The perplexing ways of PolitiFact

Yesterday PolitiFact published new version of one of its recent fact checks, changing its "Truth-O-Meter" rating from "Mostly True" to "Half True."

PolitiFact's action solidifies the charge that it rules arbitrarily and subjectively.

The original ruling on a description by the Obama campaign of Paul Ryan's plan for Medicare probably already downplayed a key misleading aspect of the original ad.  How does one make an accurate charge about "the" Ryan plan while talking about a version of his plan that has been updated?  If the relevant features of the plan changed (they did), then the statement misleads in the absence of qualifying language.  And it misleads in a way that may give the audience a picture that is the opposite of the accurate picture--that's approximately how PolitiFact defines its "Half True" rating.

PolitiFact does usually follow a good policy when it revises a story:  It keeps an archived version of the old story.  Since I have researched PolitiFact's process of justifying its "False" and "Pants on Fire" ratings, I was curious as to how PolitiFact changed the wording of its concluding paragraph to support the altered ruling.

The original version:

The Obama ad would have been more accurate if it had specified that it was referring to a previous Ryan plan for Medicare rather than the current one. We simply don’t have enough details to know how much extra money seniors might have to pay under the current Ryan plan. Still, the Obama campaign gave itself some wiggle room by saying that the plan "could" raise out-of-pocket costs by more than $6,000. On balance, we rate the statement Mostly True.

The revised version:

The Obama ad would have been more accurate if it had specified that it was referring to a previous Ryan plan for Medicare rather than the current one. We simply don’t have enough details to know how much extra money seniors might have to pay under the current Ryan plan. Still, the Obama campaign gave itself some wiggle room by saying that the plan "could" raise out-of-pocket costs by more than $6,000. On balance, we rate the statement Half True.

The summaries differ by one word.  "Mostly True" changes to "Half True" for the updated version of the story.

For the sake of clarity, is there any part of the story that provides a better opportunity to explain the precise justification for a "Truth-O-Meter" ruling than the summary portion of the story?

It's features like this that make PolitiFact's "Star Chamber" decisions look and smell exactly like subjective opinion.