Originally Posted By: Youper Denton: This past year I've been using a new to me method of comparing rifles and loads, and I would like your opinion of the value of it. I shoot a ten shot group, mark the median elevation and windage point, then measure the distance of each shot from that point. What I compare between loads is the median value of each load.
Why are you using medians, rather than means?
I can think of two good reasons: 1) Medians are easier to do in the field. Slide a card down the target until it is over the center of the middle shot (or between the two middle shots if an even number of shots), and strike a horizontal line. Turn the card 90 degrees, do the same (might be a different shot) and strike a vertical line. That is the median position of the group. No calculations. 2) Medians are robust to outliers. A flier will not pull the median, as it does the mean. If you have a flier, it's probably a "more true" representation of "center" than the mean is. That's good for sighting in your scope.
I'm not quite clear on how he is evaluating "spread", but mean distance from center, median distance from center, standard deviation, and group size all convey essentially the same information in "different clothes".
