ICON Reviews - Voice of the ICONists

Contribution Proposals
#1

ICON Reviews

(This is what we previously submitted as a Grant Proposal)

Project category: Development

Project duration: 6 months

Project description

Why Reviews?

Voter apathy and vote stagnation are inherent problems we face in the governance of ICON network. We have conducted a detailed analysis on this topic (https://medium.com/swlh/investigation-of-vote-stagnancy-on-icon-network-icx-5aab416b8b2e), and found that about 40% of total votes were ‘untouched’ in the last 90 days, while 75% were not moved (stayed on the same P-Rep). Vote stagnation and vote spreading are currently being tackled, particularly by ICONFi that is actively voting and spreading across P-Reps ranking between 23-100 whose nodes are actively running. However, these votes account for about 12 M votes (out of ~350 M), which means that the vast majority of the votes are allocated by individual ICONists.

There has been some shift in votes as new P-Reps joined. This is particularly true for incentive-providing P-Reps. There are also other P-Reps that initiated excitement around what they will be brining to the ICON Ecosystem and gained votes. However, for newly joined ICONists, it is difficult to know what P-Reps do (or even what they are), and tend to vote for Top P-Reps.

To strengthen the argument, we ran regression analyses and found strong correlations between weekly P-Rep ranking and voter ranking (prior to the introduction of ICONFi). Note that the correlation was stronger with first-time voters (right).

It has been rather onerous for voters to know what each P-Rep has been doing. This may be one of the major factors in voter apathy, because it takes too much effort and time to go through individual P-Reps, and may decide to just vote for the Top P-Reps. This is especially true for the newcomers.

Below are some of the voices from ICONists. The list goes on and on.

366×533 16.6 KB

To increase awareness and to see the reaction of the community members, Tono from our team has done a pilot test as a community member called “Shill Your P-Reps” (https://twitter.com/tonoplast/status/1300978803175432193 ). It has proven that community driven ‘review’ would be invaluable and is a truly decentralised way. However, such a twitter thread has limitations in that it is a point-in-time snapshot of what the community thinks, and unless this is done longitudinally and compiled each time, this is not sustainable. It is also not scalable.

Having a ‘ICON Reviews’ website where ICONists can leave reviews of ICON-related topics (e.g. P-Reps, dApps, NFTs), with ratings (both for P-Reps and the review itself) and other perks will make the decisions of the voters easier, similar to looking at Google Reviews for future favourite restaurants.

How will it work?

This will be a community driven effort, and no central authority will need to flag which P-Rep is good or bad, and therefore, bias-resistant. It will also be updated by the community, so ICONists will know how outdated the reviews are. The reviews will be recorded onto the blockchain, so it will be immutable and even if the website goes down, the information will be available on the blockchain.

Also, reviewers will be required to login with their wallet and send transaction fees for leaving reviews, the address will be publicly available and be somewhat traceable to prevent defamation (e.g. a P-Rep member writing bad reviews for another P-Rep team).

Tipping system will allow for the incentives for reviewers, either by the community or P-Reps, and ranking of the reviews would be determined by the ‘like/upvote’ or the ‘dislike/downvote’ of the reviews. And best of all, although small at this stage, we will be creating transactions on ICON network. Depending on the use case of the website, it may be extended for a creation of an IRC-2 token with a token economics.

Future development will also include more features such as data analytics and visualisations of the reviews/reviewers/reviewees, social media API integration and P-Rep reward tracking for automated transparency report.

As we get a decent amount of reviews, we hope that the ratings (IconistScores) could be implemented in voting wallets, such as ICONex or MyIconWallet (approval dependent). This would make an easy gateway for voters to choose based on the voice of the ICONists.

Example (may differ from end product):

1419×1593 337 KB

  • This is an illustration and the ratings may not reflect the real scenario.

A simple design on the front page with Team logos, names and addresses. No ranking will be shown and each visit may randomly shuffle the order of the P-Reps. With reviews, it can be sorted by the number of reviews, total score (IconistScore) and/or date for recency.

