An Idea to Improve Coordinape [DRAFT 2]

An Idea to Improve Coordinape [DRAFT 2]

Changes from previous draft

A number of changes have been made thanks to the amazing feedback on the previous draft:

  • Removed references to stripping L2 status for non-participation
  • Added participation requirements for L1’s and L2’s
  • Reaffirmed that Guest Passes cannot allocate GIVE (for sybil resistance)

Background

Coordinape is the product we use to distribute 4.5M BANK a season for contributor rewards. This represents around 15% of our total seasonal budget. This BANK is distributed monthly - 500K BANK respectively for Guest Pass, L1, and L2 contributors.

Contributors sign up for the coordinape round monthly and allocate GIVE to their fellow contributors. After a week, the GIVE amounts are tallied and then converted to percentages, which are then used to allocate the pot of BANK for each member circle (i.e. if you receive 1% of all GIVE, you are rewarded with 1% of 500K BANK, 5000 BANK).

The intent behind coordinape is to compensate contributors for their work the previous month. Put another way: the goal of coordinape is to encourage contributors to contribute.

The Problems

  • Coordinape is complicated, especially for BanklessDAO newcomers. There are 3 circles (one for each member level), and it can be difficult to understand how to use the system, much less how you should be allocating on each circle.
  • Coordinape doesn’t reward contribution
    • The Ops Guild recently shared that there is only ~65% of L1 and L2s who signed up actually allocated GIVE. Because of the way GIVE translates to BANK, contributors are incentivized NOT to allocate, because they get a bigger piece of the pie when they don’t.
    • Our L2s are our most active contributors, and it’s possible (and probable) that less active Guest Passes receive MORE in coordinape rounds than these extremely active contributors simply because there are more L2s sharing the same 500K BANK. I’ve personally dumped a ton of GIVE on a Guest Pass simply because I didn’t have anyone else to give it to.
  • Coordinape doesn’t encourage contributors to move up member levels
    • In May’s coordinape allocation, the top Guest Pass and L1 each received around 2x more BANK than the top L2. I’ve even seen contributors refuse L2 nominations.
    • Since Guest Passes can’t allocate GIVE, it’s desirable for them to stay as Guest Passes to benefit from coordinape without contributing to it.

The Idea

  1. Convert the 3 circles into a single BanklessDAO circle.
  2. Only distribute rewards to L1 and L2 who allocate a certain percentage of GIVE. This doesn’t apply to Guest Passes are they aren’t allowed to allocate GIVE (to prevent sybil attacks)
  3. Use pirate math to weight rewards towards higher-level contributors: Guest passes get 2 shares, L1’s get 3 shares, and L2’s get 4 shares. How this works in practice is that after all GIVE is allocated, L1’s and L2’s will have their GIVE multiplied by 1.5 and 2, respectively.

An example using 1/1.5/2 multipliers and requiring 50% allocation of GIVE for L1’s and L2’s:

POLL - How much GIVE do you think L1’s and L2’s should have to allocate to be eligible for rewards?

  • 100%
  • 75% or more
  • 50% or more
  • 25% or more
  • I don’t think L1’s and L2’s should have to allocate GIVE to get rewards

0 voters

The Benefits

  • The coordinape process will be simplified. Instead of having 3 circles, there’s just one, which makes it easier to teach people how to use it.
  • Coordinape will more closely reward contribution.
    • With everyone in the same circle, it’s easier to align your GIVE with who you’ve actually seen contribute, rather than run out of GIVE in the L2 circle and have excess in the Guest Pass level.
    • With a requirement to participate to get rewards, we’ll see more L1’s and L2’s participating in coordinape itself.
  • Contributors will be encouraged to move up member levels because of higher rewards. This means contributors will be encouraged to hold onto BANK. We often talk about utility of BANK…this is utility!
  • This could encourage L2’s to stick around longer. I can’t be the only person to notice that L2’s have a habit of reducing their commitment after a few months. These are, by definition, our top contributors. We need to encourage them to stick around and contribute.

The Impact

In the previous post, the 1/1.5/2 multipliers were vastly preferred. Another way to look at this: Guest Passes get 2 shares, L1’s get 3 shares, and L2’s get 4 shares of the “loot”.

Here’s a little spreadsheet to see the impact of this change on May’s Coordinape (see May Coordinape 1/1.5/2 Tab). Note that this simulation doesn’t take into account how things would change if all contributors were in a single circle.

