Posts

Updating the Engagement project for POB - 50 engagers gets upvotes instead of 25 from now on.

avatar of @amr008
25
@amr008
·
·
0 views
·
4 min read

Good morning to everyone , hope you are doing good .

Today it is not an analysis post , it is more of a update for the engagement project.

POB

POB activity is growing day by day , the number of posts and comments made from that particular front-end is increasing .

So I thought of encouraging people more by upvoting not 25 users who engaged but 50 users .

Why the change?

These are the reasons

  1. There are many users who are leaving 5-6 quality comments everyday whom I want to support and recognize . I want to tell them that each and every quality comments matter .

  2. This might spur more engagement as 50th position gets less upvote than 49th . So the 50th person who gets good amount of upvotes for 2-3 days might put more effort to increase engagement .

One of the main reason apart from above 2 is - @amr008.pob holds a lot of POB right now compared to most other users on POB .

We even crossed @pob-fund and are sitting at 6th position .

Right now the 100% upvote is worth 34 POB which is like 50 HIVE ( 25 HIVE for author ) .

Since POB is still in early stages we should diversify the distribution . It shouldn't be in the hands of few people . That doesn't mean we just give it away to everyone who doesn't put the efforts , let's distribute to those users who are putting efforts and contributing to the POB ecosystem .

Changes

  1. Not only the 25 to 50 users change , I just updated the script to make sure the quality comments gets picked up most and the upvotes distributes properly based on the effort .
  2. This doesn't mean those who are putting more efforts won't get more upvotes , they will . At the end of the day , it all depends on the final "quality" value that I calculate . ( I will explain this below)
  3. I might reduce or increase the number of users for POB in coming days based on certain analysis .

Today's upvotes for POB and explanation

Here is the full details of today's upvotes. Let me explain certain things a bit .

  1. Sumlen - Total characters types by user ( this doesn't include "URLs" , "Posted via proofofbrain" it is automatically removed by the script and then the characters are counted)

  2. CC - Comment count

  3. Parent_c - Number of authors you talked to

  4. Unique words - Number of unique words used ( all comments of users are considered here)

  5. Unique comments - Number of unique comments made by the user

  6. Similar comments - Number of similar comments made by the user .

    • Note: These are the steps taken to check for similarity
      1. Your first comment of the day is picked and checked with all other comments of yours . If the first comment matches 65% with any other comment , then it is considered similar .
      2. If your first comment doesn't match with any other , then it is considered unique .
        • How did I arrive at 65% for similarity = analysis of various comments .
        • Note that I don't use any third party API for this , I have written all the codes to check this myself .
  7. Ratio similar - Number of similar comments / Number of Unique comments . The higher the number this is - the less the quality value .

  8. Avg_length_c - Average length of comments by user .

All these 8 factors have various weight assigned to them which I cannot make public because users can game the system "easily" if I do it .

Now all these values are multiplied by "weights" that I have assigned to come up with the "Quality" value which determines your upvote value .

Everyday 100% upvote is divided among all these users based on their "Quality" value .

Just know that these are positives and negatives

Positives

  1. Sumlen

  2. CC

  3. Parent C

  4. Unique words

  5. Unique comments

Negatives

  1. Similar comments
  2. Ratio Similar

Neutral

  1. Average Length Comment .

Positives - You have to work on increasing these scores . The higher they are , the better your quality value is .

Negatives - You have to work on decreasing these scores . The higher they are , the less your quality value will be . It is assigned negative value in the script .

Average Length comment - This can be both positive and negative . I have set a threshold , if it is above that threshold then it is taken as positive value . If it is below that threshold , it is taken as negative value .

Example of myself

If you check my name in the list , I am ranked at 28th . ( It says 27 because it starts with 0 ). I have made 14 comments , talked to 12 authors but still 28th ?

A lot of people have done just 3-4 comments and are above me ? Yes this is because of the similar comments , I have made 3 similar comments :(( Sorry lol + the average length is not great too . That's the reason the script punished me and put me down at 28th rank .

It is the total effort made that is taken into consideration

Now this is the last thing I want to clearly specify . Sometimes you might see a @amr008.pob upvote of say 10 POB on a comment which says " Wow this provided me so much insight , thanks" and you might think , that's not fair at all .

But just know that the upvote is caluclated based on all the other comments the user has made . The script just picks a random comment to upvote .

So in this particular example , the user would have made 10 quality comments elsewhere .

I will keep improving the script to make sure the best quality comments are upvoted more and stays at the top .

That's it from me today :) Feel free to provide your feedback here in the comment section . Let us keep that engaging going .

Regards, MR.

Posted Using LeoFinance Beta