Mirror mirror on the wall who's the best curator of all?

Steemit is a reflection of all of us and curation is one of the most important parts of what makes steemit fun and profitable!

Everyone has their own methods, reasons and criteria for curation and each of us does it our own way.

Steemit will live and die by the quality of content that is surfaced by dedicated curators and each of us should tip our hat to the people who do this work by hand.

Nevertheless, the benicents project has its own criteria for curation, this criteria may or may not line up with any particular person or group's objectives.

Our goal with curation is to maximize curation rewards by picking content we believe will trend and get in before the whales do!

But to do curation successfully, we need to find the best curators, people who consistently get in before the whales, because once the whales jump in the pool, curation rewards are minimal.

Once we find these excellent curators, we want to hitch our wagon to them so to speak. We want to watch what they do and act accordingly. This has the effect of building a trail. A trail that can dynamically select the best curators and respond as quickly as possible to changing conditions in steemit world.

But how do we do this?

Well if you saw our first post, you'll notice that whales presently contribute a total of 12% of the total rewards on a post with each whale putting up between 0.8 and 2%. This gives us a great clue about where to begin.

To start with you need to fetch the top 100 posts in trending.
Once you have those, you need to examine all the votes, thankfully the votes are returned in order that they were cast.

The next thing you need to do is to examine each post's rshares.

RShares are Reward Shares

R-shares
The contribution of a vote to a post (or comment) is measured in rshares, and they are calculated from the vesting shares of curators (steem power) and the voting power used.

where Vs are the vesting shares of the curator, and rs the rshares generated. The absolute value of rs must be greater than 50 VESTS, this is a dust threshold, the minimum amount.

So in plain English what that means is that each post gains a value called RShares with each upvote and the total reward for the post is calculated from that. For right now we don't need to worry about the specifics, but if you're interested the information is at https://www.steem.center/index.php?title=Rewards:Formulas#R-shares

This gives us a counter, an accumulator and a gatekeeper function we can use.
Simply add up the rshares on each vote until the total rshares counted equals or exceeds 12% of the total rshares associated with the post.

As you sum the rshares, increment a global counter specific to each voter, this counter tells us immediately how many times that person picked a winner, i.e. upvoted content while it was beneath that 12% threshold.

Once you have that, all you need to to is sort the voters by counter, drop anyone from the list who didn't pick enough winners that today and voila! You produce a list like the one below.

NAME	Count	AVG RShares
@gerber	12	243044711264
@daan	10	197295488619
@cheolwoo-kim	9	3710707755
@loreennaa	4	105524215

So our top performing curators are whales, but these whales aren't voting full strength and that gives everyone a chance to play. The top 2 are @gerber who picked 12 winners and @daan who picked 10. Nice job guys!
However @cheolwoo-kim and @loreennaa picked 9 and 4 respectively despite being minnows, that's fantastic!

Now keep in mind, the list above is not meant to be static. It could change hourly and most definitely will change daily.

However the reasoning is sound. If you trail these four you'll find you're picking the winners well before the whales jump in and soak up all the rewards ;)

We will be releasing a trail / ditto vote app soon as well as source code on github, so stay tuned!
As always, this post is 100% Steem Powered Up!