One thing I can give credit to autovotes is that those who use them well, i.e. monitoring them often (and the authors on them) and not just doing it for max ROI (by giving very small percentage votes) do very well for retention but as said giving big votes will only be able to cover a certain amount of accounts. The downside of manual curation which is what I've done in forever now is that it goes in periods, some times you curate 10 certain accounts a lot while spreading the rest around and then a few of them drop/get forgotten/voting power just isn't there/i just don't happen to see their posts. Yours a month ago for instance I think someone had reblogged, I remembered your username from the steem.chat days and threw it a vote but somewhere along the line I must've unfollowed you (i sometimes prune my following based on inactivity but haven't done so in over a year now cause I figured it's better to just leave them there in case they come back).

Anyway, what I'm trying to say is that while focusing mainly on one community (OCD) for a while it felt great cause me and a couple other curators from the team were able to check each and every post posting there (while the rest of the curators were curating outside our community for the daily @ocd compilation posts), but then we decided that having a general "original content" community was not the best idea - thanks to those that made me realize to stop it as early as we did cause switching curation to the more niche communities has done wonders for them and for communities as a whole. Point being it gets more difficult to cover same authors more often no matter how consistent and great quality they produce. One thing that's great is that many accounts like to "spy" on what we curate and set autovotes on those accounts to maximize rewards and we usually don't care too much if posts already have a lot of rewards on if the content is great - although as a curation guild focused on newcomers and underrewarded content we tend to curate that but there's a big lack of late voting going on because they know they may end up being the last voter and receive minimal returns.

Man this comment is getting long, I recently posted about something that may help this whole dilemma I mentioned at the end of the last paragraph but it's still a very experimental service we have to monitor closely to avoid abuse, at the same time we don't want to focus curation on those using it cause it will seem biased and that we're doing it for the extra rewards - even though that has almost never been our main focus as we curate manually and 99% of our votes are not cast within the first 5 minutes.

I think the way rewards work right now we're slowly entering a phase where getting consistent good rewards will only land on a few people on the platform and getting high rewards will be even harder. I assume if Hive were to pump there will be a lot more usage of downvotes too because there's already so much talk of "overrewarded posts" even at $60 where as only $30 go to the author and half of that is stake, so you can imagine that if Hive pumps 5x again those rewards will also go up by quite a lot so your content will have to be very outstanding to constantly receive a lot of rewards. In @pitboy's case here he's shown that he really knows what he writes about and produces results, people like to reblog his content or like in this post's case share them in chats so other curators can trust the sharer and some times not even need to read every post to know it deserved the rewards it's getting. So yeah, we're kind of moving in this zone where stakeholders are kind of being forced to hold more stake and curate with it since curation rewards is now 50% and that's something that is way more guaranteed than post rewards - not just because of how often curators find you but also cause of downvotes that can happen at any time for any reason - although the latter is really a low percentage of cases.

Okay, I kinda lost track of where I was going with this so better to end it here, lol.