Depending on how you look at things, you're either about 10 months behind the times (Steemit and the Web of Trust: A Potential Love Story) or 10 years behind the times (Jaanix, the software which I cite in the course of my article).

Which I suppose I can't really mock over much, since people involved in cryptocurrencies don't tend to be extremely well versed in the historical design of social networks and all of the things which have come along on the way. Web of Trust systems have been around for a very long, long time – but we don't see a lot of direct and clear implementations for a lot of reasons. One of which is that trust is generally more complicated than "I trust you," but tends more toward "I trust you on this subject" and that introduces a level of complexity both in storing those lenses and presenting them.

As such, I think you're going to have real problems with "this is my vision and cannot be copied by another person." Ideas are cheap. They're everywhere. Anyone can have them. No one can own them. And you're a decade late getting to the buffet.

Sorry, kid. Life got there first.

But let's go up to the section which is really wrong. Demanding Web of Trust systems is perfectly reasonable. What is not reasonable is having an absolutely broken understanding of how humans interact with multiple choices, and how that interacts with their ability to rate things.

There is a vast multitude of study crunched into this particular field, and if you're really interested I would encourage you to look into those sociological studies, but I'm going to use examples taken from the field of game design because it's what I have handy in my head right now.

Ranking depends on the ability of the user to discern differences between classes at an elemental and intuitive level. If those differences are not discernible, people will not differentiate them. Speaking purely from the perspective of people ranking things for themselves, you find a significant break off if you give people more than five options. Analysis paralysis sets in, and ultimately what you find is that people really only give three: the highest, the lowest, and the most middle.

This is also a really strong pressure when you find people having a five-point scale without clear definition of qualifiers for each point. You will see a lot of fives, a number of ones, some threes, and almost no twos or fours – because they simply don't matter.

The reason that people break things down into three cluster rankings is because they think comparatively. "Is this thing better than that thing?" Either it is, it isn't, or it's about the same. That's the cognitive load that people will use regularly, usefully, and effectively.

Funnily enough, that's really all the information we need given repeated rankings to start making an ordered list. All we need to know is whether one thing is better than another thing and then we repeat that enough times and the order becomes apparent. It is the most basic of sorting algorithms for a reason.

This is why increasing the granularity of a rating system isn't going to get you any more useful information. In fact, it's just going to lose information because it is overly granular. People will not use it. It's just not really useful to increase the granularity beyond three if you're looking for an ordering or beyond five if you don't mind the fact that people will ignore two of your options.

The Fudge RPG had a very useful seven scale for rating pretty much everything: Terrible, Poor, Mediocre, Fair, Good, Great, and Superb. That correlated to effectively a -3 to +3 range. In practice, dropping Terrible and Superb was extremely common, reducing it to a five-point scale for all intents and purposes. And it worked wonderfully – because statistics, Traits, etc. had an increasing cost as you moved up the scale. There became real reason for deploying the Mediocre and the Great because they were inherently cheaper than the extremes.

Modern MMO's often have a five point descriptive ranking when they talk about "gear tiers," going from gray/very basic through gold/exotic to purple/legendary (adjust to taste for your personal preference in MMO). But these are descriptive systems not rating systems and always go along with actual material differences being described by the ranges.

When dealing with people ranking trust or ranking content for which there is no cost for giving higher ranks, the sensible pressure is to stay with a three point scale and make heavy use of it.

The rest, as they say, writes itself.

If you are really interested in ranking systems and the use of descriptive lenses to present large amounts of content in an ordered, useful way, I suggest going into history and looking very hard at the GNUS USENET reader from far back in the day. Despite not having a quantified ranking system, it allowed the user to assign weights to almost any metadata associated with a given piece of content and then present all available content with all of the rankings and weights so applied. If only we had something so useful for dealing with the fire hose of content coming out of the steem blockchain.

(Regarding web of trust systems, you might also want to read Steemit and The Dynamic of the Authoritarian in Assumption and Steemit and the Ultimate Recourse: Blockchains For Everyone!. Feel free to implement anything in there you find. Ideas are free. Free for everyone.)