Think about it for a moment. 0.003 of an inch variation is almost nothing. It's quite literally less than the average thickness of a human hair.
You also fired it in two different rifles.
And yes, any variance in methodology could be sufficient for this small variance.
