When I brought up the subject of three shot target I said
I love the folks that show me a 3 shot target (their best of course) and try to convince me that their gun is all that and a bag of chips too.
If you were to show me 5 or more we now start to have a valid data set to work with. Let us assume you claim all targets were fired in the same session. If you tell me that you are holding you crosshairs dead center on the target for each group, the targets were at the same distance, and some groups are 3 o'clock from center, some at 11 o'clock from center, etc. then I count the aggregate grouping of all targets based on target dead center. Now the extreme edges of the 3 o'clock and 11 o'clock, etc. become my measuring points.
If I overlay each target onto a master and mark the groups so I have all 15+ shots on one piece of paper, we begin to see the bigger picture. Singular three shot groups may make MOA but the aggregate may fall well short.
This goes to what several have commented on earlier. There is a big
difference in the gun itself being able to be sub. MOA consistently and the shooter behind the gun being sub MOA consistently.