HPUD vs other days - Data analysis .

Good evening to everyone . Today I will be talking about Hive Power Up Day - famously known as HPUD - and let's see how it differs ( if it differs ) from the rest of the days .

HPUD

You all know about HPUD - Hive Power UP Day . It is 1st day of every month , we are encouraged to power up our Hive on this day and we get badges from HiveBuzz for powering up atleast 10 Hive .

Now I am not going to talk about the June 1st 2021 HPUD specifically because as we all know it will be done by other users who have been taking care of it for months together now .

My post won't add any extra value if I did the same so instead of that I would like to share the data for 3 months and let us see how much the HPUD is impacting the staking of Hive .

3 months data

For the purpose of the analysis I have taken data from 28th February till June 2nd ( Yesterday ) 2021.

Let us not waste anymore time and see the data -

Daily power up data

Daily_powerup.gif

If you check this you can actually see these things

  1. There are couple of medium height bars
  2. Many small height bars
  3. One long bar

There is only 1 huge bar - 26th April 2021 . On this day we actually saw a huge staking of 800k + amount of Hive .

I actually went and checked what happened on that day and this is what I found out -

image.png
image.png
Source- Tx Id

I don't have much info about this particular transaction . I will just leave it at this for now .

But we consider this as an "outlier" which means as you all know - it doesn't fit among other observations.

So let us remove that particular day staking details and use cluster analysis to see what pattern we can see.

Daily_powerup_cluster.gif

Note: this is the exact same data as before with just 2 changes -

  1. Removing the outlier ( which is 26th April )
  2. Performing cluster analysis .

Now you can see 3 different clusters here -

  1. Red - meaning fairly huge staking .
  2. Yellow/ Orange - medium amount of staking .
  3. Blue - low amount of staking .

Now let us ask outselves these questions -

  1. Does all the red bars belong to 1st of every month ? - Answer is NO .
  2. Does all 1st day of every month has red bars ? - Again answer is NO .

Okay so that means HPUD doesn't have a great impact of amount of staking ? Well we shall calculate that later on .

Number of unique users staking on HPUD

unique users powering up.gif

I have already applied the cluster analysis here .

Now this is where you can see the real difference

  1. Are all red bars on 1st day of every month ?
  2. Are all 1st day of month fall under red bars ?

Answer for both is YES .

We can clearly see it right ? That means - there are more number of people staking on HPUD than other days .

Merging all days and checking the unique number of users

In the above chart I have taken each day separately for 4 months .

Here I have combined it which means

March 1 and April 1 and May 1 and June 1 = 1

March 2 and April 2 and May 2 and June 2 = 2

Merging_days.gif

Once again I have applied cluster analysis and the result is clear , 1st of every month clearly stands out from rest of the days when it comes to number of unique users .

Let us see the same for amount of hive staked ?

Merging all days and checking amount of hive powered up

Merging_days_sum.gif

See this is why it is important to aggregate and check the results . In the first chart we saw that the cluster analysis didn't show clear distinction between HPUD and other days but here we can see that

  1. 1st and 26th ( this is purely because of the outlier which actually shouldn't be considered ) falls into Cluster 1 .
  2. Everything else falls into cluster 2 .

This tells us that in long term perspective HPUD is having an impact on staking of Hive .

Average - Power up

Average_powerup.gif

Well it tells that 1st day isn't that big of a difference when compared to other days when it comes to average power up .

But a couple of analysis has taught me that average isn't always the best method when analyzing . Since there is a lot of variation in the amount of Hive power up , the average doesn't provide a good insight .

Median - Power up

Instead , let's see Median .

Median_powerup.gif

I don't think this even required any explanation .

This tells us that users wait for the HPUD and then stake . The median value for each and every other day is 0.003 HIVE .


That's it from me today . Let me know in the comments if you enjoy these type of posts . I will continue posting them for you if you like it.

Regards,
MR.

H2
H3
H4
3 columns
2 columns
1 column
12 Comments
Ecency