# Difference between revisions of "Talk:What is Precision?"

(Created page with "Herb, 4/19/2015 (1) Make it clear that Accuracy and precision are being defined in terms of their typical statistical meaning, not how shooters might sloppily use the terms....") |
|||

Line 1: | Line 1: | ||

− | Herb, 4/19/2015 | + | Herb, 4/19/2015: |

− | + | # Make it clear that Accuracy and precision are being defined in terms of their typical statistical meaning, not how shooters might sloppily use the terms. | |

+ | # In second paragraph accuracy is discussed as a function of the shooting system. A slip is made and precision is discussed in terms of just the shooter. Precision is a function of the shooting system too, not just the shooter. A good rifleman with a bad rifle and bad ammo will shoot groups with poor precision. | ||

+ | # In the discussion of precision the term "dispersion" is used. There is no point in adding another term meaning "precision." Define precision as "the spread of shots about the center of impact for the shot group. A small value for precision thus indicates a group with shots very close together and larger numbers indicate more spread between the shots." Use relative phrases "better precision" and "worse precision" and do away with the additional superfluous term dispersion. | ||

+ | # "Precision describes dispersion about the center point, and is independent of accuracy." This is an assumption, not a fact. It depends on the model being tested. For example let's consider a bomber dropping unguided bombs with something like the Norden bombsight. A bad altimeter setting would mean the airplane was dropping bombs too soon or too late. Too soon and bombs would have less spread but be short. Too high and the bombs would be long and more spread out. So in that case there would be a correlation between accuracy and precision. | ||

+ | # The section "Precision in Shooting" is really discussing error propagation. I'd move this to a discussion page of model. Model for precision (ignoring flyers for the moment) should be explained as: | ||

− | + | SD(system)^2 = SD(shooter)^2 + SD(rifle)^2 + SD(ammo)^2 | |

− | |||

− | |||

− | |||

− | |||

− | |||

− | |||

− | |||

− | SD(system)^2 = SD(shooter)^2 + SD(rifle)^2 + SD(ammo)^2 | ||

+ | :: Agreed. I'll put those changes in. [[User:David|David]] ([[User talk:David|talk]]) 14:55, 20 April 2015 (EDT) | ||

--------------------- | --------------------- |

## Latest revision as of 14:55, 20 April 2015

Herb, 4/19/2015:

- Make it clear that Accuracy and precision are being defined in terms of their typical statistical meaning, not how shooters might sloppily use the terms.
- In second paragraph accuracy is discussed as a function of the shooting system. A slip is made and precision is discussed in terms of just the shooter. Precision is a function of the shooting system too, not just the shooter. A good rifleman with a bad rifle and bad ammo will shoot groups with poor precision.
- In the discussion of precision the term "dispersion" is used. There is no point in adding another term meaning "precision." Define precision as "the spread of shots about the center of impact for the shot group. A small value for precision thus indicates a group with shots very close together and larger numbers indicate more spread between the shots." Use relative phrases "better precision" and "worse precision" and do away with the additional superfluous term dispersion.
- "Precision describes dispersion about the center point, and is independent of accuracy." This is an assumption, not a fact. It depends on the model being tested. For example let's consider a bomber dropping unguided bombs with something like the Norden bombsight. A bad altimeter setting would mean the airplane was dropping bombs too soon or too late. Too soon and bombs would have less spread but be short. Too high and the bombs would be long and more spread out. So in that case there would be a correlation between accuracy and precision.
- The section "Precision in Shooting" is really discussing error propagation. I'd move this to a discussion page of model. Model for precision (ignoring flyers for the moment) should be explained as:

SD(system)^2 = SD(shooter)^2 + SD(rifle)^2 + SD(ammo)^2