I know most of the potential audience for this are likely busy playing Overworld, but here’s my somewhat belated recap of the November 2022 Illuvium council election.

Several major storylines dominated this election, and we’ll cover each of them. 

• Implementation of Staking ILV to Vote
• Impact of Late Voting Restart
• First Blind Voting

Your New Council

Overall, the new Illuvium council consists of two incumbents and three newcomers to the council, but all well known names within the Illuvium community. Illuvium founder Kieran Warwick retained his council for the fourth consecutive election (since December 2021), and is now four-for-four in being elected as a candidate. This was Kieran’s second time with the highest total number of votes. Yours truly, Deraji, was again selected to serve the DAO for the second consecutive term.

For the new faces on council, lead Illuvium moderator Blickter won a seat with the second most votes, earning a seat for the first time after two unsuccessful runs (June 2022, finished 7th, September 2021). In his first ever attempt at a council seat, Animositas (Ani Knows) won a council seat with the second most wallets voting for him. Lastly, Scoriox earned a council seat in his second nomination (June 2022, finished 8th).

New Voting Requirements & Restart

Several things challenged the election this time. The first and likely most significant is the overall down market and its impact on engagement. Secondly, this was the first election after IIP-26, which now requires ILV to be staked in order to have voting power. Lastly, as an unfortunate consequence of IIP-26 passing just a week before the election, there was a very late discovered error in the code that caused certain ILV to have too much voting power, requiring a restart about 60 hours into the voting. The combination of these factors resulted in the lowest number of voting wallets since May 2021.

The exact number of wallets impacted by the restart is unknown, as I wasn’t able to download the data prior to the vote being stopped initially, but from my direct observations, there was a reduced turnout in the second election with voting also occurring more slowly. The pace of voting was very good initially during the initial voting period, but considerably slower in the restart. Anecdotally, I know at least one wallet didn’t vote in the restart, as I was personally traveling during the second vote and didn’t have my hardware wallet with me.

Despite the restart and poor market conditions, voter turnout by wallet was only slightly worse than the March 2022 election. With the exception of the May 2022 election, the number of participating wallets has largely mirrored the $ILV token price trend. As we advance forward as a more developed DAO, we will see if this trend continues.

Impact of Staked Voting

Specifically addressing the impact of staking, prior to the passing of IIP-26, I dug into the data to determine the number of past voting wallets impacted by the change to require staking. If you haven’t been following Illuvium Discord governance channels, this proposal was created and quickly passed to help counteract a potential Sybil attack. Due to the low token price and gas price, it was a greater possibility that an individual could acquire ILV and hold it across a number of wallets to impact voting power and counteract the safeguards of quadratic voting.

Assessing the June 2022 election, there were only 7 wallets out of 760 wallets that participated that were still holding ILV but were not staked. These wallets held fewer than 20 total ILV, and voted for multiple candidates in June. With over 99% of past participating wallets still staked, this proposal was deemed to have a minor disengagement impact, while offering added protection from a nefarious actor.

A significant increase in the number of wallets holding ILV was noticed in the weeks prior to the election, though a variety of hypotheses can explain this sudden increase. The non-nefarious consensus idea is that these tokens were held on centralized exchanges, such as Binance, and with the fall of FTX and other centralized entities in November, this was a safeguard by individuals to move these tokens to non-custodial wallets.

Regarding the potential for Sybil attacks, “buying” the council was as cheap as it has been since May 2021. We will consider a simple assessment of what it takes to “buy” the council, being slightly more vILV than the number received by the overall election winner. In this case, Kieran received 2729.03 vILV, which could be received by staking 1 ILV in 2730 unique wallets, and then repeating this five times to win all five council seats. With a token price around $42 on the final day of voting, the entire council could have been “bought” for around $575,000 USD. We will ignore the complexity of managing 13,650 unique wallets. Staking increased that complexity and cost.

However, blind voting actually helped eliminate the Sybil attack potential further, raising the number of votes required to win the election. In the past without blind voting, the full community knew exactly how many votes were required to gain a seat, and vote shifting was common-place, with votes moving from candidates who no longer “had a chance” to win to those in the top 6 or 7, and votes moved from those leading to those in fifth/sixth place to swing the final seat. With blind voting (which we’ll cover more below), it was unknown how many votes the leading candidate had, or even who was the leading candidate. Kieran won this election with 2729 vILV, which was even greater than the previous vILV high of 2460 vILV in December 2021, an election with far greater participation. This is another slightly more subtle security benefit of blind voting.