Next Steps

  • :white_check_mark: Work with Coordinape Workstream in Ops Guild to see if it’s technically feasible
  • :white_check_mark: Gather consensus in a forum post
  • :white_check_mark: Draft a second forum post to solidify consensus
  • :white_check_mark: Bring to Grants Committee to determine if we need to run a Snapshot
  • create a Snapshot
  • Implement the proposal for the next Coordinape round

What do you Think?

I’d love your feedback!

POLL - Do you think we should implement this proposal as stated above?

  • Yes indeed!
  • No, and I’ll tell you why in the comments.

0 voters

2 Likes

Gm y’all!
I’ll try to make some points, ask some questions etc.

Isn’t decreasing GP’s share on the coordinape contrary to this? especially for those who really contribute as a GP.

I don’t think so. I mean, finding your way to bDAO and deciding to join needs way more information to comprehend and use. But there might be a lack of source on how to coordinape and that is another thing. This tendency towards making everything easier kinda bugs me. It is not that hard to learn these stuff.

yes but that’s also about how people tend to allocate. I think fame is at least as powerful as contribution if not more. For me that’s the biggest issue of coordinape.

I’m not so sure about this. In my case, it actually encouraged me since I saw an increase in my earnings and I bought some bank and boom! also for the first issue mentioned under this title; why a gp with a higher earning in coordinape - which is their only income - is a problem exactly since he/she/they contributed remarkably and awarded fairly?

Also I cannot understand the reason behind not allocating give if you have the chance to. Why someone doesn’t allocate his/her give? As far as I understood that’s about earning more bank and if so, that MUST be 100% as these kind of drifts are definitely NOT how we dao.

on the contrary, gp’s might tend to not care much since their share will be lesser. I checked my position in the latest coordinape and according to 1/1.25/1.5 ratio, I earn sth like 30-40% less. Especially in the current market conditions, it might be way easier to give up on contributing for them.

All in all, I love the fact that coordinape is being questioned but we might be searching for the problem in wrong places or overlooking some other issues. Sorry for the huge answer and lots of tagging. I might be missing some points and/or talking nonsense but hey, there’s no stupid questions, right? (=

I can probably help with some of your questions, especially coming from experience.

“Isn’t decreasing GP’s share on the coordinape contrary to this? especially for those who really contribute as a GP.”
It will feel that way to the sectors that are currently getting the heaviest relative rewards. The trouble here is Goal of coordinape != Goal of this proposal.
As I understand it, the true objective of this proposal is to eliminate imbalances in reward allocation created by silos. The original intent of the silos was to make it easier to recognize the different sectors of contributors for the allocator, and equal weight was decided on for simplicity, though I warned at the time these imbalances would emerge.

In terms of the “Coordinape is complicated” I think links means as much about their UX being non-intuitive/ efficient and therefore friendly to repeated use. It’s not so much that it’s complex, it’s that it’s clunky.

Fame is definitely a problem with intent vs. truth in coordinape, The elements of things like sourcecred do a bit to answer this problem, but the tradeoff is they pervert incentives and create non-desirable behavior in an effort to game the system. This has been a pain point for our social systems, Current solutions are inadequate and therefore we use an imperfect proxy with the explicit understanding that the social expectation at the individual level is to uphold a voting method that recognizes and adjusts for this bias.

In terms of member levels, You touched on something that we’re having to consider as a DAO that is pretty unpleasant, but part of reality.
“Which is their only income”
I’ve been a champion of working for DAOs and living within these ecosystems since the beginning. I even did it for a while. However, I’ve discovered some truths about closed loop economies like BANK, and one of them is: If the Guest/member isn’t generating an equivalent or superior amount of buy pressure on Bank in that same time as it took to earn what they get, and they sell to live on, they create a negative feedback loop in the economic system that left unaddressed will kill the flow of that economy.
Getting back to the member level thing though, Just because the current system does the same encouragement doesn’t mean we shouldn’t adopt an upgrade when developed. The level system is long overdue for a ground level overhaul and it’s on the GSE to-do list. When that comes around, things should be simplified quite a bit as two independent systems of recognition (Membership/Contribution) are separated out of the single level system of today.

