“Did they announce their whitelist FAQ?”

“Yes !! And it’s really easy too !! Simply follow them on Twitter and join their Discord !!”

“That’s it?”

“That’s where it begins !! You’ll then have to send 420 messages every hour in their discord and tag 69 friends for every tweet !! You can also increase your chances by uttering a prayer to the founders on VC, building a shrine to the project in your room, painting a mural on your walls or renaming your firstborn to MHPTITNAR.ETHXYZ !!”

“… I’ll buy $DOGE and $SHIB instead”

TL;DR: Whitelist mechanics could be designed in a more holistic manner by combining on-chain activity of wallets, adjusting discord bots to identify keywords and adjusting metrics to focus on quality instead of quantity of contributions

Separate Note: This is a pretty long shitpost, you can scroll till you see the MEE6 for existing and proposed whitelist mechanics if you have the attention span of a chihuahua

Before the ohmmie uprising, whitelists were predominantly distributed via totally-fair-and-transparent lotteries awarded to the luckiest shitbags holding on to shitcoins, whether they took the form of launchpads or exchanges. The hoi polloi were but ignorant fools back then, offering up our sincerest prayers to the gods of RNG in hopes we would be chosen as apostles and bestowed the right to dump on our fellow plebs. It was a simpler time, we bought up any shitcoin that had the suffix of “starter” or “pad”, joined their Telegram channels and subscribed to every token launch we could.

Then one day, our lord and savior Zeus decided enough was enough and started a revolution that would kick off countless forks spanning across every corner of every chain. The details are irrelevant since anon cannot comprehend DeFi or curve or the intricacies involved in naming quadruple-wrapped-leveraged-staked-bridged-voting-power-boosted-governance-tokens. What was relevant was the by-product of that movement, the start of the rat race known as “whitelist grinding”, the beginning of the end.

The Grind Begins

Phase One: Fastest Fingers First With a Boost

The early ohmmie days were simple, anons would smash that like button and ring the bell on Twitter before flooding through the gates of discord to get their OG roles. This archaic practice has stood the test of time and we see it employed even for current launches. Its effectiveness lay in its simplicity, anons are rewarded for the commitment to the cryptosphere and the project, effectively being put on standby 24/7. This also encouraged early adoption and organic user engagement, which are strong plus points all around. Another common thing to do back then was to reward users that boosted the discord server, although this was too simplistic and is no longer in practice today.

There was no way to know back then that this small trick would mark our gradual conversion into mechanical bots.

Phase Two: GMs, GNs, Hewwo Fwens

The sharp utensils quickly realized it didn’t make sense to just be giving out whitelists to the 20,000 earliest joiners for some pretty obvious reasons:

• It was stupidly easy to bot

• It was not equitable

• There was no incentive for long term support of the project

This was the point where we started seeing the early iterations of the whitelist grinding meta that we know today. Common tactics included:

• Invite competitions - Still in practice but falling out of trend

• Participation (gm, gn, be “helpful”) - Apparently the pinnacle of identifying loyal supporters, totally not prone to sweatshop abuse, currently the de-facto standard

• Engagement (asking questions in AMAs mostly) - Still in practice but used as part of a larger toolkit now

• Content Creation (memes and fanart) - Falling out of flavor too as project teams finally realized fiverr artists can be hired by anyone, not just by NFT teams

Phase Three: Peak Engagement

After numerous trials and tribulations, this is where we’re at right now, the zenith of engagement. Whitelist grinding has effectively become a fulltime job, except that if it were truly a job, your boss would definitely have you replaced by a bot in the coming months. Congratulations everyone, a great round of applause, we’ve made it.

The ineffective tricks of the past have been replaced and the effective ones have been enhanced to maximize pain disguised as filtration of quality. Here’s what we ended up with:

• Participation remains the gold standard, daily gm/gn expected, preach the gospel in #general, good vibes only

• Engagement has mutated into its final form, games on discord are frequent, AMAs/spaces/VCs occasionally reward participation

• Collabs, giveaways, twitter raids and alpha group partnerships are the bread and butter of organic growth, and with good reason

• Raffles are frequent enough to fill up the leftover slots and allow for the illusion of fairness

• Invite competitions are still around, and most have included some form of bot filtration in an effort to maximize value

The Need for the Grind

At the moment, the grind is a necessary evil as NFTs work on a fixed supply structure and in most cases, demand exceeds supply. This means that everyone and their pet chinchilla want the whitelist because it is literally free money (at least until Winnie The Pooh mints an NFT and causes the inevitable collapse of this sector but we’ll save that for another time). Now here’s the hard part anon, should the precious and limited whitelist slots be given to:

A) a bunch of randos on twitter, or

B) only the best, long term supporters?

