Account Recovery Analysis: How much chances is an account recovered?

Introduction

---

The objective of this analysis is to determine the Steemit account recovery performance by looking into the TxAccountRecovers section of the steem database. At the same time, the chances of an account to be recovered. It will look into the steem block chain database to evaluate steemit account recovery activities based on:

• The numbers of recovery request, actual numbers of recovered and UN-recovered account within January to December 2017; and,
• The number of recovered and UN-recovered account by recovery account to determine which recovery account has a higher success percentage as compared to the other recovery account.

Outline

---

I. Scope of Analysis

II. Data Presentation and Analysis

The data presentation and analysis includes monthly statistics for recovered and UN-recovered; and, monthly statistics by recovery account.

III. Conclusion

IV. Tools and Query

I. Scope of Analysis

---

The analysis performed using the data extracted from @arcange 's SteemSQL database. There are two sets of data is used:

• A Steemit accounts recovery activities for the 12 month time frame ( January to December 2017) which involves the number of account recovery request done, average and sum of recovered and UN-recovered accounts in the period stated earlier.
• A monthly statistics that involve the TxAccountRecovers section of the Steem blockchain database to acquire data specific to the distinct recovery account and recovered status.

The data extracted is taken as the sum and average of the number of recovery request, recovery status, and recovery account for the subjected timestamp. This is done to assess the performance of Steemit account's recovery activities. The data extracted and subjected to analysis is presented using graphs and tables, as applicable. In addition, a data interpretation and summary of result is then presented.

II. Data Presentation and Analysis

---

In 2017, A total of 807 account recovery request done resulting to 450 and 357 recovered and UN-recovered account respectively. The account recovery request has a 56% chance of recovery against
44% odds to be UN-recovered. We can draw that for every 25 recovery request, 14 accounts can be recovered.

On average, there are 67 account request with close to 38 and 30 recovered and UN-recovered accounts per month. A zero recovery request tallied on February 2017 while the largest UN-recovered accounts can be drawn at May 2017. Generally, the account recovery highlighted by a 56% distribution shows an increasing trends in the number of recovered account in 2017. A highest recovered can be observed nominal to four months data--April, July, August and September. It has an above 80% of account recovered against less than 30% UN-recovered, which is a good indicator for an increasing chance for an account to be recovered. In contrast, a decreasing trend in the numbers of UN-recovered account highlighted by less than 30% in four months.

In addition, the number of account recovery request has a increasing trend highlighted by May and October 2017 with between 140 to 160 request. A substantial decrease can be observe in the the data points of November 2017, but overall, the plot suggest a positive increasing trend line for the number of account to recover request.

Majority of the recovery is done by @steem with 568 recovery transactions with an 70%-80% chance for account recovery. A 143 recovery transactions was process by @steemitadmin and the remaining 96 recovery transactions done by other accounts, with 0% and 10-20% chance of account recovery respectively for both recovery account.

A total of 434 recovery request was successful under the recovery account @steem, amounting close to 76.41% chances of success for recovery. A 143 recovery request was unsuccessful under the recovery account @steemitadmin, having 0% chance to recover accounts. Other recovery accounts accounts, such as @anonsteem, sum up to 16 accounts recovered with 16.67% chance of recovery.

III. Conclusion

---

In 2017, A total of 807 account recovery request done resulting to 450 and 357 recovered and UN-recovered account respectively. A total of 434 successful recovery was done by @steem, which is equal to 96.44% of the total recovered accounts. Generally, a 56% chance of account recovery is observed in 2017. Based on the data trends, a projected increase above 56% in the numbers of successful account recovery is foreseen with the a projected increase in the number of request in 2018. A close to 76.41% success of recovery if processed by @steem, which will increase in 2018 based on the data trends in the graph presented. By the way, it is projected that their is no changes in the success percentage for @steemitadmin recovery processing. It will remains to 0%.

IV. Tools and Query

---

I extracted the data from SteemSQL by importing it on Microsoft Excel using SQL Queries. I used Microsoft Excel to visualize and create a graph for the data. The data is extracted by using COUNT() function from TxAccountRecovers. For the monthly transfer data, I run different queries using the different conditions to extract specific set of data in accordance to different recovery account and recovered status. For recovery account you can change the condition from recovery_account='steem' or recovery_account='steemitadmin'. For recovered status, change, from either recovered='FALSE' or recovered='TRUE'.

SteemSql Query
SELECT MONTH(timestamp) AS Month, COUNT(ID) AS Number_of_Request, COUNT(account_to_recover) AS Account_to_Recover FROM TxAccountRecovers WHERE recovery_account='' AND recovered='' AND timestamp >= '2017/01/01' AND timestamp < '2018/01/01' GROUP BY month(timestamp)

In the case for the other recovery account, change the WHERE conditions.

SteemSql Query
SELECT MONTH(timestamp) AS Month, COUNT(ID) AS Number_of_Request, COUNT(account_to_recover) AS Account_to_Recover FROM TxAccountRecovers WHERE recovery_account!='steem' AND recovery_account!='steemitadmin' AND recovered='' AND timestamp >= '2017/01/01' AND timestamp < '2018/01/01' GROUP BY month(timestamp)

---

Posted on Utopian.io - Rewarding Open Source Contributors

---