Should Compound Retroactively Airdrop Tokens to Early Users?

—-Agree with most of your points. (Gas price especially)

But with respect the above, I also see the flip side where large investor groups are accruing substantially more in a shorter period of time with less risk by providing a stable currency and borrowing another. Eliminating the risk many/most early users endured.

To me those early users that contributed and withdrew their c-coins a short period of time later would likely sell their comp but those that don’t will likely be encouraged to engage and contribute to governance. Just my 2 cents. (Fighting for the old school, small timers)

Hi,

I have created a spreadsheet containing the raw data (csv) that was generated by @allthecolors. Thanks again, btw! The spreadsheet contains a formula with variables in which the data can EASILY be manipulated to change the amount of COMP that is airdropped to each address (simulated airdrop).

The formula has multiple parts, which allows for a mathematical approach to determining the balance between a social and capital based distribution method:

Variables:

  • cZero = amount of COMP distributed to addys who earned 0.00 interest
  • iLowAmt = lower bound interest earned from addy
  • cLowAmt = amount of COMP distributed to addys who earned less than the iLowAmt
  • cReduceCapital = this is the number that divides the square root of the interest earned
  • cAddSocial = excluding the lower bounds interest earned, this is the least amount of COMP each address should receive

Formula:
An Addy with 0.00 interest gets cZero COMP
Addy < iLowAmt interest gets cLowAmt COMP

(Sqrt(interest) / cReduceCapital) + cAddSocial

The spreadsheet:

The spreadsheet contains global airdrop data (left side) including the total amount of COMP distributed in the simulated airdrop, the smallest and largest single address distributions and the amount left over from the proposed 500,000 COMP, which I have labeled as Community COMP airdrop. Which I think should be divided equally and airdropped to everyone who participated on this forum post.

Using this sheet to simulate airdrops:

cZero variable

The cZero variable is included because of the multitude of addresses which interacted with Compound, but without even earning a penny’s worth of interest. This could have been due to several reasons including the possibility of flash loans being tested in the earlier days, hackers trying to find a way to break the protocol, Sybil attack, inside knowledge of a possible airdrop or many other reasons, none of which IMO deserves a substantial amount of COMP to be airdropped to each address. Also, these addresses which earned less interest have a greater chance of being abandoned, unless they are from malicious actors. I would even consider forcing those addresses with $0.00 earned interest to need to perform some type of claim, like signing a message.

iLowAmt and cLowAmt variables

These are other variables used to determine who and how much to airdrop/reward addresses which interacted with Compound in a way that couldn’t be due to flash loans, or other “same block” instances, but also didn’t risk as much capital. Also, would imagine that a greater than average percentage of these addresses are abandoned/lost. So, again, considering a claim function or message signing operation for these addresses may not be a bad idea.

cCapital and cSocial variables

These variables are directly related to capital and social weighting distributions. Instead of using a percentage, I developed a mathematical equation in which can be modified by anyone on our forum to modify for easy and quick simulations of airdrops.

Please send me a request if you would like to manipulate the data within the spreadsheet. Actually, you can just copy it, if you have any problems or questions about how to use the spreadsheet, let me know. I have pre-filled the variables where I believe is a fair distribution model.

@Fishbtc, this is able to provide 2 levels.

@cryptobuddy_1712, lets not let it expire, take a gander.

Really, using the interest earned, is taking the amount of time in consideration. I mean, someone supplying $50,000 for 5 days should earn nearly equally the same interest as a person who supplied $10,000 for 25 days. If time wasn’t already a product of interest earned the person who supplied $50,000 would get 5 times the airdrop right?!?

Just needed to ask @getty if this is helpful?

Also, would be much appreciated to hear feedback from @rleshner, @arr00, @Sirokko, @massnomis, @TylerEther and @alive. Thank you!

5 Likes

This needs to be brought up at the next dev call, or should have been brought up a the last dev call! If an idea of this magnitude (both user and COMP) fails to reach the proposal phase, that is absolutely crap. It should be required to be brought up during the call, or really needs to be a calander somewhere that members can check for important dates.

Can we do a simple proposal, just voting on that there will or will not be some type of airdrop? That should get plenty of votes. Also, I hope you already saw the spreadsheet that I created allowing everyone to be able to simulate airdrops, even without a single hint of developer knowledge.

2 Likes

Great work @CryptoCraig . I really agree with you. Lot of people are following this thread since months hoping for an action to be taken but unfortunately there isn’t any. Hope there will be some conclusion drawn in next couple of weeks from all the contributions. Thank You.

2 Likes

Thanks man, really I was just looking back at it and realized that @allthecolors data needs a way to enable the community to be able to manipulate the data in a way so that they can see the scope of the airdrop. My sheet is designed to allow anyone to be able to change the metrics simply by editing the 5 variables highlighted in yellow on the sheet. Copy the sheet and then edit the variables to get a distribution that you see fit, then share it with us!

Use my data as a guideline. Let me know if you need help creating your own airdrop simulations!

1 Like

I’ve been following this discussion for a while, and I just created an account to say that the points and suggestions that @CryptoCraig raised are in my opinion valuable and really worth considering.

Taking into account that the goal of the proposal is to improve governance decentralization (that is, wide distribution to active addresses and users) I find there is little downside to adding a lower bound that significantly reduces the airdrop to addresses with trace amounts of interest, and making that process “opt-in” instead of an actual airdrop for wallets that have a high likelihood of being lost.

I personally think there is an argument to be made for filtering out accounts with accrued interest < 0.01$ completely, for the reasons mentioned in this previous post. FYI, the total amount awarded to accounts with accrued interest in the spreadsheet < 0.01$ is 21392.5 COMP.