Wallet Size Impact of Staked Voting

Something I’ve been examining every election is the voting power by wallet size. In the past, my perspective has focused more on decentralization and the impact of Whales (>1000 ILV) on determining the council. With the transition to staked voting and concern over Sybil attacks, the perspective shifted slightly to evaluate the number and impact of Shrimp (1-25 ILV) and Plankton (<1 ILV) wallets. Examining past elections, if we saw a significant shift in balance between whales and shrimp, that may imply either a Sybil attack (significant rise of Shrimp impact), or disenfranchisement or centralization (significant rise of Whale impact).

First, let’s look at the overall vote distribution by wallet size this election. Just under 43% of the total votes cast came from wallets holding over 7500 ILV (Leviathans), while 4.7% of votes came from Shrimp (1-25 ILV).

Comparing this to previous elections, both of these values are in line with expectations. Leviathans have averaged about 42% of the vote in the previous three elections, with a standard deviation of 6%. This election was right on the average for Leviathan, even considering the decreased wallet turnout. The percentage of votes coming from wallets greater than 1000 ILV was also nearly right on the average, with 58.4% of the vote coming from such wallets, compared to a longer term average of 60.8%.

Looking at smaller wallets, Shrimp representation was actually down slightly at 4.7% of vILV, below its 7.5% average, but within a standard deviation (4.1%). December 2021 was the large outlier for Shrimp engagement, with 13.6% of that election coming from Shrimp wallets.

In summary, there did not seem to be any evidence of a Sybil attack, or significant evidence of voter disenfranchisement due to the new requirement of staking.

Considering overall engagement, below is a graph of wallet size for all participating wallets.

On Quadratic Voting Impact

Very briefly, we’ll look at the impact of quadratic voting in this election. As a quick reminder, voting power or vILV is determined by the square root of the number of staked ILV held in a wallet. For example, a wallet holding 1 ILV has 1 vILV, while a wallet holding 10,000 ILV casts 100 vILV. The largest wallet participating this time cast 715 vILV, representing 511,900 ILV, while the smallest wallet held 0.051 ILV, casting 0.23 vILV.

In total, 341 wallets held less than 500 ILV, which represents 80.0% of the wallets participating.  These individuals held a total of about 48,533 staked ILV.

For wallets over 1000 ILV, 45 such individual wallets voted, representing 10.6% of those participating. These wallets cast 4761.2 vILV or 58.4% of the total votes, while holding 1,276,285 ILV, or 96.3% of the ILV participating in this election. These values are in line with what we’ve seen in previous elections, and the engagement of Whale wallets (1000-7500 ILV) was actually down slightly as a percentage, with 15.7% of the vote coming from such wallets, below the average of 19.5% in the previous three election.

In short, quadratic voting is still working to balance the voting power across the DAO.

Blind Voting Impact

The biggest impact of blind voting was the discrepancy in number of votes between first and fifth place. This is essentially the “wasted vote” percentage, as both first and fifth place earn a spot on the council.  Obviously, this election had the greatest percentage of “wasted” votes, with Kieran taking 33.5% of the total vILV, while Scoriox earned his seat with only 6.7% of the vote, for a “wasted” percentage of 26.6%. Previously, only once had any candidate exceeded 20% of the vote (Kieran in March 2022 with 22.4% of vILV), and this was the other greatest difference between first and fifth, with 8.0% “wasted.” In all other elections, the difference between first and fifth has averaged 2.0% of the vote.

Interestingly, the difference between the final council seat and sixth place stayed relatively constant as a percentage despite blind voting, with Scoriox taking 6.7% of the vILV and Scrubadubdad at 4.4%, for a difference of 2.27%. This was one of the higher differences between fifth and sixth in history, but well within standard deviation. Considering lower voter turnout and total vILV, the gap between fifth and sixth was just about average for past elections as well, with only 185 vILV separating earning a spot and not earning a position. In the hotly contested June 2022 election, Kieran just barely edged out Sascha by 137 vILV, so the gap between Scoriox and Scrubadubdad was higher than the previous 5th/6th difference.

