How would you visualize the truthfulness of Presidential candidates?
Well, my friend +Avinash Kaushik
asked just that question in his latest email newsletter (which is excellent: https://zqi.me/akintersect
), and also here on G+ (https://goo.gl/5bW0qj
And it got me thinking. The examples that Avinash used were all based on visualizing the actual data, but what if we instead thought about it in relation to how important the truth is to us?
What would look like then.
So, I took a different look at data, which BTW is based on Politifact’s Truth-O-Meter. It has 5 levels overall:True – The statement is accurate and there’s nothing significant missing.Mostly True – The statement is accurate but needs clarification or additional information.Half True – The statement is partially accurate but leaves out important details or takes things out of context.Mostly False – The statement contains some element of truth but ignores critical facts that would give a different impression.False – The statement is not accurate.
And then they also have a 6th level for statements where a politician is not just making a false statement, but is so out there that it seems to be intentionally misleading. They call it:Pants on Fire – The statement is not accurate and makes a ridiculous claim.
So, if we were to rank these levels in terms of how important the truth is to us, we might say something like this.
We define the baseline to be ’Half-True’
, so we give this a value of zero
because it's the minimum value that we can expect.
We then give ’Mostly True’
a value of 2
, and ’True’
a value of 5
. The idea here is that we reward not just that something is true, but also that it provides us with the complete picture (or close to it).
Similarly, we punish falsehoods. So, ’Mostly False’
is given a value of -2
, and ’False’
a value of -5
Finally, we have intentional falsehoods, the ’Pants on Fire’
, which we punish by giving it a value of -10.
We then take the percentages for each category and for each candidate and multiply that with our score, and the result is the picture below.
We can always have a discussion whether the scores are using the right numbers or not. For instance, should ‘Pants on fire’ really be -10, or should it just be -5 (same as false)? But even if you completely removed that category, the outcome is largely the same.
As you can see, Clinton, Sanders and Rubio are largely in the same league. They say some true things and some false things. Keep in mind that the ‘Mostly False’ (orange) category are statements that contain some elements of truth, and Rubio actually has a fair bit of those.
In fact, if we look at the raw numbers, we see that his falsehoods aren’t actually that bad. They are still false, but it’s not that clear cut.
Cruz and Trump, on the other hand, are in a completely different league. They lie by default, with Trump rewriting the book on lies.
Clinton and Sanders is also an interesting comparison. Overall, Clinton wins because she has a larger share of true statements that are also backed up with real information. Sanders, on the other hand, has a larger share of true statements but they lack clarification. So, he is not less true than Clinton, he just doesn’t go into the details as much as she does.
If we only rank true statements overall (regardless of how true they are and without ranking their importance), we see that 51% of Clinton’s statements are mostly true or completely true, while it’s 49% for Sanders. That’s practically the same. The difference between the two has more to do with how they communicate.
And the same is kind of true for Rubio. He doesn’t go into as much detail so he ends up at 35%.
But, again, Cruz’s 'truth percentage' is only 21%, and Trump is at 7%. They are basically just making shit up.