Bad effects of the small 'curation window'.

in #steem5 years ago (edited)

STEEM is just great,
Will go to the moon,
Dont be afraid,
We will be rich very soon!

As you can see above, I really gave my best, but praising something has never been my strong point so that I better stick with what I am expert in: to criticize. :)

Curation still doesn't work as intended.


At first I would like to express that altogether my experiences with EIP haven't been that bad. I would say that the current state of STEEM is at least not worse than it was before HF 21, and some things have clearly improved (which, however, are not my topic today).

There won't be a German version this time, and instead of that you may read this comment, written in German language.

My point today is that I don't really share all the excitement about how great 'curation' would work nowadays. Yes, there is the improvement that some whales and former bid bots have changed their voting behavior, but at the same time I think most of them are just auto upvoting a certain group of authors while they could upvote far more different, and especially also not so well known, small accounts. One of the positive exceptions in my eyes is @curangel.
Anyway, it's still the case that well known Steemians earn more money with very trivial and rather effortless posts like "Wow, I sold Steem Monsters today, and their price is so high now!" or "Earn Passive Income with Bitcoin ..." than less 'famous' users could ever make with ever so elaborated and well-researched articles.

Bad effects of the small 'curation window'.


Source: pixabay.


I think one of the main problems is the 'curation window' of only five minutes. Nobody who is upvoting manually rather than in an automated way is able to find new good content within the first five minutes after its appearance ... and if he could he still had to read and evaluate the post within a few minutes (do you really think that's possible? Come onnn!). :-)

That means most curation rewards aren't earned by real, manual curation but by auto upvoting the same authors again and again, and most of the time the 'curator' doesn't even know what he is actually rewarding.

As a consequence I upvote comments.


As someone who only upvotes manually, I have not the slightest chance to find, read, evaluate and possibly upvote new posts within the first five minutes after they have been published on the blockchain (especially as I cannot sit all the day in front of my computer lurking for new posts). I know, I am a very slow person, but be honest: YOU also can't! :)

I can't spread my upvotes over the whole day but place most of them within a narrow time frame, and most of the time when finding great stuff from at least somewhat popular authors, the reward already exceeds ten or more dollars, so that concerning curation rewards it's not really worth it anymore. As people are so triggered to vote as fast as possible, my upvotes often are the last ones at all. :)
As a consequence I started to upvote either great posts of unpopular authors or comments (instead of the posts itself) in great content of popular authors, because anyway I have no chance to compete with all these auto voters and also don't want to further reward them for upvoting posts as fast as possible without reading anything! As I always want to know what I am actually rewarding, I won't use any auto upvote option myself.
From now on it will be an exception for me to upvote posts with a pending reward greater than ten dollars!

Possible solutions.


I think ...

  • Either there shoulnd't be a curation window at all, whereby then the curation reward should only depend on the vote weight but not on the date of the vote. In that case ones decision should depend on the quality of the post only and not on the fact who else has already (not) upvoted the post.
    @pharesim opposed that then the 'maximizers' would simply upvote 10 random posts per day, to minimize their effort. Possible ... but then I ask why we are having free downvotes nowadays ...? :)

  • Or we could think about a really long curation window of, lets say, a whole day or even more. In that case it wouldn't be so obvious when upvoting would lead to the highest curation reward. So hopefully it would be worth to concentrate on the quality of content again, instead of the date of one's upvote.

I tend to prefer the first option. Anyway, in my opinion the problem how to encourage Steemians to seek and upvote quality content is still not completely solved.

Sort:  

The best (Steem related) post I've read in weeks!!

I could also very well live with option 1, but have no problem with 2 either tbh.

"Either there shoulnd't be a curation window at all, whereby then the curation reward should only depend on the vote weight but not on the date of the vote. In that case ones decision should depend on the quality of the post only and not on the fact who else has already (not) upvoted the post.
@pharesim opposed that then the 'maximizers' would simply upvote 10 random posts per day, to minimize their effort. Possible ... but then I ask why we are having free downvotes nowadays ...? :)"

But wouldn't be the result of this that curators wait till the last day and curate than the post with the highest payout ?

But wouldn't be the result of this that curators wait till the last day and curate than the post with the highest payout ?

If everybody waits, nobody can know which posts will have the highest payouts, :-) and in addition, if a high payout is divided by many curators, it doesn't necessarily mean a very high curation reward for the single user ...