If you answered A, you’re an idiot, hfsp.

If you answered B, congratulations, you’ve eliminated yourself from that list because you, dear anon, are a flipper, hfsp.

For some context, designing a sustainable NFT project is easier/harder than a fungible token play depending on your perspective of whether more/less moving parts is easier/harder.

Fungible tokens are a much more established sector by this point, with a myriad of tweaks and settings that can be adjusted. These range from conceptual aspects such as vision and product design, to mathematically sound principles in tokenomics design. Distribution channels, investors, liquidity avenues, launch platforms, exchanges, campaigns, these are all things that can be tuned to give fungible projects the early boost they need and get them where they are going. And that boost makes an entire world of difference when assessing the longevity of any project.

Therefore, if we were to exclude the eternal constant which is the team, then due to the restricted nature of NFTs, the whitelisting process is THE SINGLE MOST important factor that determines the viability and success or failure of the project over an extended period of time. The selection of early stage investors decide whether the NFT can successfully navigate through price discovery, identify a strong floor and defend it to set an anchor. More flippers = shittier floor. Other potential factors would include:

• Marketing/distribution channels - extremely important but not enough on its own

• Utility - which is barren at this time aside from those of alpha groups

• Roadmap - or at least pretending to have one (mega shitpost on this but we’ll save it for another time)

• Listing - possibly in the future, we’re seeing an increased number of CEXes trying to get involved with the NFT space right now, impact is debatable

The “Perfect” Whitelister

In the eyes of the project team and “long term investors” as all anons claim to be, here are the traits of the ideal whitelister, from most important to least important:

1. Big bagger - able to fall for fomo and buy more NFTs from secondary markets, unlikely to panic sell, BAYC pfp

2. Diamond hands - hopefully they lose their private keys at some point after minting

3. Le Connoisseur - in it for the art, NFT wallet shows “next page”, might open a museum in the future

4. Community member (T1) - active contribution to the community, able to generate/cycle/redistribute alpha, propose feasible ideas, structured arguments

5. KOL - distribute news and updates (shill) effectively, best if they pfp the NFT

6. Content creator - court jester of the community, able to create memes, gifs, videos, articles, other marketing collateral to be distributed, keep morale high

7. Community member (T2) - good vibes only gang, dilute fud by gaslighting fudders, mostly a moonboi, sometimes a fudder when spooked, will pfp but has no impact

8. Discord bots - WL farmers that forgot to leave the chat, pads group numbers, at least they are quiet

9. Community member (T3) - old school wen muun gang, place a curse on the team if the floor price dips by 0.001 ETH, sustains their life force through fudding

10. Flippers/WL Hunters - helping to build opensea one royalty payment at a time, at least they contribute to trade volume which is a good metric

11. Literally invested lifesavings into a jpg - bad vibes only ganga, high propensity to smash floor by listing 50% lower

This is not an exhaustive and not an exclusive list, we fluctuate between different traits at any given point in time.

For example, you, dear anon, could be a [flipper] that got rugged repeatedly and forcibly converted into a [diamond-hands-connoisseur].

In my case, I tell myself that I’m a [big-bagger-diamond-hands-connoisseur-KOL-content-creator] before I go to bed at night but deep down I know I’m actually a [community-member-T3-who-literally-invested-lifesavings-into-a-jpg].

Reviewing the Grind

With all that said, the reason the grind sucks is that it is not effective at identifying the above traits and has descended into “busy work” for community members. At a broad level, most of the activities have no quantifiable metrics and contribute nothing to the selection of whitelisters. They only serve to churn low-value engagement by dangling a carrot in front of the masses.

