Illustrations of Accuracy vs Precision and Bias

I find the six-sigma “bullseye” illustration of the accuracy vs precision problem… somewhat imprecise.

Source: “Experimental Designs for Next Generation Phenotyping”, January 2015, DOI:10.1007/978-3-319-13677-6_2
  • (a) Not Accurate, Not Precise
  • (b) Accurate, Not Precise
  • (c) Not Accurate, Precise
  • (d) Accurate, Precise

While it’s accurate to say that a bullseye illustrates the precise problem, surely there are more precise ways of illustrating overall the accuracy problem.

I mean there are so many “quality” applications other than trying to narrow everything to a point in order to shoot or bomb it — the obvious reference of a cross-hatch.

LiDar to the rescue! I found the following vendor diagram very refreshing as it relates to big data security controls.

Source: yellowscan-lidar.com

Ok, maybe that’s still an illustration of targeting… by an over-the-horizon drone strike meant to eliminate an entire building. *Sigh*

My only real complaint though for this kind of superior illustration is that red/green are biased, subject to colorblindness. Was it too hard for them to use the word “not”?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.