When do whales upvote relative to publication time?

in stats •  2 years ago

In when do whales upvote? I looked at the upvote patterns of a subset of whales to see which days and which times they tend to upvote.

Since writing that post I've been wondering how soon whales upvote after a post has been published. Are they focusing on curation rewards like most regular users, or are they more benevolent?

To answer the question I wrote a program in Python to plot the upvote times relative to the post's publication time for the period 03-09-2016 to 17-09-2016.

The founders

Since writing the last post I've discovered that @dan and @dantheman are the same person. It appears that @dan uses @dantheman a lot more, so only data for this account is shown.

@ned and @dantheman have very similar upvote patterns. The bulk of their upvotes are after the curation penalty cutoff and before the first payout - 30 minutes and 24 hours respectively.

In the largest window their upvotes are also distributed similarly, with more upvotes being placed in the 7 hours of a post's lifetime.

@ned and @dantheman are also the whales who upvote the most after the first payout. Sadly, this is only 8.47% and 6.98% of the total, respectively.

Whale bots

These are the same bot accounts as in the previous post, with the addition of the infamous @wang.

It's no surprise that these three bots, owned by the same whale, have very similar upvote timing. Additionally, the distribution is a lot flatter than the other whale accounts.

@wang is the outlier in the data set as the only whale that upvotes more during the first 30 minutes, especially the first 15 minutes, than at any other time. This means @wang forfeits its own curation reward, giving post authors an extra boost. @wang has one of the worst reputations on steemit, yet gives the most back in terms of curation reward.

Curation

@curie and @robinhoodwhale both exist to help lesser-known authors get recognition for their work. It's worth stressing that neither of these accounts are whales, but they are helped by whales. Sadly @curie doesn't have much data yet, probably because all the upvoting is done by proxy, but @robinhoodwhale is going strong.

Others

Here are the charts for the other whales. If you think other whales should be included in reports like this, please let me know.

These charts show a big difference in vote distribution, even though the share in terms of curation penalty and post payout are similar. While @berniesanders and @nextgencrypto's distributions gradually tale off as a post gets older, @blocktrades limits voting to between 30 minutes and 3 hours.

I know that @berniesanders and @nextgencrypto allocate voting power to other projects, such as @curie, so this might be an explanation.

The remaining charts from @complexring, @pharesim, @rainman, and @smooth show the same pattern where more upvotes occur between the curation penalty cutoff and the first payout. All four show a greater number of upvotes after the first payout than by other whales (excluding @dan and @ned), but compared to other timings, it's still a very small amount. @smooth differs every-so-slightly in that more upvotes are placed between 15 and 30 minutes.

Summary

Most whales upvote after the curation penalty cutoff, taking all curation rewards for the upvote. The only exception is @wang who upvotes before the penalty more often than not. Some whales like @smooth and @pharesim do have a large portion of votes between 15 and 30 minutes, but the majority for them is still after the penalty.

Perhaps the most striking pattern is how little whales upvote posts after the first payout. There are a few possible explanations:

  1. After the first payout there's no curation reward - curators nothing for upvoting. This removes all incentive from upvoting after 24 hours;
  2. If the post wasn't good enough to vote in the first payout cycle, why would it get upvoted in the second?
  3. Posts after a certain time are no longer easy to find due to newer posts.

Whatever the reason, it makes me wonder what the purpose of the second payout is if the whales themselves seldom upvote after 24 hours. I can imagine that the same pattern will be found in most accounts on steemit.


Follow me to make sure you don't miss my next post

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Hello @bitcalm,

It gives us pleasure to inform you that you have been chosen as a featured author by the @robinhoodwhale initiative.

Learn more about the Robinhood Whale here!

We hope to see you continuing to post some great stuff on Steemit!

Good luck!
~RHW~

·

I don't think this could come at a better time. I've been feeling quite demoralized on steemit lately. Thank-you for the recognition, I really appreciate it.

wow what a great job you did here

@robinhoodwhale's stats would appear to be heavily skewed by its votes on its own comments when informing the author of their selection. Curation is manual and within the first 15 minutes would be extremely rare. Especially since there has to be at least two members of the curation team supporting the proposal to vote on a post.

·

Thank-you so much for this comment. It almost made me cry when I realised, in true face-palm style, that I forgot to filter out upvotes on comments.

I thanks the steemit devs for making posts appear in the blockchain as a comment! I shouldn't have to think about this, but maybe there's a good reason they do this.