At a much more specific level, in no particular order:

1. Twitter camping for “Discord open to first 100 members” is a decent strategy for ultra early stage hype and fomo but it will only get us so far. There’s no way to filter the members that are joining for any of the traits other than how often they check their phones. It’s also very susceptible to botting strategies and hinders the growth of the community if run for too long. Always a tradeoff between fomo and patience. [8/10 initially but effectiveness falls off very quickly]

2. Invite competitions are extremely easy to bot or game via paid giveaways, only effective if the team has very advanced bot filtration technologies for discord. This most likely does not exists because it would used by everyone if it did. Captchas are useless. [2/10 on its own, perhaps 4/10 if there’s bot detection]

3. Chat grinding (gm, gn, #general) is pretty useless for trait identification BUT has merit as a marketing tool to churn activity and engagement in the community. There’s also some potential for improvements by adjusting how exp is calculated, possibly by assigning weightages to specific keywords in an effort to reward quality discussions and reduce the pain. Highly susceptible to sweatshop operations. If NFT teams can hire their artists and devs from fiverr, WL hunters can hire content writers and social media managers from fiverr too. [5/10 in its current iteration, positive points for marketing effect, negative points for how cancerous the process is for community members, at least half of the people bitching about whitelist grinding are referring to this specifically]

4. Engagement churning via AMAs, spaces, VCs, campaigns and other team interactions is at least decent in that it is a very filtered process and can help to select the kind of community members that can contribute effectively. Very useful for whitelisting but requires high commitment and labour. [8/10 for effectiveness but basically a pseudo interview process, not feasible at scale]

5. Engagement churning via discord games is absolutely worthless. At least some of them can be fun the first few times. It is extremely time consuming for everyone involved and provides no quantifiable insights about the whitelister. [0/10 would not recommend as a whitelisting criteria, the other half of the people bitching about whitelist grinding are referring to this]

6. Engagement churning via games built by the project team is a real coin-toss. If the game is fun (..it never is) then I think it would be an extremely good way to engage with the community and leverage them for incentivized beta-testing. There is a really good amount of merit to this but it’s still in the early stages. Does not help with trait identification for whitelisters unless the game is some psychoanalysis one. On the other end of the spectrum though, if the game mfing sucks ass (looking at you cryptogene and kaijukingz) then it is a mfing chore and could even impact the sentiments of the prospective minters. It’s so bad its actually a net-negative marketing strategy. “Muhh fun is subjectiv-” just shut up your game sucks. [∞/10 if the game is great, literal endless possibilities that can be incorporated, makes me excited for the future of NFTs] [-∞/10 if the game sucks, may piss off anons so much that they write a mfing dissertation about whitelist grinding in future]

7. Content creation used to be effective before the sweatshop operations started, now it’s just an indirect way of buying fanart off fiverr with extra steps. Memes, threads, videos and writeups (pick me blz) can still be effective if there is good distribution channels for them. Good for identifying court jesters to keep morale up when the post-reveal dump happens. [7/10 for marketing effect and ease of integration, primary negative is the sweatshop possibilities]

8. Paid giveaways “follow/like/rt for $50” are absolutely worthless [0/10]

9. Collabs and alpha groups are pretty much the one and only thing in the current meta I say with absolute confidence that it is the best way to filter and distribute whitelist slots, as long as the partners are vetted to be a good fit with the direction of the project. The only drawback is that this creates a closed and exclusive ecosystem but if at all possible, the bulk of the whitelists should be set aside for these campaigns. [10/10 but they may not want you even if you want them, might stifle adoption if used excessively]

10. Other giveaways, whether through KOLs, tag 3 friends, rt, like, follow are only effective is the target audience is a good fit, usually they aren’t. Decent way to expand outreach and help get the initial wave of interest. [2/10 for its ability to bring in users when you have literally nothing, quality is questionable]

11. Raffles are the last on the list since they are usually used at the very end to fill up any gaps. Reasonably useful as it helps to diversify the community. Premint.xyz has basic criteria that can be configured as a filtration process. Helpful but not particularly effective. Naughty teams run sybil attacks on their own raffles. [6/10 if done fairly, can be significantly improved by tuning address criteria]

This is of course not applicable to second-generation series because you can just whitelist whoever is bagholding your shit, refer to invisible friends for a good implementation.

