It's not close to correct. It is wrong. The article literally links to the correct test, but the author's language cobbles together of a few sub-links worth of

It's not close to correct. It is wrong. The article literally links to the correct test, but the author's language cobbles together of a few sub-links worth of half-digested material that the author doesn't understand to avoid plagiarizing the legal dictionary. There is no quantitative assessment of how much evidence supports one side or the other in the test at all. Ignoring how ironic it is to make a mistake of this type, it's particularly grating because adding more evidence to one side of a dispute doesn't necessarily add to how much we believe their position, yet most of the article's statements regarding proof are made with that assumption. He equates evidence with proof (or at least an element in approaching proof), but fails to address the connection between evidence, pre-existing belief and interpretation. This is why people talk past each other. This is why different news orgs can 'spin' the same facts in opposite directions. Ignoring that complexity leads to mistakes. This error is consistent and the top parent comment in this comments section notes a particularly glaring weakness in the non-bayesian approach.