I've had some time to think more about @reward.app since all the comments and discussion in the announcement post recently. Honestly I was a bit surprised that there weren't as many negative comments about it than I was expecting. Some of those seemed to kind of ignore the pros and how they'd outweigh the cons and then there were a few comments that just sidetracked the issue we're trying to tackle completely and brought up a whole other spectrum of issues we can't even really do anything about. Anyway, I wanted to discuss some things in this post and bring forth a few more ideas I had about the portion that we can control (meaning the curation rewards that go through us that we then distribute extra to the authors. It's also really great that there were so many comments and even a few interesting suggestions of new additions such as being able to blacklist certain authors you don't want to share the curation rewards with and instead want them to go to the rest of the curators - although I don't see a big usecase for it now it can be something that can be added down the line for certain situations. 

<center>![image.png](https://files.peakd.com/file/peakd-hive/ocdb/cs16z1Il-image.png)</center>



One reply I received some time ago on a comment talking about the current curation activity was something that pushed me further to give this experiment a try. A user said that he really disliked being on an autovote front-runners list because it was driving away other curators from voting on his posts afterwards and that he usually wrote long form content that took a lot of time and only seeing that one curator and trail vote early for max ROI and then everyone avoid it discouraged him from continuing posting similarly. Now of course rewards aren't everything but even if you act upon it or not it can still discourage someone over a longer period of time. That's what really made me realize how backwards some curation activity has gotten where the focus is not the content or content discovery but just the highest rewards for the least amount of effort from the curator. 

As a manual curator I've always despised autovoting, not just because of them getting often times 2-4x more rewards than manual curators but also because it takes no effort - or similarly to autoposts it's just a one time thing that you get rewarded for continually afterwards. The effects it has on authors is also not great in the long run. I have nothing against autovoting working similarly to patreon (although I think hive.vote could use some more gamification and following the steps of patreon a little bit - been meaning to talk to @mahdiyari about it but have had a lot on my plate lately) and at the same time be more susceptible to downvotes and disagreement of rewards, even though that usually brings a lot of drama forward because autovotes are in one way very similar to pending author rewards - authors get used to them and no one likes losing autovotes the same way they don't like "losing" author rewards through downvotes. There's very few accounts that over time have not in one way abused autovotes to generate more and more low effort content or start posting more because they know the votes are just waiting on them thus they feel like they're missing out on rewards if they don't. This psychology exists outside of Hive as well, Twitter influencers having to tweet x times per day to remain consistent and grow that magical follower number that most of the time means nothing, youtubers having to create x amount of videos per week to keep growing in subscribers and adrevenue, etc. Like I get it, no one likes seeing numbers go down once they go up.

Without side-tracking too much as there's obviously a lot that can be discussed about the paragraph above I wanna get back to some things @reward.app could fix or at least balance out a little bit with usage and experimentation. Of course everything has a cost, giving curators a bigger share of the rewards means that some other authors out there not giving curators a bigger share of the rewards might lose out on curation themselves even if the author who does give curation a bigger cut ends up with the same author rewards in the end anyway. The inflation pool gives out the same number weekly to all accounts and it is declining in Hive every day. We need to be careful with how this is used, for one, I think there will be a lot more downvotes happening depending on the content posted while giving curation a higher cut so both authors and the curators will have an incentive to judge the content before casting their votes on it. Of course there will also be front-runners, maximizing attempts here too but since we have a lot of "centralized" control over the extra share that goes to curators there are a lot of new things that could be done with said share. 

One idea I had just today was to add additional curation formula's. Right now the default one is to give the extra curation rewards to voters in a linear fashion depending on the stake and vote weight, meaning it wouldn't count the early curation penalty or the late curation penalty. Another one that's gonna be made available is to keep it the same as the rules of the chain. So knowing that we can alter the extra rewards how we want in a way, I was thinking about different ones and one that came to mind was post length, even though I realize this could also be automated in the long run to attempt to maximize rewards but since @peakd already shows the "for example: 15 min read - 3000 words" why not make use of that to disincentivize front-running, etc. Say the curation penalty would first start at 12 minutes and gradually deplete until 14 minutes to leave some room for fast readers but anyone who voted before 12 minutes would not receive any extra curation. Or it could add some RNG to it as well, maybe a curve that goes up and down, the post would show a captcha of the best minutes to cast your vote on. I realize this last example would still encourage manual "maximization" but manual curators are already getting rekt in terms of ROI by front-runners for the usual 50%. This is also why I mentioned in a comment yesterday that this experiment would've possibly worked way better with the EIP but not 50/50 and instead 25/75 or 30/70 to give the service a bit more breathing room. 

And who knows, maybe we'll find some really great balance that would showcase that 50/50 is not the best default split but starting at a lower curation cut and letting authors decide it themselves could be more popular (as was discussed a year or so ago) and maybe it might be implemented in a future hardfork. That's what is so great about the experiment that we can try our way through and explore if there are any better balances and as I mentioned this service is something that was inevitable. Anyone could've started it at any time.

Anyway, looking forward to see how this develops and continues forward, if you guys have any more ideas you'd want to explore with @reward.app feel free to let us know and we'll see if it's worth pursuing at some point down the line.

Thanks for reading, and in case there's already some bots attempting to maximize curation rewards on this post, I'll let you humans know that I set the extra curation rewards of it to ![image.png](https://files.peakd.com/file/peakd-hive/ocdb/eF7EGCZ5-image.png)

<center> :P </center>

___

https://images.hive.blog/0x0/https://files.peakd.com/file/peakd-hive/hiddenblade/i2W9CjnS-Acidyo.jpg