(image source)
Yesterday I have decided to go Whale watching here on Steemit as a part of my ongoing pet project @robotev for a useful bot on Steemit that should help discover good posts from new authors that have not established their name on Steemit yet, but are showing potential. So I have decided to put the code of the bot to a more serious test in following most of the top Whales on Steemit and work with their votes to see how good it will perform compared to the list of volunteer curators that the bot normally follows.
The bot has been operational for more than 12 hours watching whales, so it is time to post some interesting findings and stats on how it has performed so far. It follows most of the top whales with a total of 68 users, though not all probably. Have in mind that not all of them have been active in the last 12 hours that the stats below covers, though more than half of them should have been.
Time for Some Stats
Total number of votes the bot would have made: 171
These are the number of votes that my bot would have made looking for good posts from authors that could use some support in getting encouraged and help them get noticed. This is not yet 100% accurate filtering, but most likely 60-70% of these would've been considered good hits if the bot code was optimal. Of course all the collected data goes to further improving the performance and even for just 12 hours it has helped me get quite useful data that will help.
Total number of votes that the bot would have ignored: 192
These are the number of votes that my filtering would have excluded as being made by popular authors and people with high reputation already on Steemit that don't need extra support in getting where they are already. As you can see this number is higher than the previous one and since my filtering is not yet complete and 100% accurate the actual difference is even higher in favor of popular authors.
Total number of posts that two or more Whales voted for: 92
This is a very interesting number covering only the first category of posts that the bot would have voted for tracking the ones that more than one Whale voted for... usually many of these get between two and about 6 votes from Whales. I'm sure that the filtered posts with popular authors do get much more Whales to vote for than the new discoveries.
Some Observations so Far
In 12 hours some of the Whales did not vote for more than 1-2-3 posts, quite a few did not vote for any new and unpopular authors, but only for established names. As one would expect Whales on Steemit do like to swim together and vote together, some of this is probably automated via bots as well. The number of Whales monitored that have voted for unpopular authors is 27 while the ones that voted for popular and unpopular authors is 36, so 9 Whales voted only for posts from authors with already established reputation.
Some people think that Whales do not vote for posts made from authors with a low reputation, lower like forty for example. That is not true however, they do vote for such, though not that much, but because of their high reputation (mostly) even a single Whale vote can quickly boost the reputation of a new lucky author to over 40 most of the time. Of course Whales do tend to vote more on higher-reputation authors and that probably is also one of the criteria that they do use when checking the list of posts and deciding which to open and actually read. This probably makes it harder for new authors with low reputation even if they have written something really good, unless they do come up with a catchy title or an interesting header image that is attracting attention as a thumbnail as well.
There are some concerns that there are holes in the Whale activity on Steemit, so properly timing the posts you make might help you get noticed then if posting whenever you want. While this may be true, it is also true that if your post does not get a lot of attention in the first 30-60 minutes then it probably won’t matter much when you posted it. Whales do seem to check older posts as well, so they can still spot it and vote for it should they decide it is worth of their vote.
Do note that the above information and cited numbers are based on not yet perfected filtering algorithm, so they reflect the numbers of current not yet optimal filtering capabilities at the moment!
The Conclusion
I still have some more work on the bot code to optimize its performance, but even after 12 hours of monitoring Whale activity it has helped me gather interesting and useful data that should help a lot. I'm going to continue with my Whale Watching activities on Steemit and will also post more interesting findings and stats in the future as I gather more data...
If you have a question or want to add something, then please leave a comment below.
Did you like what you have just read? Check my other posts on steemit @cryptos