At the end of last year, the Free TON community announced a contest for the best solution to improve the electoral system for Latin American countries.
The problem of the electoral system is a common problem for Latin American countries. A striking example of this is the constant tension over the elections in Guatemala.
As noted in the contest announcement:
The tabulation of voting results in Guatemala is conducted offline and has significant drawbacks due to the need for a “paper trail”.
By the decision of the organizers, the prize fund will be divided among the participants, whose works took one of the first 10 places. In total, there were 20 applicants, which means that exactly half of the participants received the reward. The amount of tokens, of course, depends on the position in the rating.
Review of the Top Three
Work № 14 eceived 8.11 points from the jury.
To make it clear how non-trivial and multi-tasking the contest was, let’s get acquainted with the work of the winner.
The author aims to develop a simple, budget-friendly and flexible enough mechanism for counting votes and checking the results, taking into account the «resistance to change» of state institutions. With regard to Guatemala, he assumed cooperation with the Supreme Electoral Tribunal (TSE) and work on the basis of Acta #4 protocol.
As the author pointed out, when developing the mechanism, he relied on the principles of management and operation in Free TON, since in both cases the basic environment is the lack of trust. To overcome the problems associated with this, he suggested using blockchain technologies, as well as monetary incentives for honest participants in the process.
Within the framework of each specific campaign, he recommended issuing a unique token called the “Democracy Token”, which will serve as an incentive and regulator of electoral processes.
The author divided the entire electoral process in the work into two stages — entering data on elections and validating these data.
The Collators register and receive the locked tokens to the account, then enter the data into the blockchain. If the entered data is correct (this must be confirmed by the Validators at the second stage), then the tokens on the Collator account are unlocked.
The Validators check the entered data. In order to minimize cases of Validators dishonesty, it is assumed that the verification is carried out in several stages. First, the Validators will be shown photos of documents uploaded by both Government representatives and Collators from among the observers and volunteers from the electoral precincts. At this stage, the Validators must enter the correct numbers into a blockchain digital Acta #4. The entered data is added to the pool of ballots and will be shown to the Validators later, along with the rest of the ballots.
Another insurance mechanism is the introduction of «artificial» ballots in the first round of the audit.
We will also show them fake computer generated Acta #4 documents, if the Validators validate the fake documents their stake will be slashed in a progressive function, meaning that each mistake will be more costly than the last. The Validators will have no way of knowing which documents are real or fake.
In the mechanics of the Validators’ process, the author suggested using the Byzantine Fault Tolerance algorithm BFT. But at the same time, to verify the document, it is needed to have not two-thirds of the votes, but two-thirds or 66% of the consensus of the locked Validator tokens for each round of voting. That is, the voice of a Validator with an excellent reputation will be more significant than the voice of a Validator whose share of rewards was reduced in the process of incorrect decision-making.
The scheme is demonstrated by an example: if 10 validators vote on a document, then in order to get 66% it is not necessary to have votes from 7 of them. The consensus will be 66% of the total share of tokens of all 10 validators. This means that 5 Validators with an excellent reputation can carry out verification. At the same time, validators will be selected randomly for each new document.
Some judges indicated insufficient description of possible risks. In particular, what if the condition specified in the essay about 66% of honest Validators is violated?! Otherwise, the judges highly appreciated the mechanics of the proposed process, incentives for participants with tokens and options to prevent fraud.
Bravo! Fully decentralized approach to elections audit. Clear and mindful description. The only statement which requires additional research is “66% of Validators are honest”, there could be a case when a political party will fund an attack on an audit.
In general, the work is a promising audit mechanism for further development and takes the first place. The structure of the winner’s essay is also typical for the other best works in the contest.
2nd and 3rd place
Works №6 and №17 scored the same number of points — 7.77.
The distribution of reward in this situation depends on the decision of the organizers. Initially, it was assumed that 27 thousand Crystals would be awarded for the 2nd place, and 24 thousand for the third.
The author started from the real data on Guatemala. Given the number of polling stations, he assumed the number of people involved in the process.
In his scheme, the author considers the fixation of documents when voting on a digital device with further verification using digital signatures, following the example of achieving consensus in a real audit under the Acta #4 protocol.
The judges did not spare any points for clear visualization and a systematic approach to the construction of the concept:
The submission provide a clear solution and fully meet contest requirements. The only step back from original contest task is that this system should be implemented with some interaction with election officials.
The issue of excluding fraud during the calculation and audit was the most important issue in the contest. According to some judges, it was not sufficiently disclosed in the work.
Handing over keys is a bottleneck. In general, the solution is not bad, but no original ideas.
However, due to its clear structure and visualization, the work entered the top three.
The trio of authors applied the so-called «architectural approach» in describing the concept, which includes the following architectures:
The approach is quite similar to the solutions from previous work. It also offers a digital multi-signature scheme for documents. At the same time, Guatemala’s Digital Voting Audit has been introduced into the scheme. The management, recording, and audit of votes are spelled out in detail.
The advantages of the idea include a step-by-step plan for the introduction of a multitasking concept that according to the authors can:
Solve the problem of ensuring independent control regardless of the level and type of voting — whether it is the election of the president of the country or the election of the best teacher in a city school.
The jury members were not original in their comments:
Handing over keys is a bottleneck. Didn’t understand why 50% + 1 signatures. Unfortunately, there are no new key ideas.
Overall impression of the work: it seems that the authors have overdone the details and introductions. Too much of the essay is devoted to describing the problems of the current system and the design of the tasks. In a word, «water» over the edge, which was noted by the jury:
Great detail and much thought put into it but it was difficult to follow. In addition it requires collaboration from the elections authorities.
As you can see, the problem of cooperation with the official election authorities is a bottleneck for all work. This is a really difficult question, the answer to which should change the voting system, making it transparent and at the same time protected from attacks by political groups, financial corporations and other clans.
Politically Correct Forum Reaction
Unlike the design sub-governance contests, which were accompanied by emotional comments with often opposing opinions, this project caused a minimum of comments. Even the jury turned out to be beyond criticism. The judges honestly argued two of their self-recusals by insufficient qualifications in this specific issue. The rest of the jury often left detailed comments in addition to their assessment.
Among the forum members, the contest received warm support and politically motivated comments.
Of course, the contest was also followed from Latin America:
Thank you Free TON community for the opportunity to continue building the #Fiscal_Digital movement to expand our borders. We look forward to sharing electoral lessons from Guatemala so that other countries do not repeat our mistakes. Carbon paper technology was invented in 1801 and yet is still the underlying guaranteeing technology used by elections auditors in most of the world. Blockchain is a better technology, offering increased speed with accurate digital custody of election documents while dramatically reducing the costs of elections, particularly in the developing world. We can rescue TRUST in election results worldwide, but we must be able to VERIFY in real-time. #Fiscal_Digital, we count the votes together!
Time will tell how realistic it is to implement the ideas submitted for the contest. In any case, the democratization of the electoral system is associated with new technologies, no matter how resistant the state systems of the world are to this process. And the very first step, which begins the road of a thousand miles, was made by the members of the Free TON community.