Repository

https://github.com/steemit/steem

This is an analysis of witness voting patterns of the accounts on the Steem Blockchain. In focus are the number of MVESTS held and the status of these accounts not casting a single vote for a witness.

---

Contents

• Introduction / Background
• General / Assumptions
• Not voting for a witness - account type
• Not voting for a witness - alternate approach
• Summary
• Queries and Tools used

---

Introduction / Background

There have been a number of discussions regarding the engagement of Steem accounts with regards to voting for witnesses. And reports are produced each month showing the overall activity relating to casting witness votes.

To summarize, there is a distinct lack of accounts actively voting for a witness, and this analysis is an attempt to find out where the 'inactive' VESTS are held.

VESTS are vested shares in the STEEM blockchain, the more an account owns, the more influence on witness voting an account has.

The aim of this analysis is to try and find the answers to the following questions.

Are the smaller accounts not voting for witnesses?

Or is there another class of accounts not voting for a witness?

And if it is the smaller accounts not voting, how much of an impact is this having in terms of VESTS?

---

General / Assumptions

• The accounts - 'steemit', 'misterdelegation', 'steem', and 'ned' have been excluded from this analysis.

• The Steem token / VEST ratio used is 491.485 STEEM per MVEST

• An account not voting for a witness is one that is not voting directly, or via a proxy.

• For simplicity, the following criteria has been used:

Steem Power	Status
< 5000	Minnow or smaller
5000 - 50000	Dolphin
50000 - 500,000	Orca
500,000+	Whale

---

Not voting for a witness - account type

To try and understand where the 'inactive' VESTS are held, we can group the VESTS not casting a witness vote into account status.

At present, approximately 95% of all accounts are classed as 'minnow or smaller'. Is this where the VESTS are held?

Status	No. of accounts	Total MVESTS
Minnow or smaller	953,657	21,930
Dolphin	686	18,411
Orca	86	21,656
Whale	12	27,178

Over 950,000 accounts that are classed as a 'minnow or smaller' are not casting a single vote for a witness. At first glance, this looks like the root of the problem. However, if we look at the number of MVESTS owned by these accounts, the total is only 274 MVESTS more than those owned by 86 Orcas - just over 1%.

Further to this, just 12 whales not voting for a witness control 27,178 MVESTS , which is 20% more MVESTS than all the 'minnow or smaller' accounts combined.

The chart above emphasizes the point further. Of the 954441 accounts not casting any vote for a witness, just 784 of these accounts control 75% of the MVESTS.

To put these numbers into perspective, and using https://steemian.info/witnesses as a guide.

• If every account classed in this analysis as 'minnow or smaller' (all 953,657 of them) voted for a new witness, this would be enough MVESTS (21,930) to place the witness at 43 in the active rank.

• If the 86 accounts classed as 'Orca' chose to do the same, the witness would also rank at 43.

---

Not voting for a witness - alternate approach

Again, 'steemit', 'misterdelegation', 'steem', and 'ned' have been removed, and the following data is of the accounts not voting for witness directly, or via proxy.

1000 of the top 2400 accounts ordered by VESTS are not voting for a witness directly, or via a proxy.

The 2400th largest account holds 7,257,524 VESTS - 7 MVESTS - 3566.420 Steem Power.

The VESTS held by the top 1000 accounts including this one, which are not voting for a witness total 69,105 MVESTS

There are 953,402 accounts below this account not voting for a witness.

The sum total of VESTS of these 953,402 accounts (with less than 7,257,524 VESTS / 3566.420 Steem Power) that are not voting for a witness is 20,071 MVESTS.

So, over three quarters of the VESTS not voting any witness are owned by 1000 accounts which are a part of the top 2400 accounts (ordered by VESTS) on the Steem Blockchain.

It is true that we have a lot of smaller accounts not voting for a witness directly of via proxy, but 77% of the Vests are held by accounts approaching Dolphin status or higher.

---

Summary

It is fair to assume that larger accounts would have a vested interest and be aware of the importance of a witness. However, a large % of the Vesting Shares not voting for any witnesses at all are owned by these accounts. Why?

Accounts classed as a 'minnow or smaller' make up the vast majority of accounts not voting for any witnesses, but with much of these accounts inactive / bots / tests, and around three quarters of the MVESTS not voting at all being held by 1000 of the accounts in the top 2400, it seems unfair to point the finger solely at this smaller accounts.

77% of the Vests not voting for a witness are held by accounts approaching Dolphin status or higher.

The larger accounts not voting for a witness are perhaps not voting for other reasons than lack of knowledge and apathy, maybe there is a wish to hide identity - voting patterns, and especially a proxy vote, would possibly give a clues to the account being an alternate account?

An analysis into voting patterns of accounts without a witness vote could be something to think about - this analyst is not venturing that far... today.

---

Queries and Tools used

The charts used to present the data were produced using MS Excel.

The data is sourced from SteemSQL - A publicly available SQL database with all the blockchain data held within.

The SQL queries to extra to the data have been produced in both SQL Server Personal Edition and LINQPAD 5. The code used to produce these results is as follows:

-- Top 1000 excluding 'Staff'
select name, cast(left(vesting_shares,charindex(' ', vesting_shares)-1) as float)/1000000 as vests,cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 as sp, proxy, (LEN(witness_votes) - LEN(REPLACE(witness_votes, ',', '')))+1, witness_votes from accounts with (nolock) 
where name in (select top 1000 name from accounts with (nolock) where (proxy = '' and len(witness_votes)<4) and name not in ('steemit','misterdelegation','steem','ned')
  --order by cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 desc
  )
  order by cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 desc 

-- Minnows or smaller
select name, cast(left(vesting_shares,charindex(' ', vesting_shares)-1) as float)/1000000 as vests, cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 as sp, proxy, (LEN(witness_votes) - LEN(REPLACE(witness_votes, ',', '')))+1, witness_votes from accounts with (nolock) 
where name in (select top 1004 name from accounts with (nolock) where (proxy = '' and len(witness_votes)<4) and cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 < 5000
)
order by cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.41/1000000 desc 

-- Whales 
select name, cast(left(vesting_shares,charindex(' ', vesting_shares)-1) as float)/1000000 as vests, cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 as sp, proxy, (LEN(witness_votes) - LEN(REPLACE(witness_votes, ',', '')))+1, witness_votes from accounts with (nolock) 
where name in (select name from accounts with (nolock) where (proxy = '' and len(witness_votes)<4) and (cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 > 499999.999)
)
order by cast(left(vesting_shares, len(vesting_shares)-6) as float)*491.485/1000000 desc 

_{This data was compiled on the 18th May 2018 at 9:30pm (UCT)}

---

Thanks

Asher @abh12345 / Witness - @steemcommunity