697×1110 110 KB

  • This is an illustration and the ratings may not reflect the real scenario.

Detailed reviews can be found for each P-Rep, with the option to upvote/downvote the review and tip the reviewer. There will be dashboards available for each reviewer and P-Reps for updating their status and tracking the progress of their review/tips.

Final Comments

As the new voters tend to contribute to temporary decrease in the vote stagnancy, it is important to steer them into the right direction. It is critical to educate active ICX holders now more than ever, as their votes will likely carry more weights on which P-Rep gets more funding that will benefit the ICON ecosystem.

We believe that ‘ICON Reviews’ will provide an easy overview of what the community thinks of P-Reps, and ICONists would have easy access to the data which will help new and old voters to make more informed decisions which would be much less taxing than before.

With a grant, we will be able to build this platform and make meaningful adjustments to our existing workload so we can commit more time to it.

Project milestones

  • Prototype development and testing
  • UI/UX design
  • Back-end / Front-end
  • MVP
  • Finalised back-end / Front-end
  • DevOps
  • Community Management & Beta Testing
  • Marketing

Funding amount requested

Total project budget - $28,000

  • Prototype development and testing: $5,000
  • UI/UX design: $7,000
  • Back-end / Front-end: $12,000
  • Community Management & Beta Testing: $2,000
  • Marketing: $2,000

Official team name

Transcranial Solutions

Team Members:

Tono is a data scientist with a neuroscience background . He has conducted a number of analyses on the ICON blockchain data and translations for the community.

LinkedIn - https://www.linkedin.com/in/sung-wook-chung/

Ted is one of the main contributors within the team. He is also the creator of Icon Transaction Extractor (ITX) and has technical experience in blockchain data extraction and development.

Steven, the latest member of the team, is a software engineer specialising in python and C++ development at Soul Machines. He also has game development experience.

LinkedIn - https://www.linkedin.com/in/sjson/

Contact information

Email: transcranial.solutions@gmail.com

Telegram: @tonoplast / @skyvell

Twitter: @transcranialsol

Public address

hx2f3fb9a9ff98df2145936d2bfcaa3837a289496b

#2

First of all I want to say that I really like this idea and think it is needed as the current system just doesn’t work.

Forgive me if I’m reading this wrong but are you only relying on the broader community with no input from the P Reps themselves? The issue that I see with this is that a lot of these great projects have multiple P Reps working on them and unless you are really looking it is often difficult to know who is working on what which may make it difficult for the people doing the reviews to have all the correct information. An example of this is that Parrot 9, Block 42 and ICX Station are all working on Bridge but I may not know that if I haven’t looked into Bridge and review one of these less as I as I don’t actually know what they are involved with.

I think it would also be good if there was some way that P Reps or sub-rep teams can post their accomplishments or projects that they have worked on also. This could give them a single place where they can advertise their worth rather than having to dig though Twitter and then people can verify and review their work.

1 Like
#3

Thank you for the feedback. Anyone should be able to contribute to the reviews, and in detailed view (at least is what we envision), there will be a open dashboard that has more information about what P-Reps are working on – this could be provided by the P-Reps and I think they will be incentivised to do so as it can lead to better rating.

We are also thinking about getting a pool of trusted ICONist to do a paid full-review of P-Reps (or even P-Reps themselves promoting). This will likely come after the launch of the product (or at beta phase).

#4

Cheers for the reply, I think having the PReps promoting themselves and having people verify what they say would be the go.

1 Like
#5

Interesting concept here. Curious, how are you going to protect this from abuse? Let’s say…someone creates multiple wallets with a couple of ICX in each wallet. Then that malicious actor fakes review, may it be bad or good.

2 Likes
#6

Hi Ian,

Thanks for the comment. I think it is not going to be easy trying to police it. We will need to find patterns to see if someone is using multiple wallets for spam comments, or potentially more straight forward way is to connect the dots between wallets. I don’t know of a perfect solutions for this and we are internally discussing the very issue.