The give allocation requirement is a bit heavy handed and creates unnecessary admin overhead imo, because of the way that coordinape treats give allocation. I think awareness is a big reason for this.
Coordinape uses a system that treats unallocated give like burned tokens, which means if you allocate 10 give out of your thousand, those 10 give are treated with 10% weight/give. If you allocate any give at all, you’ve fully allocated. If you allocate no give, then your weight is equally distributed among the weights of those that did, essentially strengthening their votes weight. I’ve missed allocating in the last two rounds, despite it being on my to-do list, because I’ve been doing other things in the DAO. By enacting a limit like this, it would potentially negate all my other contributions because I missed a deadline. That seems to me against the spirit of the original intent of coordinape. In practice, It’s going to add admin overhead where there need be none, because even receiving at all is gated by signing up in the forms each month I think that is enough of a minimum activity barrier. If they contributed nothing of value, their peers will not reward it, and those who attempt to game the system in that way will walk away empty handed.
Awareness around that function I think could be better documented as a takeaway though

For the final point: That’s the short term symptom of the long term effects of the “Which is their only income” problem I mentioned earlier. We need to develop methods of compensation for those who must sell that keep our economic bandwidth from being eroded, and still meet the needs of the contributor.

I’d be curious how you guys think my solution for that issue looks, which is found here: An Idea to Improve Coordinape - #47 by AboveAverageJoe

Finally, I love the deep thought and responses, It’s encouraging to see these posts taken seriously.

5 Likes

The shares quoted correspond to 1/1.5/2 multipliers.

Shares corresponding to 1/1.25/1.5 multipliers would be 4 shares (gp), 5 shares (L1), 6 shares (L2).

===========================
Please ensure we have 1k+ GIVE to allocate. Wouldn’t want to try to split 100 GIVE 20+ ways (congrats! you get 4 GIVE :sweat:)

Sorry, that was a typo in the post. 1/1.5/2 multipliers were the preferred ones.

Fixed it!

1 Like

I really like that proposal because of the simplification of the coordinape process and because you incentivize users to get to L2 for higher rewards → Holding more bank, adding value to bank.

In terms of how much needs to be allocated: I would not advise going for 100% because there are rounds where this would force some “random allocation”

1 Like

Thanks for your feedback @okachu ! I’ll try to respond to your concerns.

BanklessDAO seeks to onboard 1B people to crypto, and to do that we have to reduce as many barriers as possible. It feels like you’re coming at this from the perspective of someone with technical knowledge who lives in a western, english-speaking country (correct me if I’m wrong) - have you thought about how difficult it would be to use coordinape if you weren’t as well-versed with technology, or if your english wasn’t as good as it is? How about if you looked up a tutorial in your language only to find that BanklessDAO uses coordinape differently than EVERY OTHER DAO?

UX is the biggest gap between web3 and mass adoption. If we want to have mass adoption, we need to make things simpler. In my opinion, the complexity of our coordinape system doesn’t add much at all, and it greatly decreases participation. What this proposal aims to do is reduce needlesss complexity. Can you tell me why 3 circles work better than 1 for BanklessDAO?

I used to think that too, but some more time in system, I changed my mind. I also tried to think of ways to decouple “fame” from coordinape, and I haven’t had any luck. Do you have any ideas?

There are lots of reasons people don’t allocate GIVE - it’s time consuming (this is probably the biggest reason), they forgot, there’s other more important stuff.

The cynical reason is because if you don’t allocate your GIVE, you are actually increasing your share of the pie. Example: Coordinape with 3 people, and 2 allocate their give 50/50, and one doesn’t allocate. The one person who didn’t allocate has 100 GIVE and the two who did have 50 GIVE. This is a simplistic example, but you see that the current coordinape economics actually incentivizes people to sign up and not allocate.

…OR, would they realize that if they hold BANK, they can get more rewards? Some may stop contributing, and others will attempt to elevate to L1. We don’t know what will happen…let’s find out =)

Don’t be sorry! This is contribution and I value it =). Only together can we achieve our. And hey, now that I know who you are, I’ll be sure to recognize your contribution in coordinape. By speaking up, you are increasing your own fame (through contribution) so maybe fame’s not so bad a metric :wink:

1 Like

You’re very welcome @links and thanks for that great response. Let me speak some more :smiley: :

I can completely understand your point but I was talking for the people like me actually. I mean like the ones who are already working here and knows many. In terms of zero-knowledge people, you’re right and that might be sth. i missed. But isn’t there a timeline between joining to a DAO and benefiting from coordinape where people get to learn things? Also you’re right about the ux and language perspectives but might that be about coordinape itself to develop? I’m sure we’re giving looots of feedback to them.
I’m Turkish btw. (=
And I don’t have any objections to one big circle.

