Curation strategies for minnows

A few weeks ago I pointed out that curation rewards are very undervalued as a revenue stream, and potentially very rewarding for small investors. I think some were in disbelief at the numbers. If you haven't seen them, I suggest you read the article.

Now, common counter-arguments were:

1) "It needs a lot of time to do this"

2) "It requires skill (be good in curation) and curation is difficult"

3) "It needs good IT skills" (bots)

4) "The investment is in Steem Power and its price fluctuates / we don't know what Steem prices will be in the future"

5) "The numbers cited for top curators are difficult to achieve"

...so, let's take these one by one.

1) Does it require a lot of time?

In my experience, in order to achieve something above 20-25% performance per year, no. I've been hitting over 25% (effective nearer 30%) in annualized returns, which places me near the top-end of the board, even by clicking blindly with less than 5-10 minutes per day. Some weeks I'm nearer 40% (effective much more).

This number would be higher if I also took care of my voting power to be nearer 100% (which I let it slide down to 80% - not good for efficiency) and if I didn't upvote articles that I consciously support without expecting a return.

How did I do it without investing any time?

I used simple criteria, like reputation, title, money scored at that point of my vote etc. 

If reputation is high, it means the person involved is earning a lot, so (s)he's typically getting upvotes. If the title hints at good content, all the better.  If the upvotes are low (<1$ - preferably <0.1$), then it's worth it. If you know some authors (by memory) and whether they are making money, it helps too. If what I see doesn't match my criteria, I just scroll further. 

That's the "blind system" and if it can make it to top rewards, then one can imagine if they spend a bit more time in this... As a minnow your vote is probably worth nothing (~0.00$), so you also don't need to carry the cognitive load that is associated with the implications of what you vote. Whales however are more conscious of that, and therefore have to be more careful.

2) Is curation difficult?

First, let's define curation for the minnow / small user:

Curation, on our level (my vote is currently worth 0.00$ - so I consider myself one), is about trying to find and upvote content that will gather high value upvotes, and do so before others.

Two-three months ago, this was quite harder because there was a lot of piling up of whale upvotes on the same material. The spreading of whale votes in order to increase user retention and make the platform a more viable environment for users, combined with curator teams searching over the material for worthy content, has made it easier to find content that will be upvoted.

Right now, if you find something that is good, it has a very good chance of getting upvoted because curation actually works better than ever. But if it has 5$ or 10$ of upvotes already (and it maxes out at 30-40$), you'll not get the same as if you click when it has 0.05$ on it. 

The sooner you go in, the better. But you must try to limit self-damage by not going in too early within the first 30 minutes. 

Blogging income is divided as 75% (author) - 25% (curator). What you get when you curate, is a slice of that 25%. 

Your slice of that 25% is maximized based on four things:

a) When did you upvote (ideally after the first 30 minutes)

b) The amount of money the post had when you upvoted (ideally 0.00$)

c) The amount of money the post ended up with (ideally hundreds or thousands of dollars)

d) How charged your voting power was (ideally nearer 100% - btw, this is something you can see in steemstats.com)

When I'm saying "try to limit self-damage by not upvoting early", I refer to (a).

In the first days, it didn't really matter when you voted. Now it does, because bots started upvoting everything automatically and it was decided that if a post is upvoted very fast, then most of its curation income would go back to the author, to penalize them. This would allow humans some time to actually read the material posted before voting.

That's why there's a scheme in place, where if you vote in the first minute, ~97% of your potential curation income will go to the author. In the second minute ~94%... in the third ~90%, etc etc... It's like +3% per minute. In the fifteenth minute 50% of your curation income will go the author... In the 29th minute 97% will be yours (and 3% to the author) and after the 30th minute 100% will be yours. 

So, ideally, you want to vote after the 30th minute. But since others will do so too, you then start to think "well, if I go in during the 29th minute, then I'll front-run the others and make more" (that's factor #2 - you want to be voting before other heavy votes start piling in). Of course others will adjust to this as well, and ultimately what decides it, is your level of risk taking.

That's more or less the recipe of success:

...and you then wait for the whale or community-delegated curators to upvote what you did. Even if half your choices are flops, or in some cases even if 3/4ths are flops, it doesn't matter. Those that "win" will pay. That's how it goes.

3) Is curation without bots, doomed?

No. Bots have certain issues in detecting quality which humans do not have. How can a bot tell a nice photo, a nice article, etc? It can't. In the future, AI-bots will be able to make these distinctions, but for now we have if-then-else bots. Bots can work around the clock though, but since our votes are limited that's not a problem. Even if you were over the monitor all day long, your votes would be limited in what you can vote before exhausting your voting power. So that works in favor of the human voter. 

4) SP value uncertainties 

That can be a factor, but keep in mind one can multiply their SP over the long run based on curation alone. Besides, what does anyone have to lose by clicking upvotes? Time? They can do a blind-upvote strategy by simply going through the titles and be over with it in a few minutes. It's definitely better than not-voting. Even if you have +10/+20% in SP from curation, it's not bad.

5) Is top curator performance difficult to achieve?

I think percentages around 25-40% are quite realistic. I mean there are bots right now who vote blindly based on sets of rules like "if post is over 30min, reputation is over X, and post value is below Y, vote it" - and get over 50% in terms of annual performance. So it's definitely doable for much lower percentages than that.