Preliminary Results Of The Fuckup Files Experiment - Or Why We Need A Mix Of Tipping And Reward Pool To Strengthen The Voting Algorithm
I'm half way of the experiment started a couple of weeks ago, in which I try to understand if "putting skin in the game", (giving money in the process of voting, from your own account) can strengthen the voting algorithm in Steemit.
I published so far 5 out of the 10 articles containing various entrepreneurship mistakes, and all these articles were set up as "Decline Payout". In other words, I didn't participate in any way in the rewards pool. Instead, people who appreciated the posts, sent me SBD or Steem, in various amounts, based on their preferences and, of course, possibilities. I kept 20% of these proceedings and the rest was redistributed equally to all contributors. The full methodology of the experiment is explained here.
So far, the experiment generated a total of 30.6 SBD and 177 Steem, meaning people from the Steemit ecosystem paid out of their own pocket this amount of money.
The average redistribution was 0.723 SBD / person and 3.903/person (each contributor received both SBD and Steem regardless if they contributed only SBD, only Steem, or both).
Here are the 2 most important charts:
Total contributions per post:
Average redistribution per post, per person:
Starting with Episode 4, Hardfork 17 / 18 was implemented, which completely changed the rewards distribution and people didn't seem too willing to contribute anymore, not knowing when the rewards they're expecting will come and wether those rewards will be predictable or not.
Preliminary Conclusions
- people are more willing to contribute from their own pocket if they like the content
- if there's no predictability, people will stop putting skin in the game
- if the contribution will be rewarded, somehow, the size of the contribution will increase (people see this as an investment: vote with more money and you will get back more money)
These partial conclusions seem to validate my initial observation, which is:
voting should be an expression of personal contribution per post, and that contribution should generate a return on investment from the reward pool, relative to that contribution.
In the current structure of the voting algorithm, everybody is fighting over the reward pool based on the previous stake in the grand total of available tokens / Steem Power, which is not necessarily related to an actual post. This encourages arbitrary flagging and doesn't support contribution or growth.
I'm very curious to see how this experiment will end up.
I'm a serial entrepreneur, blogger and ultrarunner. You can find me mainly on my blog at Dragos Roua where I write about productivity, business, relationships and running. Here on Steemit you may stay updated by following me @dragosroua.
That's not an observation, just a hypothesis.
It's an observation based on the general behavior of people engaged in specific interactions. If there is something you "put on the table", the value and the power of the interaction changes completely.
In the case of content voting in Steemit, putting "skin in the game" means contributing with a certain amount of value from your own account as a token of appreciation (pun intended). The more you contribute, the bigger the reward should be.
The reward pool should be a direct consequence of people interaction, not just a pile of "free money", waiting to be distributed by whales. It's the interaction that creates the value of the tokens.
At least that's my opinion. We can always agree to disagree.
it's ok, you're right.
These are very interesting results and I wonder what the rest of the experiment will look like once the reward pool gets to "equilibrium". I will continue to play out of sheer curiosity.
Glad to see things are going well @dragosroua.
And you are generating some entertaining commentary. ;-)
Even though this is not a double blind, experiment with placebos and control groups and endorsement by the ethics committee and overseen by a professor with 412 years experience in the field of this particular research, it's still interesting to find out if people will part with some cash for your content. Especially when they don't have to.
I find it really interesting that people will focus on the minutiae of the methodology of an informal experiment as if is was a government funded research project costing millions, instead of looking at how it is progressing.
I don't think lives will be saved by what you are doing here, but it may spark some discussions, some ideas, some further research. Or it may just give you a different way of pulling in some cash.
At least you are doing something.
thanks for sharing.
@dragosroua
steem on.!
very interesting ..of course to make this a valid experiment many lesser known accounts, with average # of followers would need to be used. As it stands, your results are anecdotal ..nevertheless it touches on something I had been advocating (on deaf ears) ..the benefits of users being able to add STEEM directly to a post's reward, to be enjoyed in-part by curators, and serving a promotional function simultaneously ..while taking pressure off of the rewards pool.
I wouldn't call the results anecdotal. Contributors opted in, as opposed to the so-called "whales experiment" in which they were basically forced in.
anecdotal in the sense that these were user responses to your posts ..any pattern of behaviour is necessarily relational to 'you' and your posts, therefore a much wider sample of posters is required
The Experiment was mathematical and voluntary participation was not required by a large number of users ...yours is behavioural and thus large numbers are required, in this case on both sides of the equation.
:)) yeah, right.
This fkn guy is laughing at me :) ..someone help me out here ..AM I WRONG!!!!???
I'm not laughing at you, I'm laughing at what you wrote.
The so called "Experiment" was arbitrary and even among whales the consensus was lacking. People woke up one day being flagged and the flagging "hysteria" only grew larger. It wasn't announced in a hardfork, it wasn't backed by Steemit Inc, not even informally by @ned, it was just a group of guys deciding to play along with other people content. I don't see how this "Experiment" can be useful, let alone how it can generate mathematically proven results.
In the experiment I'm running, people can opt-in or opt-out as they wish. Hence, the results are valid. I agree that we need a bigger sample to infer a reliable balance between contribution ("skin in the game" and reward pool, though.
I keep missing it, coming on the day after. I don't suppose you could extend each post a day? Or wait to do the follow up posts until every third one?
Ack. I'm sorry, because I really wanted to be a part of the entire thing, but my inner clock has always been problematic, and recently life is a bit hectic so I lose track of things more easily. If you waited on the follow up posts, it would give me, and probably others, the opportunity to contribute to the last post, and the one before even. Since they've changed the payout periods, I've readjusted so that I'm no longer operating on having to get to posts within a day. But that means I'm missing out on this. My own fault, I know. I'll check back in.
Will try to extend a bit the time. Maybe 48 hours instead of 24. Thanks for supporting this :)
Oh and, I have begun branching out of my steemit writing routine. No longer will there be only chapters to my books :) You are under no obligation to look at this, but it's basically the beginning of a series of posts I will be doing on my area. This is all about the scenic aspect (waterfalls in this case)
https://steemit.com/life/@dreemit/anniversary-on-a-motorcycle-part-two-anna-s-falls-and-excerpts-from-the-novel-reborn
Great! And no problem, I'm really interested in the results :)
I know you mean well but this is a terrible idea.
Why is this a terrible idea?
It is the basic premise of some other attention based platforms which are in the works right now.
Of course it is. Just because you say it is, right?