In any case, everything is up to date. You were 100% correct. @robinhoodwhale doesn't upvote in the first 15 minutes at all.

I'm glad there are people who think a bit when they read my posts. I make mistakes like anyone else, and I'd rather they be caught than mislead people.

·
·

Wow, what a quick turnaround! Thank you for this. Such an interesting post.

·

Curie also votes on some of Robin Hood's posts. We're all in this together.

·
·

I suspected as much. I'm surprised curie upvotes so little though. I would expect curie to vote more in order to power up and become stronger, and then be able to help more.

Great analysis !

Most whales upvote after the curation penalty cutoff, taking all curation rewards for the upvote.

This also increases the curation reward for the early curators.

·

This is a great point. An altruistic whale could be seen as a whale who either upvotes very soon to give their curation reward to the author, or a whale who votes very late to increase the reward of previous curators.

·
·

It would be interesting to see your analysis of the voting patterns regarding comments/replies.
Your analysis is only for root posts, right ?

·
·
·

Very smart question. Most people just look at the charts and the conclusion.

The intention was to make it about root posts, but I forgot to filter. Luckily the comment from @bacchist alerted me to my mistake, and the charts are now up to date.

Great critical thinking. I'm following you now.

·
·
·

I wouldnt find that analysis interesting though :) but nevermind me...

·
·
·
·

From the whitepaper:

One of the primary goals of Steem's reward system is to produce the best discussions on the internet

That's why I find that interesting.
I don't see good comments being rewarded

Good discussion requires back and forth posting. When you reply to someone else they get 50% of any payout you receive in that thread [...]

This hasn't been implemented yet, or I don't understand it right ...

Until then it would be nice to see, why and how comments get a reward.
Especially from whales.

·
·

I think I missed the memo on the timing of a vote...I thought early voting was the key to curation rewards. This is saying you need to wait 15 minutes to avoid a vote "penalty." Would this have something to do with the 5 vote limit for users above 300 SP? Help. :0)

·
·
·

The penalty is at 30 minutes. It was added because the bot @wang (I think) gamed the system and was upvoting as soon as posts were published. Keep in mind that those who vote sooner are rewarded more than those who vote later. It's a reverse auction. Yes, it's also complicated.

·
·
·
·

So I guess the question is, WHEN should I vote? Can you point out a good source of info. on the subject, for a minnow? I guess it depends on the article, whether it's new or hot....but I thought I had this figured out. Now....not so much.

·
·

I'll be following you. This post was certainly enlightening. You put a lot of effort into it.

So hepful your stats. Thanks so much for the effort. Dolphins can also have an impact which would be also interesting to see.

·

Absolutely. It's not all about whales. I'll do some stats in future to concentrate more on the other user types.

Great post and analysis.

it makes me wonder what the purpose of the second payout is if the whales themselves seldom upvote after 24 hours

I had exactly the same question in mind. As there is no curation reward after first payout, there is no real incentive to upvote "older" posts.

The drawback is that the very good posts are forgotten as soon they pass the 24h milestone. This lead to over-production of new and new posts, with lower quality, which is quite the opposite of Steemit goal.

I have not seen this in the past two months I have been here. Great info and I have saved this post. Thank you. You have a new follower.... :)

Sauber!

Am I reading something wrong or is there something wrong with Smooth's charts.... They don't seem to suggest the same thing... The first suggests that smooth votes mostly between 15-30 minutes while the pie chart suggests they vote mostly after 30 minutes.

·

Maybe I need to explain the chart better in the post. The bar chart shows vote in the 30 minute to 24 hours range in one hour bins. If you sum the values for rows 3 to 26 in the bar chart, it equals the 30 minutes to 24 hour wedge.

·
·

Oh OK, it just didn't look that way at first glance but I see now. Thank you

Some nice stats and one of the best I've seen for a while. Great post and deserves an up-vote. Following and looking forward to reading more of your stats. I joined STEEMIT a few weeks ago and I have seen a remarkable increase in the quality of posts. An amazing concept. I recently posted a couple of articles about STEEMIT, the future of BLOCKCHAIN, Social Media CryptoCurrencies, and one about Dan & Ned. You may find them interesting to read. You can also catch us Twitter✔. Cheers. Stephen

That's a lot of data gathering. Thank you for posting

·

I updated my post to add that @curie very likely only uses proxy voting at the moment.