At some point, abusive wallets could be blacklisted, but that doesn’t stop from creating others. In terms of reviews and ratings, it could be evaluated using the whitelisted addresses (all data will be available on the blockchain but frontend would show whitelisted ones).

I think the whole point is really to have no censorship, and hopefully with more good reviews, those abusive ones will be negligible. I think this is why we could start off with selected pool of good ICONists give review and see how it could work, but that can also be quite biased. Having specific number of ‘upvotes’ might help weigh the ratings.

1 Like
#7

Two suggestions to prevent manipulation of review comments, etc.(Must be KYC certified to give p-rep score)
First, it is divided into KYC authenticated comments and unmarked comments.
To get rewarded for comments, they stake a certain amount of icx in the app (ex 10icx)
This staking vote votes for developers.

#8

KYC seems a little much, and I don’t think people will be inclined to leave reviews. It should be as easy and seamless as possible because it’s quite likely that it will just be an empty website with barriers like that.

Staking to get rewarded hence encouraging ICONists to leave honest reviews might be something that can be worked on, but may not necessarily stop trolls from leaving comments.

I don’t know if we are being overly concerned about these potential abuse. Just thinking out loud, how is it any different from, say a manager in Restaurant A leaving bad comments about Restaurant B because it’s a competitor on google reviews. Or customers leaving bad reviews with multiple account.

#9

I think this is a great idea and absolutely needed for the community and new joiners. As someone who follows news about Icon probably more than the average holder, I also struggle to understand which P-Reps to vote for, and which ones to avoid.

I also suggest some kind of mechanism where P-Reps can provide regular updates (I would suggest monthly) on:

  • What they have achieved (in last time period, and since beginning)?
  • What are they working on now?
  • What is their roadmap?

I think this self reporting on progress might be more important as a mechanism than a review system.

If they cant clearly articulate in this way the value they are bringing, and why they should be receiving rewards, then they do not deserve votes.

Agree there might be some malicious activity to tank/boost some P-Rep scores on a platform like this, so some preventative mechanisms will need to be built in. I wonder if there is a way for trusted parties to get higher weighting in scoring or at least provide independent assessments of P-Rep activity, although this goes against principles of decentralisation.

2 Likes
#10

I actually like the idea of KYC for the reviews. You could have both a Verified Iconist Review and a General Review section similar to how Rotten Tomato’s has a Critic and Audience score. That way people can choose to KYC or not but it will help identify if spamming reviews is happening with lots of random wallet addresses. I could just be over complicating things though.

3 Likes
#12

Yes, you are trying to good work. Voter apathy and vote stagnation in the governance of ICON network is a big issue. P-Reps in the lower ranking are suffering and also they do not have the financial power/muscles to do something big to be visible. Good to know that ICONFi is actively voting and spreading across P-Reps ranking between 23-100 whose nodes are actively running. I need to spend more time on this in order to understand how it is going to help more for P-reps in lower rankings?

#13

I think KYC with verified IDs for the reviews maybe help?

#14

Yes, I think KYC with verified IDs for the reviews maybe help!

#15

Interesting idea

If we want it to be a trustful and useful source, I believe it should include mechanisms to ensure that all reviews are trustful and the scores not manipulated.

I understand the concerns around spam, it can really kill a review Dapp. Saw some suggestions about KYC. Personally, it would personally make me feel quite uncomfortable to go through a whole KYC to “just” leave a review.

To keep it permissionless and avoid spam, would it make sense to only display reviews and the average score of users with min 100 ICX staked? Then using a filter button one can see a larger set of reviews and non-adjusted scores.

I guess you’ve noticed this, MyIconWallet by ReliantNode has a nice Discover P-Rep overview section. Does the ICON ecosystem need a standalone review Dapp? Would it be possible for users to directly review their P-Reps inside existing wallets? Have you talked with these already? Your project would benefit from an existing user base, and they would have a better app with more data in it?

1 Like
#16

This is not a bad idea. We’ve discussed internally the possibility of having a pool of verified and trusted community memers leave reviews. And have them share a pool of icx for this effort. This would get us some quality data by individuals who knows the ecosystem well. The issue was that we want everyone to be able to leave reviews. So this might be a good option - having serperate review pools for “verified” and “unverified”.

