Rashomon Revisited

“His argument for these is pretty simple: they ensure that we have a financial stake in being accurate when we make forecasts, rather than just trying to look good to our peers.” – Nate Silver, the Signal and the Noise

Long time, no update. Not that it’s that important though. But I’m making good progress and very pleased with what I have at the moment, though a lot more work needs to be done. Rashomon is making headway…

If you don’t understand what I am talking about, I’m referring to a previous post called Rashomon, or Fake News and What To Do About It. As a refresher, my thought was that we aren’t treating the news and information ecosystem as what it actually is, which is a market. Furthermore, the news and media organizations actually supply a commodity of questionable value, and that the way we deal with commodities of questionable value is by alleviating the inherent risk by betting. So what we should do is create a betting market on the content of the media itself.

Think to children, who often when a friend is making exaggerated claims will say “I’ll bet you didn’t” or “I’ll bet you did”. Generally, this is the child’s way of testing social claims – dare the opposing party to perform and setup an equitable trade to alleviate the risk. Pretty simple.

This isn’t a new idea. Prediction markets have been used by Google, Eli Lilly and Company, DARPA and even in the Vatican during the papal elections. It has even been applied recently, through online platforms like the now-defunct Intrade.com and PredictIt.com. Robin Hanson speculated on a form of government based off this principle, called Futarchy. Actually, Hanson was who Nate Silver was referring to in the quote at the beginning. I would even argue that derivatives and futures markets are variations on the same theme. You can pretty much bet on anything, right?

Well, if I said, “I’ll bet you the sky is blue”, would you take that? Probably not, because we know that the sky is blue. This generally leads people to believe that placing a bet means that we are betting on an event, not a state. This is why most of the markets I described above are events-driven. Who will be the next president? What will natural gas go for in two years? Will Britain leave the EU? These are events that neither party know the outcome to with binary results.

My point is simple. We shouldn’t limit our bets to future events. Time is not what we are betting on – we are betting on uncertainty. I won’t take the “Is the sky blue?” bet because I am certain it is. I am not certain, however, whether the media is accurately reporting or not.

The obvious wrench to throw here is that when we are talking about fake news and filter bubbles and echo chambers is that we cannot come to a consensus as to what is right or not, hence allegations of a post-truth or post-fact society. My answer was that we should at least be knowledgeable as to how far we are on a spectrum. I think most, maybe not all but most, would agree that Fox News is more to the political right than CNN.  You don’t have to agree that Fox News is Center or Right wing, just that relative to CNN it is more to the right of CNN than it is to the left of CNN. Using a relative rating system, we can start more accurately contextualize our information intake and our media diets. This market also exposes you to a variety of sources, since article selection should be random, and the call of a financial incentive should bring players in spite of their own political leanings.

But now the question is, and the thing I am most apprehensive about, how do I maintain the integrity of this spectrum? And also, issues don’t exist merely on a political spectrum. Many issues are multi-faceted and cover several spheres of influence. How do I account for those?

The obvious answer has been the political chart, a graph with a political axis and an economic axis. And please do not point me to the Political Chart Memes page, for everything you might hold sacred and good, please don’t derail this thought just yet. The political chart organizes several thoughts well, but does so according to power (political spectrum) and value (economic spectrum). It does not account for an informational spectrum. My thought is further depth of analysis might come through a three-dimensional understanding of a view, so instead of a 2D political chart, we would have a 3D political sphere.  There are a few ideas going on there, so let’s dissect this a little further.

What are the extreme ends of this information spectrum, i.e. how is this spectrum defined?

I think one way to dice it would be inductive vs deductive reasoning. What I mean here is that I often see one method of processing information as collecting facts or data and then extrapolating from this data as much as possible. This is what I would refer to as inductive. On the left end of the spectrum in the US, I see this in our technocratic liberal elite and priest-class scientists, and on the right, I see it in our neo-conservative military and defense strategists.

The opposite of the “Data-is-God-analysis-is-flawed” way of thinking is deductive, when knowing a few principles or axioms can lead one to extrapolate out to some version of the truth. On the left, I see it in the historical materialism of Marx and on the right in the Austrian School’s praxeology.

This tends to be the method of the public because it isn’t the public’s job to know a bunch of facts about a particular issue, so it is easier to rely on a useful heuristic and connect the dots on the way. The ‘elites’ tend to use inductive reasoning because it is their job to know the ins and outs of a subject. So, implicit in this axis is the level of trust in the self and the crowd vs. the experts.

Obviously, this is a spectrum as few would claim to be entirely inductive or entirely deductive.

The other idea behind this model is that a spherical model of politics, much like an extend version of the sometimes-maligned horseshoe theory, now exists. Imagine that the equator is the political spectrum and the Prime Meridian is the economic axis. What goes directly through the intersection is my informational axis. Every political issue can be defined by longitude (the political or social stance, either right or left), latitude (economic stance, planned or free-market), and depth-from-center (informational stance, inductive or deductive).

This is very useful to my news betting market idea because it provides some self-reflection as to how it is consumed and might help map out  any data gathered from such a market. But I am also keenly aware that I am reinventing the wheel to some extent and am thus calling out for feedback.

Reach out, I would love to hear more on this issue.