Enhancing the Whitelist

The often employed tricks are usually only effective for identifying whitelister traits 4 to 9. By combining them with some extra mechanisms, we can re-design whitelists to achieve the following:

• Accurately identify the traits of prospective WLers for a healthy pool of minters

• Encourage contribution of high quality content in the community

• Generate hype for the project

In designing these mechanisms, I have tried to keep them flexible enough such that the implementation can be adjusted to meet the technical or other bandwidth limitations of the team. I initially made the assumption that all project teams include at least one competent developer, in which case the technical difficulty of the concepts proposed below would all be set to 1/10 when compared to smart contract development. However, I have since re-adjusted the difficulty levels in consideration of the fact that outsourcing development work is apparently a common practice.

Mechanism #1: Address Analysis

This is, in my opinion, the key missing ingredient in the current whitelist meta. In summary, addresses can be analyzed based on their activity in the cryptospace. The metrics and weightages can be dynamically adjusted to align with the team’s vision and the implementation is reasonably trivial - base script can be hammered together in a day with Etherscan API, tuning based on criteria would not take more than a day. Alternatively, just pay for nansen NFT Leaderboards and work within the scope of their available data. Conceptually, the end result would be something similar to DegenScore, but geared towards NFTs instead of DeFi.

Here are some metrics that could be useful, sorted by implementation difficulty ascending:

1. Address Age → 1/10 difficulty | 1/10 usefulness

2. Transaction Frequency → 1/10 difficulty | 2/10 usefulness

3. ETH Balance → 1/10 difficulty | 5/10 usefulness

4. Total NFTs (Curated*) → 2/10 difficulty | 7/10 usefulness

5. Unique NFTs (Curated*)→ 3/10 difficulty | 7/10 usefulness

6. Address USD Balance (ETH+ERC20s) → 4/10 difficulty | 6/10 usefulness

7. Total Address USD Balance (ETH+ERC20/721/1155) → 6/10 difficulty if you have an elusive Opensea API Key, 8/10 otherwise | 8/10 usefulness

8. Mint-to-Sale duration → 6/10 difficulty | 10/10 usefulness

9. Purchase-to-Sale duration → 6/10 difficulty | 9/10 usefulness

* A curated list could be defined prior to the checks to target specific NFTs such as BAYC, CryptoPunk, Azuki, etc.

As mentioned, the concept is highly customizable and the list above is not exhaustive. Weights and scores can be assigned based on a combination of metrics, and auto-win/auto-lose filters can also be added. As an example, an address that has >100 ETH could be automatically qualified and an address that has a consistent MtS duration of <1 day can be automatically disqualified.

A further consideration is that this mechanism can work complementary to existing whitelist procedures as a secondary filter, or on its own as a standalone whitelist process. It’s important to note that if it is used as a standalone process, the whitelists set aside for this mechanism should be limited to an upper bound of 50% to 75% of the total supply, as it may result in a walled-garden if used excessively.

Mechanism #2: Revamping Discord Chat Grinding

The bulk of the complaints regarding chat grinding is because it is an extremely mundane and repetitive task that isn’t meritocratic in nature. However, the marketing benefits are enormous due to the amount of activity it churns out. In my opinion, minor tweaks can be added to the chat grinding bots for significant benefits:

• Reduced exp for target keywords - “gm” / “gn” / “hello” / “good/nice project” / “when” / “promotion” / etc.

• Increased / Reduced exp for message length - ideally on a cumulative normal distribution, exp increases based on message length up to a threshold before diminishing returns, this should be used as a multiplier in addition to other filters

• Increased exp for target keywords, used in tandem with message length multiplier - “roadmap” / “utility” / “discussed” / project specific keywords such as their name or upcoming products / etc.

• Increased exp for consistency across the day - a multiplier that increases the exp of a message based on time since last message, this rewards participants who are active throughout the day and penalizes those on a shift or schedule

Mechanism #3: Transparent Raffles

This is just a slight upgrade but raffles can be used in tandem with a less stringent version of Mechanism #1 for qualification. This eliminates the possibility of sybils. Address submissions should be shuffled and the results should be published publicly, then an offset can be applied based on the output of a VRF. The offset mechanism is not important as long as it is provably fair, it could be as simple as adding the VRF to a sorted list or using the VRF as the first index, or as complicated as using the VRF as a seed fed into a pre-disclosed hashing algorithm.