In my opinion, blind voting was extremely powerful in the November 2022 election, particularly considering the low turn out. In the past, it was quite difficult for a single wallet to earn a position, both from a combination of the vote being visible, while also a greater turnout dictating a higher number of votes for the fifth and final council seat. Scoriox’s 546 vILV was the lowest for any winning council member in the DAO’s history, with the previous low of 801 coming in May 2021, when only 183 wallets participated. Given the low value for fifth place, two wallets (715.5 and 597.7 vILV) could have single-handedly selected a council member this time. However, due to blind voting, both wallets selected Kieran, largely contributing to his 33.27% of the total vILV. Interestingly, if both wallets would have switched their vote, Kieran would still have earned a seat, and actually would have still received the most votes with 1415.91 vILV.

Preferences By Wallet Size

As I’ve done in past elections, I also looked at each relative voting block to see if there was a significant difference in candidate preference by wallet size. Overall, Kieran was dominant across all voting size groups, winning with wallets above 1000 ILV, and all groups between 1-100 ILV. At worst, Kieran finished fourth in a voting block, assuring himself a council seat.

Blickter earned his seat with a very strong performance among wallets above 1000 ILV, and not finishing outside of the top 5 in candidates with >100 ILV. Interestingly, he was outside of the top 7 for all wallets between 1-100 ILV.

I (Deraji) was fairly consistent across groups with one exception, finishing in the top 4 for all voting block with the exception of Whales (1000-7500 ILV), where I received none of the votes from the 28 whales participating. I did win with my Fish (100-500 ILV). I guess I need to work on my Whale appeal.

Animositas earned his spot also with consistency, and aside from Kieran, Ani was the only candidate to place in the top 5 in every demographic, winning with Big Fish (500-1000 ILV).

The largest wallet voting for Scoriox held only 3546 ILV, and was the smallest “big” wallet of any of the new council members. Scoriox earned his spot with a strong showing in wallets between 50-1000 ILV, and finishing just outside of the top 5 with Whales.

Of the candidates that didn’t earn a spot, Scrubadubdad had the strongest showing, finishing as high as third with Whales, while also finishing in the top 5 with Lobsters & Shrimp (1-50 ILV). Artemis, Lucius and Vetemor were the only other candidates to finish in the top 5 in any demographic.

Taking a more zoomed out approach to the data, Kieran obviously topped both wallets above and below 500 ILV, with Blickter finishing second with >500 ILV wallets, and 6th with wallets under 500 ILV. The other closest candidate, Scrubadubdad in sixth place overall finished fifth with wallets under 500 ILV, but seventh with wallets over 500 ILV. All other elected council members finished above Scrubadubdad with voters under 500 ILV, and Blickter finished just behind him with 146.5 vILV for Blickter from this group compared to Scrubadubdad’s 163.2 vILV. However, Blickter had 1166.3 vILV from wallets >500 ILV, compared to Scrubadubdad’s 197.7 vILV.

Completely Removing ILV - Wallet Only Analysis

Removing any impact of ILV holding and only evaluating participating, Kieran again dominated this election with a total of 92 wallets selecting him, or 21.6% of participating wallets. Animositas jumps from fourth to second with this perspective, with 68 total wallets selecting him. Deraji and Scoriox were next, with 56 and 54 wallets voting for them, respectively. 

The next tier of candidates all received between 20 and 27 wallets, with Scrubadubdad at 27, followed by Blickter, Artemis and Vetemor. Again, Blickter held just 4 fewer wallets than Scrubadubdad, but a significant edge in wallet size for these participants.

In total, 23 candidates received votes, with 17 candidates receiving 2 or more votes.

Going the Other Direction - Impact of ILV

If we instead look at how the election would have gone if there was no quadratic voting and look purely at the power of held ILV, we see a bit of a different picture, but with the top 4 remaining the same. Overall, Kieran again dominated, but accounted for over 71% of the total ILV participating in this election. Blickter finishes second again, but with about 750k ILV fewer than Kieran. Deraji and Ani again finish third and fourth, though Artemis gets the fifth spot with 1.2% of the ILV (15,716 ILV) compared to Scoriox in sixth (11,743 ILV). Scrubadubdad drops to ninth in this metric with 10,394 ILV.

Regardless of how the election data are evaluated, the system selected the top five candidates by community preference, with only Scrubadubdad entering the conversation for potential other winners based on alternative analysis.

Summary

Despite the challenges of the restart and broader market sentiment/engagement challenges, the community stepped up and selected their new council. I view this election as a success for blind voting, and positively look forward to elections being held in this manner to accurately judge the community’s preference for candidates.

Thanks for reading and feel free to share your thoughts or other aspects of the election you’d like to consider in the future.