Experiment #8: Harmony Flash Bounty
A potential explanation for a significant drop-off in active wallets.
(originally written by Donatus Prince)
Our goal with this flash bounty was to analyze Harmony wallet activity to give a potential explanation for the significant drop-off in active wallets noticed from May 26 to May 27, 2022:
Usually, our bounties run through an organized system every week: announcements on Fridays with a set of bounties that will be due a week later.
And in our regular bounty programs, all valid submissions are rewarded.
With the Harmony flash bounty, we gave analysts extra days to complete it (a total of 11 days), and only the top 3 submissions, as determined by a committee review, were rewarded.
The best submission received 300 USDC, and other submitters also earned 100 USDC each.
Submissions by the flash bounty contributors include:
- Transaction Comparer by ant
- Active Wallet Drop by ant
- Harmony Activity Drop by jackguy
- Harmony Lost Users by pinehearst
- Harmony Wallets by issui
- Harmony Mystery 1 by potmo
Contributors could find the evaluation criteria for the flash bounty here.
The primary ask for the flash bounty was to:
- Investigate the sudden decrease in active wallets
- Check if there was a network of wallets that stopped operating
- Check what these wallets were doing before the late May drop-off
- Check if those wallets were now active on other networks (Ethereum, etc.)
Some other potential lines of hypotheses that could be explored included:
- Did a large number of DFK players stop using the Harmony implementation?
- Was there a protocol issue or migration that could have created this drop?
- Did these wallets resemble human or bot activity?
We only promoted the flash bounty in the MetricsDAO Discord instead of through the standard bounty portal, tapping our active community directly to promote a faster turnaround.
The consensus from these analysts is that active wallets comprised new users performing their first and only transaction on Harmony, and 99% of these new users only completed one transaction.
Findings also indicate that majority of Harmony’s new users were bots, and the growth observed was highly conflated for a while.
We are excited about the flash bounty experiment for Harmony because we saw community members discussing their approaches and collaborating instead of competing against each other.
Analysts in our community had spent time already working on the Harmony data, so the flash bounty was more effortless for them to complete quickly.
The bounty created a sense of fulfillment and increased community engagement and participation.
We want to continue experimenting with incentives that promote collaboration instead of competition between analysts. The best part of MetricsDAO is our community of brilliant analytical minds, and bringing everyone together always brings about more robust analyses than if we’re all working in silos.
In this case, the wisdom of the crowd helps to break biases and provide more well-rounded answers to complicated questions like this one.
MetricsDAO is uniting the best analytical minds in the space to build the future of crypto analytics.