2 Likes

Agree with that as well

Thanks very much for sharing this tool to help put the interest data into the community’s hands! I went back and forth about including this step in my work. The reason I held back is my agreement with the grants committee to deliver the interest data and not to conflate it with the discussion of a distribution model. So I really appreciate your effort to fill in this gap (and tagging of key community members to continue this conversation!)

1 Like

My recent suggestions are more geared around distribution methodology and timeline than comp allocation. I’m onboard with the distribution allocation suggested by @allthecolors with your recommendation of minimum qualification interest of greater than .01.

Provided we have the ground work for the comp governance distribution not sure if those in the position to bring this to vote are ready to do so. @rleshner ?

Even once a distribution and mechanism are decided on, there is still smart contract development work ahead of us before a proposal can be submitted. I’m optimistic that once we develop a clear specification for what needs to be implemented, we’ll be able to recruit someone experienced with COMP’s smart contracts to bring it to life.

2 Likes

@gabrocheleau, just make a copy of the Airdrop spreadsheet and create simulated airdrops within google sheets or excel. This is a tool for the community, I just wanted to give everyone an idea as to what an airdrop could look like. On the spreadsheet is now a description of each variable making it easier than ever for community members to link to their distribution simulations.

Play around with the tool, it may take you changing every variable 2 and 3 times before you get to the point where you feel like “that’s it!”

The ONLY reason this was created was to make it as easy as possible for community members to visualize what an airdrop looks like with their ideas. If you need any help PLEASE let me know, I will be more than happy to help. There have now been 23 link clicks and I have yet to see another simulation ran with the tool, which is kind of upsetting to me because I want to get the community involved. After all, Compound is a DAO and what is a DAO without a community?

1 Like

we cant change variable

thanks for your tools.

Thanks, yes I actually made a copy yesterday and started playing with it. I’ll try to come up with something that makes sense to me and share it here.

@bulajacky You can’t change variables on that document, you must first make a copy. To do that just log into your Google account and click File > Make a copy.

image

2 Likes

ok
thank u
i will try it later

1 Like

Hey @CryptoCraig!

So I’ve played a bit around with the numbers. The issues I was trying to address were the following:

  • As previously mentioned, I do not see value in distributing COMP to addresses who have earned less than 0.01$ on the protocol. For the reasons you mentioned (likely flashloans, attacks, tests, etc.)
  • To me, it also did not seem reasonable than an address right below the iLowAmt (e.g. 0.49 interest earned) would get 5 COMP, while an address right above (e.g. 0.50 interest earned) would get 5 times more: 25 COMP.
  • I felt like the distribution curve could be “flattened” and the median received amount by lower interest addresses increased. I believe this to be consistent with the spirit of this airdrop.

With that in mind, and keeping the amount of COMP distributed to a max of 500k, I have done the following:

  • Set the COMP airdrop to 0 for addresses with < 0.01 interest accrued
  • With this freed COMP, I increased the social rewards to 30 COMP
  • I have increased the iLowAmt to 1.00
  • Instead of giving a fixed amount of COMP to addresses below the iLowAmt, I have set it to Max(cLowAmt, cAddSocial * interest earned). Basically, this means that the rewards start at 5 COMP for 0.01 interest earned, but quickly increase to 30 COMP as one approached the iLowAmt. This removes the huge discrepancy described above and I believe makes much more sense.
  • Lastly, I have flattened the distribution by adjusting the cReduceCapital parameter to 12.5.

So in practice, this means that:

  • An address that earned trace amounts (< 0.01) of interest gets 0 COMP (vs. 2.5 COMP)
  • An address that earned between 0.01 and 0.15 of interest gets 5 COMP (same)
  • An address that earned 0.49 of interest gets 14.7 COMP (vs. 5 COMP)
  • An address that earned 0.50 of interest gets 15 COMP (vs. 25.28 COMP)
  • An address that earned 1.00 of interest gets 30 COMP (vs. 25.4 COMP)
  • An address that earned 200.00 of interest gets 31 COMP (vs. 30.60 COMP)
  • The address that earned the maximum interest (3271357.72) gets ~175 COMP (vs 748 COMP)

Here is the document:

Let me know your thoughts!

2 Likes

Thanks for putting this together. This step is crucial in turning hypothetical discussion into reality.

1 Like

I like this distribution model as it targets the "middle " more effectively which will achieve the goal of decentralizing the protocol. In my opinion getting voting rights to users who legitimately engaged with the protocol is top priority. Distributing 5 ETH to 150 addresses is much more effective than 1 address getting 750.

Omitting addresses who contributed trace amounts also helps with this IMO for the reasons you mentioned.

1 Like

i lend my eth at comp… but that time they only offer 0.01% interest… of course i eligible to get 0 comp LOL

I would argue that even at 0.01% interest (was it really that low for the whole duration of the loan?), one would have to lend 1000$ for ~40 days to reach 1 cent worth of accrued interest. A penny of interest is really not that high a treshold imho, but this is worthy of discussing.

2 Likes

Early users should be defined as anyone that genuinely interacted with the protocol prior to the COMP announcement. For example, Uniswap gave UNI to anyone that made a trade on their platform prior to a certain date.

Excluding genuine users that either did not deposit enough funds to earn an arbitrary amount before a certain date would disservice to those early users. I argue that if any user of compound deposited any non-dust amount of funds for any non-negigible amount of time prior to the announcement is an early user.

Example: Alice deposits 1 ETH one week prior to the announcement, but doesn’t earn a full 0.01 during that time due to low interest rate. Alice still holds her Compound ETH today. Isn’t Alice an early user that took risk with her capital to engage with Compound?

I urge the group to prioritize COMP distribution to all early genuine users, not just the extremely early or very wealthy ones.

4 Likes