I say devalue the system, it doesn't mean the price will go down.
If the price doesn't go down, or doesn't go down enough as a direct result of the action taken, then self-voting is successful self-enrichment. The math on this is simple: If I vote myself $400, then I'm $400-X richer, where X is the amount that my stake depreciates as a result. If X < $400 (and it appears to be far, far, less than $400 in practice), then self-enrichment is the most logical, rational action to take. This is true for both n and n^2.
That is what you are missing. The incentives are misaligned which makes the system essentially broken. Both n and n^2 are broken. True they are broken in different ways.
BTW, n^2 doesn't not enforce proof-of-brain. That is the whole problem. It enforces proof-of-large-wallet. Something that enforced proof-of-brain would be great! I would be all in favor of it. (Actually I do think there is something that might: reducing the cost of downvotes.) I don't see n^2 doing that.
It isn't sufficient to argue that n is no good. You also have to make the case n^2 is indeed good, or at least that on net n^2 is better than n. The evidence for that after a year or so of experience with n^2 is far from compelling, and there is no logical case for it either: Proof of large wallet does not imply proof of brain.
Also, the white paper (and steemit blog post, etc.) argument about simple interest is wrong in the presence of downvotes, as I suggested earlier. You can always upvote yourself, but that's not enough to actually get paid! If someone else downvotes you (because your upvote does not meet their proof-of-brain standards), you are no longer guaranteed simple interest (and indeed your valuable vote power has been wasted, a potentially-worse outcome than voting for someone else and just getting the curation reward). You are forced to, at a minimum, create-and-upvote something that meets others' proof of brain standards and is not downvoted.
if all rewards are allocated through proof-of-no-brain
This theory is largely refuted after a year anyway. A lot of rewards are recovered by stakeholders via self-voting, but not all. There is still a very significant quantity of voting on merit, including when I have voted for your comments in this thread (and you for mine, and finally both us voting for others). In addition, posts with high merit such as Steem Monsters and other significant projects, get large amounts of merit votes.
So we can see that linear weighting does not reduce to simple interest in practice, and likely never will. So yet again, the theory is wrong because it does not describe how actual humans act.
What we see is might even be interpreted as evidence of stakeholders regulating the size of the reward pool in a somewhat healthy way (giving out more when there is more merit to justify the inflation, less when there is less) rather than inflating just for the purpose of inflating.
True. I haven't claimed the opposite. The wealth is dependent on the demand if the demand doesn't change then this factor is moot but the demand always fluctuates over time.
Short time this is true, but the longtime life of Steem is dependent on the market's demand over an extended period of time.
Under linear rewards, those who exclusively upvote themselves can hope to benefit from the Steem inflation minus the witness rewards. There's potential for these people to be flagged but then their potential reward would return to the reward poo,l where they will be available to anyone, including the other exclusive self-upvoters.
Currently, some exclusive self-upvoters are getting flagged every day. Those who flagged aren't getting richer and their flags are getting proportionally smaller compared to the rewards.
Under superlinear reward, @freedom could create 10 posts, occupy the whole trending page, get much more than the yearly inflation but that would decrease the demand for Steem to the point of not being the most monetary rewarding strategy.
Or @freedom could look to reward the people who will bring value to Steem and make Steem the center of attention of a growing amount of people, making Steem ever more valuable.
If any of the whales would decide to give themselves too much of the reward pool, bigger whales or combination of whales would have the power to raise their flags against those actions.
What is important is not merely whether demand fluctuates, but whether and how much it fluctuates (or can be reasonably expected to fluctuate) as a direct result of an individual's actions. Fluctuation, in and of itself, does not tell us anything about the best actions to take. I could do something absolutely horrific qualitatively in terms of Steem's value, and Steem could still go up in value! I'm only small stakeholder, even if, relatively speaking, one of the largest. When it comes down to it, my little actions just aren't that important (and the same can even be said for @freedom, etc.)!
Long term is just the sum of a series of short terms. If you want the long term incentives to work you have to fix the short term incentives. That is a mathematical fact.
Not necessarily. We've already discussed this.
Or even if so, @freedom could easily create 100 sock puppet accounts, post cookie cutter content that appeared to be meaningful and upvote those instead.
on the overall value propositionBut even if that weren't the case, @freedom is only about a 3% stakeholder (as far as we know). Whatever @freedom does even with n^2, still has limited effect (see above). The perfectly rational thing for @freedom to do is to maximize @freedom's wealth with self-enrichment, which is pretty much what @freedom has always done (first with dumb but exploitative curation bots, now with bid bots). If you don't think so, then you are arguing that someone who seems to be a pretty smart dude and objectively one of if not the most successful Steemian who turned nothing into $20+ million by behaving very smartly at every stage is and has been doing something very, very wrong. I don't think so.
My very serious suggestion to you would be to spend some time with a pencil and paper and work out some very specific scenarios in terms of hypothetical (but plausible) stake distribution and also hypothetical (but plausible) numbers on how much these individual decisions effect the overall Steem value. If you do this you will find that the way the numbers work out, it is almost impossible for self-enrichment to ever not be the best individual strategy regardless of the reward curve (for each stakeholder who isn't already almost completely disenfranchised by a heavily top-heavy weighting system).
If this has been done early on, the white paper and overall design would not have made such obvious blunders. Either no one did this, or they didn't care because the goal was to launch a coin with a compelling-sounding story to cash out on a big pump, whether or not the mechanism for what it claimed to be trying to accomplish actually worked.