50/50 curation and bots getting more?

I wrote a post last night about 50/50 curation and many people are against it so, I decided to do some numbers. I am not a numbers person so I kept it simple so I could run a scenario.

There are good conversations for and against on that post. There are many, many more on this post by @theycallmedan.

My goal with this isn't to increase my curation returns - it is looking at options that will make the platform a better place. In my opinion, if people entered into the platform and 50/50 was already in place, there would be no argument about it at all and, most people would be happy with it, like they are on smoke.io - most of who are also on Steem.

Aligning incentives across all stakeholders is paramount to a healthy system and, they are not aligned at the moment for nearly anyone.

This is the scenario I ran:

If a bidbot has a $100 vote and 10 votes a day. It offers at least a 10% return after curation which means, it has to vote at 150% of the bid price. It offers all profits to the delegators and takes no cut.

All things remaining equal (no other voters):

The bid is $66.
The vote comes in at $100
Seven days later, the payout is made. The bidder gets 75% = $75 payout = profit of $9 = 13.6%
The bot gets $25 curation + $66 dollars and returns $91 to delegators.

total sends = $660
total payouts = $750
total curation = $250
delegators get $910 / authors 90

This is under the current 75/25 split.

Under 50/50 all things remaining equal to get the 10%(no other voters):

The bid is $44.
The vote comes in at 100 dollars.
Seven days later, the payout is made. The bidder gets 50% = $50 payout = profit of $6 = 13.6%
The bot gets $50 curation + $44 dollars and returns $94 to delegators.

total sends = $440
total payouts = $500
total curation = $500

delegators get $940 / authors 60

Currently, a voter with a $100 vote gets $25 back on the vote. They delegate to a bidbot, they get 91 dollars back on the vote. A difference of 66 dollars.

Under 50/50, Currently, a voter with a $100 vote gets $50 back on the vote. They delegate to a bidbot they get 94 back. A difference of $44.

Uh oh... not looking good for 50/50??

Time to add some risk

Is this enough to convince more manual curators? Probably not. However, what it does do is close the gap by 33%. It also means that for the bidder, there is a higher risk as there are less flags needed to put the returns into minus and as a curator gets 50% curation, flags are a bit cheaper too - meaning better content? Flags also drop down the ROI for delegators through curation - as well as less people taking the risk to buy on crap?

Then, with higher curation return, there is added incentive to frontrun the bots for curators (auto or manual) and for small accounts, they can get well over 100% curation return voting before the 15 minutes, something the bidbots can't do because of a loss of curation. So, more accounts will vote before the bots lowering the bot curation return significantly.

With the current system, the "send" amount is safe and locked away for delegators and that is $66 for the bot, 9 for the author. The risked amount (what people can frontrun) for the bidbots is 25 dollars.

With 50/50, the locked away amount is $44 for the bot, 6 for the author, but the exposure is $50. That is the amount that manual curators are able to now vie for through curation activities and the smaller curators are going to smash it in the first 15 minutes. What this means is that the authors can get more as there are more voters and they are guaranteed 50%, while the bidbots are going to take less and with more frontrunners earning more, it will be even less than they get today percentage wise.

Curators for the win?

For example, I use one bot, @ocdb, which has numbers very similar to the ones above. Here is a selection taken from the last post it voted on and you can have a look at all curation here until payout. There is a queue on @ocdb which means it always votes late however, the earliest a bot will vote is at 15 minutes because it has to get the full curation return without putting it back into the pool.

While @ocdb which voted:
3169.6 min - 43.637 SBD - 0.000 SBD - 19.254 SP - 18.1 %

The highest curator percentage:
7.2 min - 0.005 SBD - 0.003 SBD - 0.027 SP - 211.1 %

That is a much better return than delegating to a bidbot. Sure, you think it is not much but out of the 134 voters, only a handful didn't beat the 25% curation return with even large voters like

@redes
14.8 min - 0.740 SBD - 0.007 SBD - 1.795 SP - 99.5 %

@acidyo
138.6 min - 3.424 SBD - 0.000 SBD - 3.808 SP - 45.6 %

getting much higher returns voting before the bidbots. What happens if there is a 50% curation return? Will the bidbots be able to still offer the same return on investment to delegators if their curations are cut? Are they able to offer a 10% return like @ocdb does, which takes zero cut of the sends or curation? Will the curators who are curating benefit more?

As I see it, the more of the voting value that is "at risk" on the chain, the more it is exposed to curation and flags. The curation distribution goes wider and with small accounts getting a higher advantage by voting early on posts, there is more incentive to power up.

More votes on authors?

What could happen is that the initiatives that fight bot abuse will be twice as effective in flagging and therefore several factors more in changing behaviors. This could mean that the posts that do get paid votes on are higher quality and, they are more likely to have more voters on them.

Every vote that frontruns a bot eats into the curation of that bot and with 50/50, the majority of the bots earnings will be coming from curation rather than sends to the bot. It makes frontrunning them twice as lucrative for curators, and the author (unless flagged) will still get the same 13.6% return. Bidbot earnings have to go down, even though "all things remaining equal" would see them increase because, not all things would be equal as there is added incentive to trim their curation returns by others.

Unless there is only the post and the bidbot vote alone after 15 minutes, there is no way the bot will get a 50% return and with so much incentive to trim, it will be trimmed. This closes that gap between selling votes and curating again.

If rather than the 50%, the added frontrunners trimmed the bot to 30% curation.

It would look like this:

The bid is $44.
The vote comes in at 100 dollars.
Seven days later, the payout is made. The bidder gets 50% = $50 payout = profit of $6 = 13.6%
The bot gets $30 curation + $44 dollars and returns $94 to delegators.

total sends = $440
total payouts = $500
total curation = $300

There is a $200 deficit. Where does it go? To curators that aren't the bidbot. But, they have also added more value to the post itself of which the author gets 50% guaranteed - instead of just getting the $6 dollars bought.

As I see it, the more active voters the better, and while I am not great with numbers, the 50/50 incentive makes curation much more attractive while trimming the value returns of bidbots. Rather than returning 91%, they might get trimmed to 70 - 80%. That might not seem like a lot but, it closes that gap further as the front runners are going to be able to get more than the 50% and maybe average 60-70%. Still not closed is it?

But, there are also other factors to consider than straight voting as there are increasing ways to use one's vote to earn SMTs, there is value in being able to curate comments, there are initiatives that will offer SMTs for delegations and several other factors that each close that gap a little more.

While this won't "kill bidbots", what it will do is return more risk to the expected ROI, lower the returns and perhaps return some of the delegating power back into the active voting pool maing bot votes a little less powerful. Even if the old delegators set and forget on autovotes for a 50%-ish return or to frontrun bots, there will still be more votes being spread across posts that aren't paid for.

In my opinion, this is worth the experiment.

I am not an economist nor a mathematician though.
How many of you are?

Taraz
[ a Steem original ]