OK good point, but it will be two advantages stay in my opinion:

  1. You see if a post has reached the minimum payout of 0.02 (more important for smaller accounts).
  2. You get the reward much quicker if you vote at the last day than at the first day.

Well, right, but don't forget what my idea is about: to make it less attractive to vote as fast as possible, even without reading anything.
I wouldn't mind if people voted late to get a faster payout, that's alright. :)

There shouldn't be any curation window. We have life outside of the steem world.

Posted using Partiko Android

Good point!

I would like to start with the statement that it was no mistake to vote for this article with the Autovoter and to have read it in peace only a few hours later. ;-)
That's how I've been doing it for many months now anyway. Therefore the update feels rather good for me regarding the vote. So far at least.
Of course you can't study longer articles in 5 minutes. And what is, if 2 good, longer articles appeared at the same time, then one of them would have lost in any case. On the other hand, the 30 minutes were already tight. When I was new here, I found it totally illogical that not even older articles are still to be voted. This circumstance makes the whole thing here a rather fast-moving thing. But it is just like that.... ;-)
I also have to admit that I'm not the expert who thinks about how things could go better, but that doesn't mean that I only want to behave like flotsam in the ocean.

Beginnen möchte ich mit der Aussage, daß es kein Fehler war diesen Beitrag mit dem Autovoter zu voten und ganz in Ruhe erst ein paar Stunden später gelesen zu haben. ;-)
So handhabe ich das ja sowieso schon seit mittlerweile vielen Monaten. Daher fühlt sich für mich das Update bezüglich des Votens eher gut an. Bislang zumindest.
Natürlich kann man längere Artikel unmöglich in 5 Minuten genau studieren. Und was ist, wenn 2 gute, längere Artikel gleichzeitig erschienen, dann hätte auf jeden Fall einer davon verloren. Auf der anderen Seite waren dafür ja auch die 30 Minuten schon knapp bemessen. Als ich hier neu war, fand ich es total unlogisch, daß nicht auch ältere Artikel noch zu voten sind. Dieser Umstand macht das ganze hier zu einer eher schnelllebigen Sache. Aber ist halt so.... ;-)
Ich muß auch gestehen, daß ich nicht der Experte bin, der sich zu dieser Sache die Gedanken macht, wie es besser laufen könnte, was aber nicht heißen soll, daß ich mich nur wie Treibgut im Ozean zu verhalten gedenke.

I would like to start with the statement that it was no mistake to vote for this article with the Autovoter and to have read it in peace only a few hours later. ;-)

What if it had turned out to be a very bad post when reading it later? OK, you could have withdrawn your upvote then, but lost curation reward by doing that. Also rather often auto upvoters don't read the posts later but never.

And yes, of course it's true, I also profit from auto votes.

And what is, if 2 good, longer articles appeared at the same time, then one of them would have lost in any case.

If there wouldn't be a 'curation window' (or a very big one) both articles could be upvoted with full strength.

On the flipside, you would be 'punished' for manually finding a post early and voting for it, then.

I think I have a knack for who I can bet on the autovoter and who I can't. So far I haven't had to withdraw a single vote. ;-)
That means, no contribution was so "bad" that I had to withdraw my vote. And yes, I would like to give out even more 100% votes. A little too low votes I have already increased several times. Then I have renounced the rewards, which is also nothing so much in the weight of my manageable VP. One must die a death, nevertheless, always. I cannot possibly please everyone. My method simply seems to be the best for me, which of course does not mean that my approach is also suitable for other Steemians. If you should consider an autovoter, I can recommend steemauto.

Ich denke, ich habe ein Händchen dafür, wen ich auf den Autovoter setzen kann und wen nicht. ;-)
Bislang mußte ich noch kein einziges Vote zurückziehen. Das heißt, kein Beitrag war bislang so "schlimm", daß ich mein Vote zurückziehen mußte. Und ja, ich würde gerne noch viel mehr 100% Votes vergeben. Etwas zu niedrige Votes habe ich schon öfter mal erhöht. Auf die Rewards habe ich dann verzichtet, was bei meiner überschaubaren VP aber auch nichts so sehr ins Gewicht fällt. Einen Tod muß man doch immer sterben. Ich kann es zudem einfach unmöglich allen Recht machen. Meine Methode scheint für mich einfach die beste zu sein, was natürlich noch lange nicht heißen muß, daß meine Herangehensweise auch für andere Steemians die passende ist. Falls Du doch einen Autovoter in Erwägung ziehen solltest, kann ich Dir steemauto empfehlen.