The first thing comes to one’s mind is to create a separation between work and what we refer as “fame”. But when thinking, this might bring some micro-management issues which no one would ever want here, presumably. Even if it’s ok with the people, that’s a huge load of work. So for now, unfortunately no (= But still, I think couple of people here has the same concern so let’s see if someone would come with a reasonable solution.

I’d suggest a “no give no receive” option, at least to a certain percentage but to quote from @AboveAverageJoe, “By enacting a limit like this, it would potentially negate one’s other contributions because he/she/they missed a deadline. That seems against the spirit of the original intent of coordinape.” That makes perfect sense, I must admit. I think somehow people should be more aware of coordinape’s significance, understand it better and act accordingly. Maybe we’re missing some source of information on coordinape as I mentioned before.

when focusing on increased share of L1s and L2s, you’re right. Thinking that GP’s will be more enthusiastic on spending time, working more and getting rewarded more, what you say makes more sense. I think I just saw the glass half empty. but yes, we don’t know what will happen and let’s find out (=

hahaha! Fame is not so bad of a metric as long as “through contribution” part exists. Thanks in advance for the gives :smiley:

1 Like

I love the revised proposal, Links.

I’m swayed by AAJ’s concern. I believe he’s pointing out that the calculation mechanics make the no GIVE / no receive argument moot. In any case, with 300+ contributors, the effect of one person voting with all, partial or no GIVE is minimized. Therefore, I believe there should be a low threshold (or no threshold) for giving in order to receive. I voted 50%, but could be convinced that the requirement should be removed entirely.

2 Likes

I appreciate all your work on this. Only one thing jumped out at me as a question mark. In thinking back on my last Coordinape round I know that much of my L2 GIVE was allocated to role holders in the guilds I’m active in, as a way of tipping them for jobs well done in these roles. Contrast that to when I GIVE to GPs and L1s. My story for them is that bounties aside they are not being compensated for a lot the things they do, and so depend more on GIVES for their reward. In the case of L2s who are role holders, who are already getting paid for the role, it doesn’t seem fair to me that their GIVE would receive a higher multipler (2x) than GPs (1.5) and L1s.

2 Likes

Would you consider changing the way you allocate GIVE? Give more GIVE to people who arent holding roles?

To be honest I think this change will make people rethink how they GIVE, and that could be a good thing :slight_smile:

1 Like

Yes, that is the simplest remedy. Blinding flash of the obvious really.

2 Likes

The goal of coordinape is to reward active members, if you’re having L2s, L1s and GPS in the same circle it’d get too overwhelming for mostly the guest passers who sometimes find it difficult to navigate through the coordinape system, and also this will reduce their ability to actually receive gives as L2s and L1s might allocate to themselves and then have a little portion left to share to the GPS. I am a GP myself and I depend on the Give I recieve to move up to an L1 status, this will be affected greatly. Unlike the 3 level system that gives L2s and L1s separate give for each of the circle.

  1. It should be noted that L2s and L1s are mostly role holders who are very well compensated for their role while the GPS are mostly not.

  2. L2s and L1s don’t necessarily need the coordinape system as much as GPs do, this is so as to improve the number of menders of the DAO as I see a lot of GPs transitioning to L1s if the system is left to be as it is.

  3. There should be other ways to get L2s and L1s participate in coordinape, imo most of them don’t participate in the coordinape round because they’re actually okay with their role compensation and don’t need the Gives as much as GPs who are not role holders do. Taking away this system will see a decrease in number of L1s and we all know that is not a good metric for the DAO.

1 Like

Thanks for your feedback !

Can you provide any data for this claim?

In my previous draft, I suggested that I see many Guest Passers gain roles and then become L2s quite quickly after. So is it that L2s are more likely to gain roles, or that roles create L2s?

Many guild coordinators (like the 2 most recent Ops guild coordinators) started as Guest Pass, and the current research guild coordinator is still a guest pass. So clearly the DAO is happy to give roles to Guest Passes. Why do you believe that roles mainly go to L1/L2?

3 Likes

Ruminating on this over the weekend I have a fresh concern, and potential, unintended consequence. By introducing a quantitative multiplier into Coordinape, a community practice for recognizing value, I am concerned that the process unconsciously conditions/codifies hierarchy into the relationship system we are creating. It’s quite subtle, yet potentially insidious, relative to the change we are endeavoring to BE together. In saying so I reference your tweet of last week, “The next evolution of crypto startups is to BE the change you are building. If you aren’t building with blockchain values then you aren’t building for blockchain.” As I understand it a core blockchain value is movement toward the flattening of organizational hierarchy.