2 Likes
#17

Thank you for all the feedback. Much appareciated.

We agree with this 100%. The data should be high quality - it should accurately represent the state of the P-rep ecossytem. We are currently discussing how to make this happend, and appreciate all improvement suggestions we can get.

I think this will have some effect on spam. But if you really want to manipulate scores, it can still be done with wallets >100 icx. This could be somehow mitigated by only allowing one particular address with 100 icx to review one particular P-rep only once. Although this also has it’s own drawbacks. Appreciate the suggestion.

We also have thoughts in this direction. It would be really valueable to have quality review data integrated in e.g. myiconwallet and eventually be able to leave reviews in the wallet as well. This is something we have not yet discussed with them (although we are in contact with them and discusssing potential collaboration here).
The main reason we are also thinking about a standalone webapp is because we have a larger vision for the iconreviews project. In addition to reviews, we want to have more data categories quantifying P-rep contribution. Categories such as social media presence, number of transactions generated by a P-rep. Any areas which can be quantified and represent contribution I guess. Tono is also working on a way to track where P-rep funds end up using graph theory. This is all part of a larger vision.

1 Like
#18

In most cases, people don’t want to do research and vote for Top P-Reps. It might still be the case although now we see more votes flow to incentive giving P-Reps, but also some of these P-Reps are really contributing. Still, there are many ICONists who care very much and vote fot those good P-Reps. Many do not have time to go through all the list and dig through information, so this would be a decentralised centralised place for it. We are planning to make it as easy as possible for ICONists so that they are well-informed, which indirectly help shift votes from bad actors / not contributing P-Reps to those who are good and contributing to the ICON network.

#19

I guess you’ve noticed this, MyIconWallet by ReliantNode has a nice Discover P-Rep overview section. Does the ICON ecosystem need a standalone review Dapp? Would it be possible for users to directly review their P-Reps inside existing wallets? Have you talked with these already? Your project would benefit from an existing user base, and they would have a better app with more data in it?

We have not quite ironed out the details, and we did get in touch with them to see whether we could expand on it. The reviews are not mean just for P-Reps though, it’s for all ICON related things but it will start from P-Reps and move on from there. The backend work is also substantially different (as far as I can tell), so we are internally discussing this issue.

We would of course like to work with ReliantNode on this, and we will think of a way to integrate it if feasible.

1 Like
#20

We came up with a possible solution to some of the problems raised – mainly for review spams or others alike.

  1. We will block wallets that are new and have no history of voting in the past.
  2. We also make reviewers ‘stake’ to the platform when they leave review, and it will be up on the website for the amount of ICX staked – i.e. 1 ICX = 1 day, 100 ICX = 100 days (+ unstaking days). These can also act as the weight of the review.
  3. If it is a positive review, then the ICX will be delegated to the P-Rep (or give the option). Otherwise, and if it is a neutral / negative review, we will delegate the ICX to our node.

We could sort by the ‘power’ of the review (ICX staked) and the number of upvotes it gets. When we get to create $ICP token (ICONist Contribution Point) down the road, we can be airdripped to the platform-stakers (this is a separate thing that we would like to do to help spread votes). Potentially in the future, we could have better prize for those honest wallets (old with a lot of voting history).

I think this way, spam will go away quickly unless they are loaded + they have to have staking history (not decided how long but I imagine at least a few months, someone considered ‘active’). Plus, it also means that any P-Rep that did some bad things in the past (hence bad reviews) would have the opportunity to change.

There could be a punishment mechanism where the staked ICX won’t get released and wallet silenced for a certain duration (burning seems a bit harsh) if the review is considered abusive, but hopefully it won’t come to that.

#21

I’m not too technical so apologies if this dosent make sense but couldn’t the review website be built with some sort of read only API so wallets etc like MyIconWallet but not just limited to them could plug in and see the reviews in an interface in their wallet but if someone wanted to add a review it would link them to the website to update.

1 Like