·
·

Yes, I just meant respect that they vote so early and reward the authors more!

Interesting post....

I think the biggest reason posts don't get upvoted after 24 hours is because they are so hard to find after that long... there is an ever increasing supply of new and fresh posts coming at us at all times no matter how great the previous ones were... maybe the reblog feature will change that slightly?!

·

I wonder how many people would still upvote even if it's in their feed from reblogging. It costs voting power.

·
·

True... do you? I have upvoted a couple already that are in my feed that were reblogged but I don't look at them as favorably as new posts presented by the people that I follow. I personally wish there was a tab for all of the reblogged posts and then a tab for all the fresh posts... ie separate them.

·
·
·

I do but sometimes, esp. if my voting power is down, then I don't. Terrible isn't it?

I don't understand the curation time windows? So it's beneficial to wait to upvote? So then we we should wait to upvote our own post?

·

During the first 30 minutes of a post's life the curation reward from upvotes is shared with the author. The closer to publication time, the most of the curator's reward is given to the author.

I'm not sure what the effect upvoting on your own post has. I recall a post that said it's better to wait, but in my experience I haven't noticed anything, and I never got around to finding out more.

Interesting stats.

Whales really do hold our destiny in their hands (flukes?). They can vote mainly for their own benefit or choose to 'bless' minnows with a big payout. How they behave will affect the future of Steemit. I know a lot of people feel it's rigged and that will put some off. I can see that there are efforts to share the bounty more by rewarding good content. I'm hoping that and adjustments to the algorithms will create an economy that will attract good content producers. That's what it needs to succeed.

A lot of posts may not get noticed in the first 24 hours. They may get 'reblogged' later and I hope that will result in them gaining votes. I think it should be possible to earn on posts beyond 4 weeks as they will still appear on Google after that and may still have value then.

·

This post is focusing on when whales upvote, in general, relative to a post's publication time. It's almost impossible with this data to say why the whales vote in this way. Since there's no curation penalty after 30 minutes, it makes sense that most whales vote at this point. You could ask that since whales have already made it, why aren't they more altruistic, sacrificing their own curation reward for others?

I don't think posts should be rewarded after 4 weeks because that's probably not good for the steemit economy. Don't ask me why, it's just a feeling. I agree that the reblog feature will help get posts "out there", and it emphasizes the importance of building a good following rather than trying to "get rich quick".

·
·

I realise we can't deduce intention from this data. Thanks for doing this anyway. More data is always a good thing and I'm sure people will draw their own conclusions. Cheers

Nice analysis

Excellent information as always! Thanks for the awesome post!

·

Thanks! :)

This article has great data that I have not seen elsewhere. Thanks for this!

I've gotten a few (bigger) votes after the first pay out period, the second pay period just affords the opportunity for posts that were missed to get rewards. I like that it's an option.

·

An option that few people take. I'm not against it, but I'd like to see people use it more often.

@wang is the outlier in the data set as the only whale that upvotes more during the first 30 minutes, especially the first 15 minutes, than at any other time. This means @wang forfeits its own curation reward, giving post authors an extra boost. @wang has one of the worst reputations on steemit, yet gives the most back in terms of curation reward.

I feel bad for @wang - he was the first bot that welcomed most people and the data shows he was actually attempting to give something back to the authors too.

·

I don't think @wang is really thinking of other users, it's just the method the bot writer chose to try and maximise curation rewards. The effect of which is that the curation reward is split with the author. From an author's perspective, compared to other bots, it's certainly better. I agree that @wang's reputation is a bit extreme considering other bots out there.

·
·

Yes that's what I meant. The operator obviously considered this and decided to be more altruistic than the others but the reputation is not consistent with that.

@bitcalm, great post. Didn't see it till after the first payout....Upvoted anyway! Keep up the good work. I know that grind and the feeling of relieve almost when a post gets some love.

·

Thanks, but it hasn't paid out yet :) It got extended because of a whale upvote I think.

General advice: label your axis.

·

I know I committed a cardinal sin of graphing. I took the labels out because at first the y-axis differed for each user. When I fixed that, I forgot to put the labels back in, and it takes 20 minutes to generate.

Dude you rock man ! Love your Stats , please keep this up ! you got me as your follower

·

Thanks. It's fun making them :)

I'm also curious if anybody receive at least 1 dollar after first payment is made