x

avatar
(Edited)

x



0
0
0.000
10 comments
avatar

I think that a 2nd price auction model would incentivize splitting the stake of high powered accounts:

Let's say that my account can vote with a strenght of 2000 rshares and there is a post with 2 votes worth 100 rshares. I am incentivized to vote with the same 100 rshare strength. Otherwise my vote becomes an outlier if I vote above or below the current distribution.

If my goal is to maximize my curation rewards it is better for me to split my stake into 20 accounts and vote with an equal amount of voting power with all the accounts on the same post. Or I could split my account in 4 and vote with a strenght of 500 each and turn the other votes into outliers.

Or even better create a high number of accounts with a moderate amount of voting power and turn everyone else into outliers.

0
0
0.000
avatar
(Edited)

Another possible consequence of this model is to incentivize last minute voting. Just wait until the last moment before the voting period and vote with a strenght just below the highest one as opossed to the current state of affairs where you are incentivized to vote around the 15 minute mark.

0
0
0.000
avatar

There are different ways you could implement it, but the way I did it still uses time of vote in determining rewards. Voters get credit for their own voting time with the next lower rshares, and it uses the same square root weighting as today. So early voters would still get the lion's share of the rewards. It probably wouldn't make sense to wait 'til the end.

You're right about the possibility of splitting one stake into multiple accounts. I mentioned that in the article. But, if you have a small number of accounts, you have to sacrifice a large percentage of rshares. If you have a large number of accounts, there's less of a built-in penalty, but it's easier for others to find it and downvote it. (and someone would have to decide if the largest downvote is also going to be ignored as an outlier. I did not include downvotes in the simulation.)

0
0
0.000
avatar
(Edited)

From a game theory perspective we probably need to formalize the model with a payoff matrix to see how different strategies playout. Although with the complexaties of the steem reward pool it's an ardous task.

0
0
0.000
avatar

I guess the thing to do is take a careful look at the paper with the Steem voting analysis, from my post the other day, as well as the original 2nd-price auction analysis by William Vickrey that gametheory.net references, and try to combine the methods for an analysis of 2nd-price inspired voting.

Probably not an effort that I'll have time for in the near future, though.

0
0
0.000
avatar
(Edited)

This is slightly off topic but Derek from Veritasium uploaded a video today that touches how Youtube ranks the content using it's algorithm, I find it interesting.

Actually it's somewhat on topic because Youtube utilizes it's algorithm to curate content for it's users.

0
0
0.000
avatar

Thanks. Looks interesting. I'll try to find time to watch it later today.

0
0
0.000
avatar

Just watched it. That was interesting. Thanks for the link.

Depressing points about the click-baity titles and clickable subjects, but I think he's right. It may not apply quite as much here since our feeds still follow the subscription model, and since trending posts rely on whale-votes or bidbots. But I'm sure clickable subjects and click-baity titles probably help a lot here, too.

0
0
0.000
avatar

Speaking of algorithms, have you checked Steeve? According to the app it uses AI to curate Steem content although I haven't looked much into how it works.

0
0
0.000
avatar

I tried it a while ago. Not recently, though. I like the concept, but I've been using steempeak, almost exclusively, 'cause I really like the grid view, and the split-screen editor. I guess for half-a-year or more. I check in on busy every once in a while, and I should probably also get in the habit of checking in on steeve periodically too, to see how it is progressing.

I guess clickable topics and click-baity titles would be more important there....

0
0
0.000