ShotIQ Learn/📡 Launch Monitor Metrics

How Many Shots Before Your Average Means Anything?

You hit ten 7-irons. The launch monitor says 154 yards. Last month it said 148.

Six yards. You changed something two weeks ago, and here's the proof it worked.

Except it almost certainly isn't. With a typical iron spread, ten shots cannot detect a change smaller than about six yards — which means a six-yard "gain" is exactly the size of difference you'd expect to see from an identical swing on a different Tuesday.

This is the single most useful thing a launch monitor owner can understand, and almost no device or app will tell you, because "your average moved but we can't say it means anything" is a difficult sentence to put on a screen you paid for.

Why your average moves when your swing didn't

You don't hit one carry number. You hit a spread.

A decent amateur's 7-iron might average 150 yards with individual shots landing anywhere from 138 to 162. That spread — the standard deviation, usually ±6 to ±9 yards for an iron — isn't a flaw in the golfer or the device. It's what hitting a golf ball is.

Now take ten shots out of that spread and average them. Take another ten and average those. Those two averages will not be the same, and they were never going to be. The average of a sample bounces around the true value, and how much it bounces depends on exactly two things:

• How wide your spread is. More scatter, more bounce.
• How many shots you took. More shots, less bounce.

That bounce has a name — standard error — and the practical consequence is blunt. Before you can say a difference is real, it has to be bigger than the amount your average was going to move on its own.

The table

Here is how big a change has to be before ten, twenty or fifty shots can actually prove it. Read down to your shot count, across to your spread.

Smallest carry change that can be distinguished from normal variation at 95% confidence, comparing one session against a 40-shot history.

Anything smaller than the number in your box is inside the noise. Not "probably noise" — indistinguishable from it, on the evidence you have.

Which column is yours? If you don't know your spread, use these as a starting point: a wedge is often ±3 to ±5 yards, a mid-iron ±6 to ±9, a driver ±10 to ±14. Wider than that and the issue isn't your averages, it's your strike — see carry distance and smash factor.

The comparison you're actually making is worse

The table above compares one session against a long history. That's the generous case.

Most golfers compare last session to the session before — ten shots against ten shots. Both numbers are wobbling, so the wobbles add:

Ten against ten, on a normal mid-iron, needs 7.5 yards before it means anything. On a driver at ±12, you need over eleven yards — which is most of a club.

This is why comparing today to last Tuesday is the worst way to track progress, and it's the comparison every launch monitor app makes easiest.

So how many shots do you actually need?

Turning the question round, for a mid-iron at ±8 yards, against a 40-shot history:

• To prove a 10-yard change — about 5 shots.
• To prove a 5-yard change — about 15 shots.
• To prove a 3-yard change — about 97 shots.

That last one is not a typo, but it isn't the whole story either. It's high because the history is only 40 shots, and a comparison is limited by both sides of it. Grow the history and the session you need shrinks fast:

Which is the practical answer, and it isn't "hit more balls today". Small changes are detected by accumulating history, not by hitting a huge session. Forty more shots spread over four visits are worth more than forty more shots this afternoon — they build the thing every future session gets measured against.

What to do instead

Hit more shots per club, and fewer clubs per session. Ten shots with four clubs tells you almost nothing about any of them. Thirty shots with one club tells you something real.

Compare to your history, not to last time. Your baseline should be every session you've logged, not the most recent one. It's a wider net and a steadier number.

Watch your spread, not just your average. A session where your average held but your spread tightened from ±11 to ±7 is a genuinely better session, and averages alone will call it a draw. Consistency usually improves before distance does — if you only track carry, you'll miss the first evidence that something is working.

Be suspicious of a great session right after a bad one. If last week was unusually poor, this week will look better whether or not you did anything. Averages drift back toward normal on their own; it's called regression to the mean, and it has convinced a lot of golfers that a swing thought works.

Ignore your single best shot entirely. Your longest drive of the day is the far edge of your spread, not your distance. Playing to it is the most reliable way to come up short on a course.

Where the causes live

None of the above tells you why a real change happened — only whether it was real. When one is, the explanation is usually upstream of carry:

• Club head speed — effort, fatigue, how warm you were.
• Smash factor — where on the face you struck it. Same speed, different contact.
• Launch angle and spin rate — delivery and conditions.

If carry dropped and club speed didn't, it wasn't effort — it was strike. That distinction is the difference between resting and practising, and you can only make it once you know the change was real in the first place.

For the full picture of every number on your screen, start with the launch monitor metrics guide.

How ShotIQ handles this

ShotIQ runs this test on every club, automatically. Open any club and it compares your last session against your pooled history — excluding that session from its own baseline — and reports which metrics moved further than your own variation explains.

Each row carries its noise band, so you can see exactly what your shot counts could and couldn't prove. A change has to clear two bars before it's called real: statistically distinguishable from your normal variation, and big enough to notice on a course. Things that pass only the first are labelled too small to feel rather than sold to you as progress.

And when nothing moved, it says so. That's usually the honest answer, and it's worth more than a number that flatters you.

Frequently asked questions

How many shots should I hit per club on the range? For a session you intend to draw conclusions from, at least 15 with a single club, and 20 to 30 is better. Below about 8 shots, only very large changes — 7 yards or more on a mid-iron — can be distinguished from normal shot-to-shot variation.

Why does my launch monitor give different averages for the same club? Because you don't hit one number, you hit a spread. Individual shots vary by ±6 to ±9 yards on a typical iron, so any average built from a handful of them moves around too. The device is measuring correctly; the average is just a small sample of a wide distribution.

Is a 5-yard gain in carry distance real? It depends entirely on how many shots it's built on and how wide your spread is. Against a solid history, a mid-iron needs roughly 15 shots before a 5-yard change can be distinguished from noise. Comparing two sessions of 10 shots each, it cannot be.

Should I compare my last session to the one before it? It's the weakest comparison available, because both averages carry their own error and the two errors add. For the same session size the difference is substantial: ten shots measured against a 40-shot history can detect a 6.1-yard change, while ten shots against another ten needs 7.5 yards. Same session, a fifth less sensitive, purely because the thing you're comparing to is wobbling too.

What is a good carry distance standard deviation? As a rough guide, under ±5 yards for a wedge, under ±8 for a mid-iron and under ±12 for a driver would be considered consistent. Tightening that spread is usually worth more strokes than adding distance, and it improves first.

Does this mean my launch monitor is inaccurate? No. Device accuracy and sample size are different problems. A perfectly accurate launch monitor measuring ten real golf shots still produces an average that moves, because the shots themselves vary. This is about how many measurements you need, not how good they are.