Putting It All Together

A combination of the above three proposed mechanisms and the ones that are still effective today (AMAs, Collabs, Alpha Groups, Content Creation) would be an ideal direction for whitelists to take in that it strikes the delicate balance between qualifying minters and generating hype for the project. These concepts are also not intended to be a one-size-fits-all approach and should be adjusted for alignment of visions.

Lastly, in case it was not obvious, all mechanisms should not be publicly disclosed to avoid gaming the system, at least until after the whitelist process has concluded.

Thanks for sticking through this long post!

If you like being insulted in creative ways, anime-crypto memes, occasional insights into blockchain related tech, unactionable and downright useless alpha, follow me on twitter @0xS4gas (pronounced SAH-gas, the ‘4’ is 1337). You can also send shitcoins to S4gas.eth / 0x00001cF7FC5cdFA1be94A3359c7659ec50c60000 so that this ape can get some bananas.

“Hey anon, have you heard of rarity ranking and scores ??”

“What are rarity scores”

“It’s an arbitrary number used by uncultured people who can’t appreciate art to decide which pieces should be worth more or less, even though they’re all worth the same amount of nothing :)”

“THATS ME”

“Simply look up any collection at [insert favorite rarity sniper here] and you can see your NFTs ranked by their rarity”

“But how does it work”

“Who cares ?? You just buy what they tell you to because the highest ranked items all look the best !!”

TL;DR: The lack of optional traits and the total number of traits are factored into rarity calculations. In theory, a blank NFT with no traits would be the rarest piece of every collection.

Over the past couple of days, I built a rarity sniper and compared it to publicly available ones (nanocube/raritysniper/traitsniper/etc.). I thought it’d be a quick and easy project but instead it took a journey of self discovery before I got it right.

I’ll be using Azuki as the primary reference since its pretty and I like it. [I do not own an Azuki yet so its fair game for me to shill it]

I’ll also be using nanocube as the primary reference since they include two methods of calculating rarity score (classic and trait normalisation).

Attempt

For my implementation, I hackerman-ed the following:

• Downloaded the metadata for every tokenId

• Extracted the traits of each tokenId

• Calculated the rarity of each individual trait based on the number of times they appeared in the entire collection (rarity = traitCount / totalSupply)

• Multiplied the rarity of all traits for each tokenId

• Took a moment to feel good about my script kiddie skillz

I was pretty pleased with myself for having generated my own list in such a short time until I realized the hit rate was completely off.

Revelation

The abysmal performance of the first generation came as a pretty big shock so I went to take a look at the top few pieces on nanocube to find out why.

The highest ranking piece (Azuki #2152) immediately struck out since it looks ordinary as shit and caused me to seriously think if there was a bug or advanced anti-rarity-sniping mechanism at play. However, a deeper dive into the scoring system revealed the answer, in that the two highest ranking traits were - <Quantity of Traits: 5> & <Hair: none>.

This means that the principle applied to such rarity scoring mechanisms factor the lack of a trait as a trait in and of itself, which makes sense from a certain perspective.

If that doesn’t make sense to you, dear anon, then consider this example instead:

If you were to attend a crypto conference of 10,000 people and everyone else has a brain, you would be the most special person in the room because you, dear anon, do not have one.

Conclusion

The impact of such scoring mechanisms means that NFTs with fewer attributes, whether by sheer luck or intentional design, would rank higher on rarity lists (except in a situation where the majority of attributes have a <50% chance of being generated).

This may seem counter-intuitive at a glance but there is precedence set by collectors of artwork, trading cards and figurines where misprints or defective pieces are able to command higher prices due to perceived rarity or scarcity.

This also means that the rarest piece of most NFT collections would in fact be a blank drawing, although in practice, the artwork is usually fully designed despite having empty metadata (The Other Side is one example of this).

More commonly though, NFT collections are using attributes like “Name” for special or one-of-one pieces, which applies the same logic and bumps the rarity score significantly as these are often unique to only a very small number of pieces.

If you like being insulted in creative ways, anime-crypto memes, occasional insights into blockchain related tech, unactionable and downright useless alpha, follow me on twitter @0xS4gas (pronounced SAH-gas, the ‘4’ is 1337). You can also send shitcoins to S4gas.eth / 0x00001cF7FC5cdFA1be94A3359c7659ec50c60000 so that this ape can get some bananas.