As a personal example of what I am talking about here I have participated now in 2 Coordinapes. As an L1, it has never occurred to me that I be higher up than a GP or lower than an L2. To the contrary, I have regarded L2s as being further down a path that we have have equal opportunity on. The difference here is that of context and orientation. Do we orient to fellow bDAOists as on a path together, or on a ladder. I contend that participation over time in a system that distributes rewards by this method risks embedding the ladder in the lenses through which we view each other.

2 Likes

That’s interesting - any ideas on how we might measure this?

Personally I don’t see any conflict between this proposal and blockchain values, but it would certainly be good to avoid that kind of insidious mindset that L2’s are somehow “better” than lower levels. L2’s are simply people with more community acceptance in my mind, and ANYONE can become an L2 (i.e. they are not blocked by anyone).

You may know that I don’t believe that decentralization means direct-democracy OR everyone having a say in everything. I think decentralization is about resiliency and pushing authority to places where it’s most appropriate via incentive mechanisms. This proposal seeks to incentivize contributors to hold BANK and to stick around as they progress.

1 Like

Anybody knows how KPI’s from UMA are working? I think will be a good idea to have the coordinape rewards as KPI’s and release them if you reach some goals/achievment?

In that case it wouldn’t be coordinape at all, but some other compensation mechanism. I’m not against it, but the whole point of coordinape is that you DON’T have to figure out stuff up front. It’s a post-effort compensation mechanism while most others are pre-effort

1 Like

My worry is that one circle for the entire membership of the DAO makes the process unwieldy and more difficult for allocators. I personally like the fact that I can do my coordinape in stages, L2 first, then L1, then if there is time and someone is worthy, give some guest passers some.

Strongly agree that we need to steer BANK rewards to the most valued and active contributors however, and we can accomplish that easily by just upping the BANK allocation pools of L1 and L2 a reasonable amount to incentivize leveling up in the DAO. This is crucial. If people are declining L2 status for higher coordinape rewards as an L1, this is clear evidence that the current structure fails to align incentives toward members becoming core contributors.

One possibility to make this more fair is to use the cohort sizes to determine rewards share. What we want to balance is the ‘per person’ pool of BANK by dividing by the cohort size. So if the L1 cohort has N1 people signed up and the L2 cohort has N2, a balanced application of rewards would be true if rewards pool P2 = P1(N2/N1).

To incentivize leveling up, maybe we could choose it such that your coordinape rewards double when you make L2, in that case, P2 = 2 * P1(N2/N1). Similarly, to incentivize guest pass to get L1, you could say P1 = 2 * Pg(N1/Ng). And the total pool of BANK given each period of course would be Ptotal = Pg + P1 + P2 so solving the three equations above with given numbers of participants would give you the respective pool sizes. For example…

Say Guest Pass has 100 people signed up, L1 has 25, L2 has 50. You have a total of 100,000 BANK budgeted to give out. What is the size of each coordinape pool if described as above? The algebra:

Ptotal = Pg + 2 * Pg(N1/Ng) + 2 * P1(N2/N1)
Ptotal = Pg + 2 * Pg(N1/Ng) + 2 * [ 2 * Pg(N1/Ng) ] (N2/N1), reducing to
Ptotal / [1 + 2 * (N1/Ng) + 4 * (N2/Ng)] = Pg
100,000 / (1 + 0.5 + 2) = Pg
28,571.4 BANK for guest pass pool

P1 = 2 * Pg(N1/Ng)
P1 = 2 * 28,571.4(25/100)
14,285.7 BANK for L1 pool

P2 = 2 * P1(N2/N1)
P2 = 2 * 14,285.7(50/25)
57,142.9 BANK for L2 pool

And to double check, 28,571.4 + 14,285.7 + 57,142.9 = 100,000! :white_check_mark:

It will still be coordinape compensation mechanism but with a twist, like there are taks about not dumping the bank straight away, you get bank options that can be set as we desire. For example have them ‘unlocked’ (claimable) at x date if bank is at 0.1$ then your 1 bank kpi is worth x amount of bank. I’m sure this can be explain much better from somebody from UMA (possible we can set up a call with them?).

2 Likes