Hello. My Submission is:
Free TON Subgovernances efficiency analysis
#3 Contest proposal: Subgovernances efficiency analysis
SG Efficiency analysis (EA) in current situation on Dec 2020 must be maximally automated with software tools. EA results must be stored in widely used formats like PDF and Excel to give users possibility to additional analyze by yourself.
Table of content:
Dataset link - https://mega.nz/file/J25SBYiJ#Sz7QmxXKrMnpOmJIv5-JyFoaiB3XMsqt-v0-IeLCfwM
To analyze something we need some data. To collect data we need understand what data we need. To decide what data we need, we need define our goals.
Result of any subgovernance (SG) is an additional users , additional features including collaborations and additional TON Crystal capitalization .
Features can be measured also in users who need this features.
So efficiency of SG is a complex of:
- number of additional new Free TON users ( NU ),
- number of users of a new features ( FU ).
Problem of a FU is a what if some feature used in core of Free TON network. This means all of users need this feature. All users is a current number of Declaration of a Decentralization (DoD) signs (link). In fact some users can work with Free TON without signing DoD, but we need some method to count users. All core features can’t be compared each other, and additional core feature not increases number of users. Core features can be compared in work hours . All other features must be measured and compared in users . Probably in some defined period of a time.
To measure FU, simplest method is a count users in some entry point like a web statistic or a logical bottleneck point of a new feature. For example contract source address in a blockchain when doing payments in TON Crystals. Also problem is an extract active users from all counted users.
Current number of DoD signs – 8406
Current forum users – 2844
Free TON capitalization can be calculated with a middle price of a 1 TON Crystal in USD or other fiat money on market multiplied by emitted (link1 and link2) or planned to emit TON Crystals. Because market specific, speculators, regulators, real infrastructure risks, many events in same time - we can’t compare capitalization. Only way, if possible for exactly feature, is a measured attration or income of a fiat money. In same time we can measure our costs/rewards/outcome in TON Crystals. Fiat money can be converted to crystals and wise versa by current market rate.
Current exchange rate TON_USD ~ 0,8 USD
Current TON Crystal emitting (except subsidy) ~ 105 547 634 TON Crystals
We can calculate CAP = 0,8 * 105
547634 = 84
438107,2 USD, but this is a lie for now, because exchange rate will fall if all users will sell received TONs and will rise if new users will buy TONs to pay for any goods in market. Also current exchange rate not include value of already developed software, cost of current validation and many more. Basically, CAP will rise if Free TON network will be useful for end users .
So I propose concentrate on user counting and classification . In same time I fully against any “Know Your Client ” (KYC) or personalization for people, not for public companies. Public companies/organizations must be transparent. For classification I mean analyzing sources and destinations in payment transactions in blockchain – creating links.
My idea for all TON news users counting and classification will be in Suggestions chapter below. Here we try count SG users, Proposals users. To count SG users we need list of SGs. So Plan is:
Get list of SGs
Get list of resources of each SG (link)
Get list of links/transactions
Count users per proposal/contest/group etc
To analyze SG efficiency we need sources. For Subgovernances we have such sources as:
- SG forum media (text, images, audio, video)
- SG official telegram group media (text, images, audio, video)
- List of SG jurors (telegram names, wallets, real names if available, all nicknames of each juror, chairman date start, chairman date end)
- List of SG proposals
- List SG jurors decisions/vote results (proposal, juror, vote)
- List of contest applicants (telegram names, wallets, real names if available, all nicknames, promise additional users/features, deadlines, payment plan)
- Blockchain FROM and TO address from transactions
Found SG (see method in Appendix A below):
Let’s try to analyze Free TON Wiki Subgovernance
Here (link) can be found its resources:
But problem is in some users in same time registered in all Telegram groups and Sum does not mean anything. Getting user list is not a simple process. There are several variants to do it – manual text selection (see Figure 1 ) with only names in Telegram web, manual click on each user with copying needed data like name and ID, use Telegram API (need write program), use Selenium or other browser automation engine like XHE (need write program). I copied only names from 2 groups because have no time (members in other groups was hidden). Then I use “remove copies” in Excel. Number of uncial user names was 254 .
Also this users is a content makers, not a final Free TON users. We need count final users by web statistic methods. Web statistic methods will give us not only map of auditory, but also understanding what articles is more interesting for people, how people find Free TON resource (reference URLs, search queries/keywords) and many more. For now seems like Wiki and Forum doesn’t use any web statistic code. So in fact we can’t compare this SG because don’t know how many unique users read Wiki and uses Free TON. Also there are some delayed and hidden effect in media distribution. Some people can read Wiki, think, discuss Free TON with friends and after a long time come back and start use Free TON.
Let’s finish try to analyze Free TON Wiki Subgovernance and back to common principles.
To analyze external partner/resource/collaboration we need count users from this partner. As I describe above, to do this we need web statistic . To divide and count traffic from different partners we need different welcome pages in main Free TON web site. Also short URLs redirection pages with counters can be used to public promotions on partner resources. Some short quiz can be added on DoD signing page – users can click on partner logos while answer “How You know about Free TON”.
To analyze partner, his wallet address can be used to check his transactions to new/promised users. For example one of a condition to partners must be to public his wallets addresses which was used to transfer TON Crystals to new users. Another condition is adding especially formed Free TON community web statistic code to his resource where from new users goes to Free TON.
People can better analyze information in graphical view. Blockchain can be visualized like a Graph . This is a non-trivial task. Current number of blocks is about 123
431932 – big volume, a lot of time to query all blockchain transactions, high load, big problem is a correct distribute nodes in space, but result can be very impressive . For example last.fm was visualized here (see Figure 2 ). Wallets will be nodes and transactions will be chords. If we create same visualization for blockchain and we will know wallet of a partner, then we will see influence/benefit of each partner for Free TON network. Process of creation such graph can be replayed with time axis where adding new nodes/wallets and chords/transactions is a time they was added into blockchain. So we will can analyze any moment in past when was activity with any partner. When replaying process of creation such graph, new wallets and transaction can be more lighted for a while. Wallets can be clustered in visualization by partner or by other logic. Graph can be visualized with a part of all blockchain started from observed/partner wallet.
In fact we can’t analyze SG or partner without access to web statistic.
Blockchain/partner can be analyzed offline or online with Graph Visualization tool.
- Add Yandex Metrika or Google Analytics to forum
Yandex Metrika (https://metrika.yandex.ru/)%20) (see Figure 3 ) or Google Analytics (https://analytics.google.com/) is a good decision for getting more statistic on forum.
Free Ton forum based on Discourse engine, so to analyze forum we can use documentation from Discourse (https://www.discourse.org/).
Here (https://meta.discourse.org/t/setup-google-tag-manager-for-analytics/47335) officially described how to add Google Tag Manager to Discourse or
By Alexey Komarov
Some sources must be collected manually and some can be collected automatically. List of all SG available at https://forum.freeton.org/ (see Figure 5 ). SG parts and URLs on common forum can be found manually by next keywords in lower-case manner:
To reduce manual work we can get list of categories in Chrome browser developer mode. Copy category html block (see Figure 6 ).
Use XPath with one of an online XPath parsers (here or https://www.freeformatter.com/xpath-tester.html). To find all subgovernances use XPath (see Figure 7 ). Paste html into XML Input field. Use XPath expression to get list of titles and expression:
and to get its URLs: