Up until now, I have been shooting a minimum of ten 5 shot groups (50 shots total) for each data point in a data set. So if I am testing accuracy of a pellet at 10 different velocities, I am shooting 500 shots to collect the data. I am taking this great number of shots so that I can try to tease out the real data from errors causes by changes in wind condtions and shooter error.
However, I am currently thinking that 3 shot groups would be a better way to collect precision data. Rather than collecting 10 data points with 50 shots using 5 shot groups, I would be able to collect 16 data points with 48 shots using 3 shot groups. The reason I think that this would be better is 2 fold:
1. The odds of the subtle wind changes between a 3 shot group would be less than during a longer to complete 5 shot group. Therefore 3 shots would reduce error.
2. Any changes in the wind or shooter error that cause a flier (oulier) would have less effect on a data set of 16 than a data set of 10.
Since the consensus here seems to be that 5 shots or more is best for precision data, I wonder what I am missing?
However, I am currently thinking that 3 shot groups would be a better way to collect precision data. Rather than collecting 10 data points with 50 shots using 5 shot groups, I would be able to collect 16 data points with 48 shots using 3 shot groups. The reason I think that this would be better is 2 fold:
1. The odds of the subtle wind changes between a 3 shot group would be less than during a longer to complete 5 shot group. Therefore 3 shots would reduce error.
2. Any changes in the wind or shooter error that cause a flier (oulier) would have less effect on a data set of 16 than a data set of 10.
Since the consensus here seems to be that 5 shots or more is best for precision data, I wonder what I am missing?