This thread reminded me of an article in American Rifleman 20+ years ago. I had to go look it up.
A well known gunwriter tested a custom rifle in 280 AI built by a famous gun builder known for his long range rifles.
The question he asked was "How much accuracy should one demand in a big game rifle?"
The builder didn't let a gun out of the shop that shot over 1 MOA for 3 shots, and most shot 1/2 MOA he claimed. The test rifle averaged .358" for the builder and .400" for the gun writer.
But he tested the rifles overall consistency by using the same backer consisting of two separate groups. One had 82 shots in it that measured 3.189", the other had 89 shots in 3.375", showing that a 1/2 MOA group could be anywhere within that 3" group under reasonable conditions.


"I was born in the log cabin I helped my grandfather build"