Politi-Score: The Beast is Yet to Come

The website The Daily Beast recently jumped on the PolitiFact scoring bandwagon complimented with a gallery pictorial 1 to 10 ranking review of what it describes as the “Most Ignorant Politicians.” I beg to differ with that, as I believe those that are listed may just be dumb…like a fox, as they say. Their “scoring” methodology, which is more akin to a "Pants-On-Fire Meter," is explained as follows:

The Daily Beast tried to measure the most fact-challenged politicians in America, who, based on their public statements, apparently need to brush up on their high-school reading. For our data, we relied on Politifact, the Pulitzer Prize-winning website founded by the St. Petersburg Times, which independently verifies or corrects statements made by public officials. We limited our pool to well-known politicians with at least six statements inspected by Politifact, and then assigned a point system based on Politifact’s Truth-O-Meter designations:

0 points for True or Mostly True statements

1 point for Half True or Barely True statements

3 points for False statements

5 points for Pants on Fire statements—the most absurd, ridiculous statements

For the final ranking, we took each politician’s point total and divided by the total number of statements analyzed by Politifact for each politician. Care to take an educated guess at America’s Most Ignorant Politician?  Click here to find out.

I’m assuming the “True” and “Mostly True” rulings, even though scored zero, are included in the “divided by all rulings” number, which would serve to mitigate the overall calculated number. But you will notice, looking at the scores as I calculated below for the Top Ten with my dataset as collected through the end of 2010, that from an intuitive standpoint, the Daily Beast scores (“DB-Scores”) are not easy to interpret. Here are the top ten “Most Ignorant” using the Daily Beast scoring applied to my through-2010 database, with the Politi-Score next to it. While the Politi-Score trends in the opposite direction to the Daily Beast scores, it does not correspond exactly. This is because the “score 5 for Pants on Fire” really skews its overall calculation, as can be seen by Michele Bachmann, who has received a LOT of Pants on Fire rulings, and has a 3.92 overall grade towering way above everyone else.

The names highlighted "pink" (red) show the Republicans.



Tina Brown, Editor
of The Daily Beast

My caveats were: (1) I used all PolitiFact state affiliates (Daily Beast’s analysis may have only included the rulings of PolitiFact National), which means I have those lesser known politicians, although I wouldn’t consider Rick Perry, who some are looking at as a possible 2016 presidential contender, as a lesser known; (2) as noted, my data only extends through the end of 2010; and (3) I included all those with 10 or more rulings. It would take some time to combine the 2011 database with the 2010, as their formats are completely different. That said, the Daily Beast’s inclusion of Paul Ryan had more to do with the recent national noise surrounding his budget/Medicare proposals, because through the end of 2010, he only had TWO (2) rulings. And even six rulings as of March 28 (the date of this article) are nearly not enough to get a comprehensive score.

On the flip side, the Daily Beast didn’t provide a measure of its lowest scores, or a listing of the most truthful politicians. In a sense, from a numerical viewpoint, doing that brings to light the flaw of its scoring system:  a politician could technically have six or more Trues and Mostly Trues, which would be an overall score of zero; so, unless there is at least one "Half True" ruling, the scoring is useless. I suppose since the Daily Beast's intent was to look for "liars" this was the way to do it; but I think readers would be interested in knowing the other side. Here they are listed, like the least truthful above, using Daily Beast’s scoring method, with the corresponding Politi-Score (fortunately, no one attained that situation of zero-out). The Politi-Score in this case has more of an inverse relationship because fewer False and Pants on Fire rulings were skewing the numbers on the “fibbers” side.

Getting back to the understandability of the scores, using 100 as a baseline for "True", it's easy to discern, for example, that former candidate for Florida governor Alex Sink is, on average, a "Mostly True" politician, and Congressman Ron Paul is not far behind....or as in the "Least Truthful" listing, that Michele Bachmann has hot pants (lol) or Rick Perry is about half way between Barely True and Half True.  The Daily Beast score, however, does not tell you anything definitive about the truthfulness itself except when ranked with the others.

I’m glad, however, that the Daily Beast started this foray into attempting to measure PolitiFact’s rulings by politician with the disclaimer that it “decided to have some fun” and claimed only to have done the measure as a result of a Newsweek poll. When I started my Politi-Score I did it similarly by only scoring well-known politicians or pundits with a large number of rulings. This is not an effective method if one is taking it seriously, which I explained in a later post.  It is becoming increasingly apparent, though, that more and more of the media are becoming aware of the PolitiFact rankings as a measure of politicians' truthfulness.