I work on the reddit operation for NEAR, and as some of you know we pay out tokens to people who get their questions answered on our regularly hosted AMAs. I am the team member in charge of sending out the rewards to users. We pay out more NEAR to the “top” comments, a designation we originally adjudicated based on upvotes. When we were using this system, I noticed that the same few accounts were consistently being upvoted above the others in what appeared to be a coordinated fashion. Clearly, some enterprising fellows had banded together to game the system. This was not particularly surprising, and we honestly should have seen that coming. In response, I simply changed the format so that AMA hosts choose the top comments, rather than relying on internet democracy.
This was all well and good, until I decided to take a look at the transaction histories of those I had paid out. I found that many of the users I was paying out were immediately sending their rewards to other accounts.
In many cases, the same accounts that were cheating in the “failed democracy” system were sending NEAR back and forth between one another. I deduced that many of these users are probably the same person, creating multiple accounts to optimize their rewards. It is also possible that a more complex conspiracy is afoot. This requires investigation!
This morning, I began the time-consuming process of mapping out the conspiracy. It was a very straightforward process of following the money. We require that reddit users fill out an airtable form to qualify for rewards. This serves as a database of near wallets attached to each account. By analyzing the transactions immediately following my pay-outs, and cross-referencing that with my list of near wallets and accounts, I was able to piece together a picture of which accounts were in on the racket. I have added each of their usernames to my “shitlist,” and will be discussing next steps with the NEAR core team on Wednesday. I realize that this forum post is likely encouraging the suspected perps to create new and increasingly devious ploys, but such is the price one pays for transparency.
Now, to my point. As I was doing this investigation, I realized that this entire investigative process could totally be a dApp on NEAR. I suspect it would be fairly straightforward to write a code that can identify suspicious flows of NEAR between different accounts, and create some sort of flagging mechanism. That way, community members hosting regular giveaways can identify and disqualify members who show signs of collusion.
I mentioned this to a few NEAR dev friends of mine who have encouraged me to post this idea in the forum. Perhaps one of you coders out there might like to take up this challenge and build what I propose we call Narc Engine: a tool for catching bad actors exploiting online giveaways. Godspeed.
