for three years.

That gap is the bet. Not because he is a bad player — he is a tour professional having a wonderful week — but because the market is pricing Saturday as though the two days that just happened tell you more about tomorrow than the three years that preceded them. They do not. They tell you about six percent of it.

Every player-event in our database where a player had at least 40 rounds of history elsewhere. Horizontal axis: how far above his own established level he played over the first 36 holes. Vertical axis: how far above that same level he then played in round 3. If form carried, the green line would track the orange one. It is flat.

02 — THE RULESix percent carries. Ninety-four doesn’t.

Run the regression properly — predict a player’s round-3 score from two things, what he has just shot over 36 holes and what he has done over his career — and the two coefficients are not close:

Predicting round 3 — 14,575 player-events
Input	Weight	Std. error	t
What he shot in rounds 1–2	+0.046	0.015	3.0
His long-run baseline	+0.739	0.040	18.5

Both are real — the hot hand is not nothing, and a 3.0 t-statistic is a genuine signal. But the split is roughly 6% this week, 94% career. A player running a stroke a round above his level projects to keep about five hundredths of it.

Of the 358 players in our sample who arrived at Saturday more than four and a half strokes a round above their own level, the average round 3 was — to the tenth — exactly their own level. Not a fraction of the form. None of it.

Forty-seven percent of them actually shot worse than their career baseline on Saturday. Paul sits beyond even that group, at +5.7 strokes a round.

Which raises the obvious question: if none of this is real, what is he actually going to shoot on Saturday afternoon — and what is that worth?

Keep reading

The rest of this one is free too.

It just costs an email address. Drop it in and the projection, the price and the staking appear right here.

The projection, and the three numbers it is built from
Why the last tee time on Saturday is worth two thirds of a stroke
Our price, the edge against the market, and how we are staking it
The two ways this bet loses

One email address, no account, no card. We send the week’s analysis and nothing else — unsubscribe whenever you like.