The Incentives of Steemit - Part 1 Unitentended Consequences
Incentives
In order to best use the Steemit platform, to make money, to create and share content, and to build an all around better community, we must first understand the incentives for different types of 'Steemit Behaviour'.
When most people think of incentives, we think about very specific and focused rewards designed to illicit a specific behaviour from others. We are very familiar with the most common kind of incentive - a contract job for a set amount of money. Paint this fence, and I will pay you $100. Pretty straightforward, and very hard to 'game' this type of incentive. Either the fence is painted, or it isn't.
But what about complex behaviour? Or goals that are subjective? (reward "good" content, for example?). How do you build a system where the best content rises to the top, and the worst content is ignored? How do you combine that with a crypto-currency that you would like to be mass-adopted and also that holds its value? Perhaps a crypto-currency that you would also like to be traded, invested, held, spent, etc just like a 'regular' currency. How do you build the proper sets of incentives to make all of that happen?
Well in short, you can't!
"Incentives matter; not just the ones you like to think about, or fit your argument. All of them matter." Matthew Zietzke
The trouble with trying to incentivize complex behaviours or outcomes is that with complex systems, there are always unintended incentives also created. People naturally seek out all of the possible ways in order to capitalize on a reward - even if it runs counter to the original intent of the incentive.
A great example of this happened to the Los Angeles Police Department back in the early 1990's. There had been a string of high-speed police chases that had ended tragically with innocent bystanders getting killed. There was a lot of pressure on the department to make changes to their pursuit policy, because the public was being put at risk by all of the chases.
The department came up with new guidelines. If a suspect vehicle raced into a residential area, or into busy areas of town, and was traveling over a certain speed, then the officers were instructed to stop the pursuit. The policy was to let the perpetrator go for the sake of public safety.
So, what happened? Well, fatalities caused by high speed pursuits went up!! The criminals learned that if they raced into a residential neighbourhood, or towards a school or shopping district, that the police would stop pursuing - and they would get away! So that is exactly what they did. They drove even faster, and drove towards the very people that the new policy was trying to protect. Also, criminals who would not normally speed away were incentivized to do so - criminals who had never tried to run in the past were now speeding off towards the closest elementary school.
Eventually, the department had to change their policy yet again - but this time, they went the opposite direction. Instead of calling off pursuits, they now put in place a policy that stated that they would NEVER call off a pursuit, regardless of where it led, or how fast it got. What happened? Well, fatalities went back down as the criminals learned that there was no point running. No matter what they did, they were going to get caught. The number of high speed chases declined, and so did the fatalities.
<br/
Stay tuned for Part 2 - where I will delve into the Steemit incentives, and their unintended consequences, as well as some of the community initiatives designed to "correct" those incentives.
Thank you for your support,
Snubbermike
I wonder how incentives would change if we ended delegation. There would be less large bots voting but then again auto voting services will remain and only certain authors would make the majority of the list.
Or maybe we need incentives for flagging inappropriate or overvalued content.
How do you determine what is overvalued content?
You upvoted your own comment (something I don't have a problem with) instead of finding a quality post to upvote. In many people's minds, this makes it overvalued...
I don't like flagging on the site at all, but that is just my preference...
Self upvote of comment is to attract attention and jump over all the thanks posts.
There is definitely a gray area of what is overvalued. A model could be developed comparing Steemit Article Quality verse Freelance Writers. I do not even know what freelance earns per article but I am sure someone knows the answer.
There are definitely some black and white cases. The other day I saw a post commenting on one individual with $28 comment self upvotes which were all just thanks or other spam.
Don't need to end it completely, just disincentize it by making it less worthwhile.
Power in your own hands is more lucrative than power loaned to someone else.
The incentive of money is a motivator but users quickly find out that earning money on steemit is as hard as an MLM scheme and we lose alot of users that way. the retention of new users issue will become prevalent in the near future as people begin to look for a facebook alternative which steemit is not.
i look forward to your next post!!
I started on Steemit about a little over a month ago and haven’t been on Facebook or YouTube since it’s been a great replacement for me it’s just made a bigger broader friends list for me finding following people from all over I’m drawn for the content followed YouTubeers over to here and I just love it
Yes I’m new here and reading your posts has helped me tech handicapped mite add figure out what I’m doing here and how this entire thing works the chatting and interaction with everyone is rewards enough for me
"Incentives" caught my attention. Been here for a month+ and I noticed all kinds of behaviour here at Steemit. Bit like in the real world. Sometimes, I think people don't really tell the truth. Anyway, I keep on looking..
I definitely think there is a pool of minnows like me who are trying really hard. I'll keep trying and won't give up or even consider yet. The people I knew who joined the same time I did gave up and I think that's for many. I don't get it all the time, I figure put out content that's different and hope to meet others who enjoy the same kinds of things. But the stuff I read makes me wonder if i'm going to find out there's no way organically? I hope not.
Why does steemit make it so that the choice for 100 percent steem power payouts will actually make you lose money?
The payouts also do not correspond to what you invest and supposedly win, with the bots. im still trying to figure tht one out.
The only ones in the bot equation who are guaranteed a positive return are the bot owners.
Most people are not educated about operant conditioning, validation strategy, psychology, split testing, etc. I keep trying to educate people that improperly tested conditions, regulations, guidelines, "great ideas" are not so great afterall— but it's against the "guidelines" for me to include diagrams I didn't myself create, and I don't have the energy to get into explaining a college degree worth of information to people.
I'm looking forward to what you have to say in part two. I'm going to try to make some of my own diagrams about these topics also— because the ignorance is painful.
I look forward to seeing what you create (diagrams). Followed
I like how data speaks for themselves, one can never predict how a policy works out until the effects are seen.
Yeah seriously. There must be incentives for people to continue on steemit but thats gonna take time. For now, good contents are not rewarded appropraitely and a few who knew about steemit earlier have monopolized the system. I am really hoping that new policies would be adopted soon to make the system fair. Looking forward to the part 2.
Well, your story proves that good incentives works well.
The problem with Steem is that it rewards good content rather than good content that is hard to find which leads to rewarding those that were rewarded before since they are more likely to write worthy content rather than someone new who wrote only few things so far.
But the point was that those who tried to set up the system first got it wrong. They expected fatalities to drop when they no longer took pursuit. What they didn't realize was that they were inadvertently setting up an incentive to flee at high speed (the chance that the criminal would get away).
We must always keep our eyes open for those unintended consequences...
Standardizing subjective behavior/perspective is just not possible: some people value memes more than 1000-word posts, and vice versa.
Exactly, so what happens in that environment? People seek out the easiest way to grab a piece of the rewards pool instead of striving to create the best content.
More in Part 2 - and probably Part 3
Communities with like interests are the key, i think. Also, education on how everything really works. Like the. 02 minimum vote burn rule. anything less is not paid out. That should be told to you during account set up describing payout system. It took me over a month to get to 100 percent voting power. Would have been quicker if i realized most my votes didn't financially benefit those i voted for.