If you should consider an autovoter, I can recommend steemauto.

I use steemrewarding to follow my manual upvotes which I place with one of my small accounts.

I have stopped discussing the rules since a long time. You probably are right, but what will be the consequences? The main problem for the price of steem is the absence of big investors. If they won't appear, the value of steem will go to zero and the blockchain has to be terminated, because the witnesses will not continue their work without earnings.

Don't get me wrong, I didn't claim, just to change the 'curation window', and then all problems would be solved. It is an idea which could (but not necessarily would) improve the chances of 'quality content' getting higher rewards (I am curious what other users are thinking about it).
And In theory investors could for example be attracted by a better looking trending page ...

Changing many small things sometimes has an even bigger effect than changing one big thing ... That's why I care about details like this.

Loading...

Hi Jaki,
du magst mir verzeihen, dass ich eine Antwort besser auf Deutsch formuliere. Vermutlich wird es außer dir eh niemand lesen, von daher ist es den Aufwand wohl auch nicht wert.
Meine Laune ist gerade eher auf einem tieferen Punkt angelangt, was diese Plattform hier betrifft, dennoch werde ich mich bemühen deinen Post sachlich zu betrachten und nicht zu sehr eingefärbt von meiner momentanen Gefühlslage.

Zum einen Stimme ich dir zu, das Curation Window von 5 Minuten ist einfach viel zu eng gefasst und macht es bei längeren Texten schlicht unmöglich, den Beitrag erst zu lesen, zu bewerten und dann zu voten, vom Entdecken dieser erstmal ganz abgesehen.
Auf der anderen Seite bin ich mir ziemlich sicher, dass dies nicht die Quelle allen Übels ist, bei Weitem nicht. Ich habe deinen Artikel ca. 7 Stunden nach der Veröffentlichung entdeckt, er wurde zu diesem Zeitpunkt bereits von fast 180 Accounts gevotet und war bereits bei 19 Dollar. Da ich eh nur sehr geringe Beträge Upvoten kann, sind meine Curation Rewards auch entsprechend niedrig, so kommt es für mich nun wirklich nicht darauf an, wann ich vote. Jetzt versuche ich mich aber in die Lage eines Wales zu versetzen, der sein Investment schützen, bzw. ausbauen möchte. (Sehr einfach und plakativ geschrieben, aber sonst wird das hier ein Roman ;-))
Der Wal denkt sich jetzt „Verdammt, wäre ich da doch mal eher ‚reingegangen‘“ und richtet sich vermutlich ein Autovoter für dich ein und stellt ihn auf vielleicht 4 Minuten. Nun freut sich der Wal in der Zukunft, weil er immer untern den ersten dabei ist und da deine Artikel ja immer mächtig abräumen, braucht er auch gar nicht mehr überprüfen wie das so läuft. (Bis zu dem Punkt sind wir uns sicher einig, wie gesagt, sehr vereinfacht)

Jetzt stellen wir uns vor, das Fenster läge bei 24h. Der Wal denkt sich: „Oh blöd, auf welche Zeit stelle ich denn jetzt meinen Autovoter.“ Je nachdem wie faul oder fleißig er ist, wird er vermutlich ein bisschen ausprobieren und sich dann irgendwann auf eine Zeit einpendeln. Vielleicht ist der Gewinn nicht mehr so große, aber immerhin bekommt er ein gutes Stück vom Kuchen ab, weil der Post ja immer noch hohe Rewards bekommen wird, darauf vertrauen ja alle.

