Three shot groups are nonsense

[Hsing-ee]

On a large proportion of YouTube rifle reviews the tester uses a three-shot group to demonstrate accuracy. Does anyone else think this is knackers? In Walter's book there was an interesting piece on calculating the true centre of a group and it needed 25 shots to give statistical significance.

While I think that 25 would be tedious to watch and also quite hard to do in a string for the shooter, I think that the minimum number of shots for assessing accuracy ought to be 10 or 12, even if two different aiming marks are used.

What d'yall think?

[Jonjon79]

If I'm testing groups on a new rifle, my preference is 10 shot groups.
After that, to reassure myself of pellet choice, I put a bunch of marks on paper with a marker pen and take 1 shot at each.

[Weasle]

If I'm testing groups on a new rifle, my preference is 10 shot groups.
After that, to reassure myself of pellet choice, I put a bunch of marks on paper with a marker pen and take 1 shot at each.

Agreed...

[JustinThyme]

5 shot groups when deciding on pellet choice ..... three not a real indicator due to operator error and the occasional flier and ten a waste when five should provide enough info

[Rickenbacker]

A group should be ten. By law.

[bozzer]

Three shot groups are plenty.

If you shoot more than three shots and any open up the group ... then just ignore the ones that open up the group and measure the best three ... and that's how accurate you can shoot a group.

[Blackrider]

3 shots with centrefire, any more than that gets expensive.
5 with rimfire and 10 with air.

[Hsing-ee]

Three shot groups are plenty.

If you shoot more than three shots and any open up the group ... then just ignore the ones that open up the group and measure the best three ... and that's how accurate you can shoot a group.

Why waste ammo? Just shoot one pellet. Lets all be in the next Olympic team!

[rhyslightnin]

On a large proportion of YouTube rifle reviews the tester uses a three-shot group to demonstrate accuracy. Does anyone else think this is knackers? In Walter's book there was an interesting piece on calculating the true centre of a group and it needed 25 shots to give statistical significance.

While I think that 25 would be tedious to watch and also quite hard to do in a string for the shooter, I think that the minimum number of shots for assessing accuracy ought to be 10 or 12, even if two different aiming marks are used.

What d'yall think?

No., Not nonsense..... Let me explain...
You know a group is pants of the first 3 are all over the place & you have done your bit.. and in a similar guise you have to do the first 3 of any 10 shot group well to make a good 10 shot group

[lensman57]

On a large proportion of YouTube rifle reviews the tester uses a three-shot group to demonstrate accura;cy. Does anyone else think this is knackers? In Walter's book there was an interesting piece on calculating the true centre of a group and it needed 25 shots to give statistical significance.

While I think that 25 would be tedious to watch and also quite hard to do in a string for the shooter, I think that the minimum number of shots for assessing accuracy ought to be 10 or 12, even if two different aiming marks are used.

What d'yall think?

When we were at UNI we were told that a minimum of 7 sample points are required to make a valid statistical graph in our measurements. I guess that you could apply the same logic to our pellet testing.

A.G