Editing Accurizing (section)

===Defining accuracy===
[[File:AmmoTestResults.png|thumb|right|200px|Graph showing the results of an accuracy test using 3 different revolvers and 7 different brands of ammunition.]]
Even defining accuracy can be problematic. An example of this can be shown by the following tests, run by ''Performance Shooter'' magazine in December, 1996. The magazine was testing seven brands of [[.38 Special]] [[wadcutter]] rounds in three different [[revolver]]s, a [[Smith & Wesson Model 686]] and Model 52, and a [[Colt Python]] Target model, with six, five and eight inch long barrels, respectively. Ten groups of five shots were fired and measured from each revolver with each ammunition. Click on the image at right to see a larger view of the graph of average group sizes for each type of ammunition and each revolver. The average group size for the overall test was {{convert|2.85|in|order=flip}}.<ref name=ammunitiontest>{{cite web |url=http://www.gun-tests.com/performance/dec96specialammunition.html |title=Performance Shooter .38 Special wadcutter tests |accessdate=2007-08-30}}</ref>

Based on [[average]] group size, the winner was the Model 686, which shot an average group of {{convert|2.69|in|order=flip}} across the brands of ammunition, with a standard deviation between ammunition types of {{convert|0.54|in|order=flip}}. However, the Model 52, while shooting slightly larger groups at {{convert|2.88|in|order=flip}}, was far more consistent across the brands, with a [[standard deviation]] of only {{convert|0.30|in|order=flip}}, and was the most consistent performer of the test. However, if the ammunition was tuned to the gun, the clear winner was the Python, which averaged just {{convert|1.69|in|order=flip}} with its favored brand of ammunition. The Python was also by far the pickiest, however, turning in the largest groups at {{convert|6.08|and|4.0|in|order=flip}} averages with its least favorite brands, for a standard deviation of {{convert|1.6|in|order=flip}}.

Based on this test, answering the question "Which is the most accurate?" becomes a matter of opinion. The 686 shot the best average groups.{{clarify|reason=what does this mean?|date=August 2018}} However, as the Python showed the best performance with one brand of ammunition, it might be the best choice if that brand of ammunition were acceptable for the application in question. If a consistent supply of ammunition were a problem, then the 52 might be the best choice, since it showed the least sensitivity to differences in ammunition.