Das nächste Szenario ist wohl, dass es keine Fenster gibt, die Rewardskurve ist aber noch die gleiche. Der Wal denkt vermutlich: „Bloß kein Risiko eingehen, besser nur Posts upvoten, die auf jeden Fall über die ‚Hürde‘ kommen oder besser schon sind.“

Als letztes Beispiel, kein Fenster, kein Rewardkurve, sondern eine Gerade. Die Rewards wären dann ja theoretisch immer die gleichen, egal ob ich beispielsweise einen Kommentar alleine vote oder auf einen Post, der schon extrem hohe Rewards hat. Ich lasse mal die Downvotes außen vor, da die ja auch noch einfließen müssten. Nun kann der Wal sich denken: „Super, ich bin frei und kann voten was mir gefällt der Gewinn bleibt ja der gleiche“ Er schaut sich um und entdeckt ein paar tolle Post, votet diese mit 100% und alle sind glücklich. Ich möchte das wirklich gerne glauben, aber er könnte auch denken: „Super, ich bin frei, ist doch vollkommen egal, was ich vote, der Gewinn bleibt ja der gleiche …“ Es wird vermutlich beides vorkommen.

Oh, das ist jetzt ein bisschen länger geworden, als beabsichtigt (da war es wieder mein Problem mit dem Kurzfassen). Ich denke, es wird zu wenig gelesen und ich verstehe das sogar, denn auch mir fehlt häufig die Zeit. Ich habe mal eine ‚Testballon‘ dazu losgelassen, um mir ein Bild zu machen, aber das führt jetzt hier zu weit. Auch ich sehe im Augenblick leider keine Ideallösung.
Liebe Grüße

Loading...

the problem how to encourage Steemians to seek and upvote quality content is still not completely solved.

It's not solved at all. After the hardfork and the birth of this #newsteem I thought that this is it. Owners of bid bots changed to manually upvote posts on Steemit and we won't have shitty posts boosted on the trending page or circle jerking. The part with boosting shitty posts on the trending page is kind of solved although some posts in my opinion have too much weight on them comparing with others that might have spend an hour or more to collect information and serve us a well documented article. Not in a thousand years in my opinion a vacation post with four photos is worth $40-$50.

Regarding circle jerking unfortunately it has survived but somehow is not something that involves only a group of whales or something similar. Now "thanks to" to the upvoting window you mentioned about the former bid bots in order to maximize their curation have simply put some users on auto vote and the job is done for the day.

I have had some upvotes from bid bots as well but not too many and nothing spectacular and it's probably that my blog is not spectacular :)). I would be for sure more motivated to invest much more time in my posts and really work for my blog but when I see almost daily payouts of $0.5 or 0.6 and very often zero comments on the post I get discouraged to put any more efforts.

The most upvoted posts nowadays seem to be the ones in the #travelfeed category and people started to post from their vacations three years ago or make some ten posts from a vacation to get some attention which leads users to try and fit a pattern which is not quite a "decentralized attitude" in my opinion.

Bottom line is the same convenience that auto votes offer to maximize curation profits that narrows the circle of users having a chance to big upvotes. Thanks to it quite a few good bloggers remain under the radar while others are constantly auto voted no matter what they post about. Steem has some work left to be done on it to become great. It's just a few tweaks here and there.

Good points!

(By the way I am also making different posts of one single vacation from time to time, but my main motive is to save all these precious memories on the blockchain - and the pics at least on a second database.)

I have such posts as well, but never got to the trending page... Lucky me :))
The idea is that there are tons of good meaningful posts besides the photography ones that get unnoticed thanks to auto votes determined by the curation window that you wrote about and also by maximizing curation revenues through this damn window. Bottom line this place will never be perfect in my opinion so we will have to learn to live with it as it is.

It will never be perfect, and every interesting suggestions has also negative aspects, but I still think we should try to improve things as good as possible.

These are some interesting ideas.

My idea would be to enable a function that lets you set all rewards to beneficiaries after a certain point.

For example if rewards > 10 SBD set beneficiary @steem.dao 100%

This would serve to purposes:

  1. It would allow authors to set the desired value for their blog or post and gain a better sense of quality as their target is eventually realistic.
  2. It would allow curators to better determine if a post is over rewarded or not and move on.

One could also incorporate it with your idea of changing the curation level or window accordingly. I can also suggest that we can even allow authors to set the curation window and whether it is diminishing or not or adjust it based on their average rate of curation so really popular bloggers can experiment and such.

Interesting ideas!

No matter how long the curation window is, bots will always win, so I think maybe solution 1 to completely remove the curation window and make votes not dependent on time (order) would be a reasonable solution.

Even if bots would always win anyway (I am so sure if that really has to be the case), at least auto upvotes wouldn't always win against manual curation anymore.

But yes, maybe no 'curation window' would be the best solution.