Kin Improvement Proposal: KRE 3.0.1

Written by: Will Gikandi

Chief Technology Officer at the Kin Foundation

April 14, 2021

KRE 3 has shown promising results in stimulating the economy and giving apps more freedom to innovate within the ecosystem. However, a number of apps have confided with us some difficulty in controlling/ maintaining an accurate Active User count (AU) due to bot behavior and other factors.

While apps have been making their best efforts to address this, we have also been working internally on some ways to reduce the impact of bots and churns on the KRE and to also make sure developers are rewarded fairly for the number of Active Users they bring into the economy.

Part of our efforts have included analyzing the data, which has led to the following proposal.

We believe that a minor change in how we define AU will help apps across the board, and also minimize the effect of bots, even as apps continue with their best efforts to reduce their activity. The proposal is currently open for comments before being incorporated to the KRE.


Kin Improvement Proposal

The purpose of this document is to outline a proposed improvement to the existing KRE algorithm.

versioning:

  • 3.0.1

Proposed by:

The Kin Foundation

Abstract

KRE 3.0.1 is an update which helps app developers manage Active User counts (AU) for their apps. The Active User count is used to determine a cap on the maximum Active User Balance (AUB) of kin an app can be rewarded for.

The current definition of the AUB cap for an app i is:

100000*MAU

where MAU = Monthly Active Users for app i. A user is an MAU in app i if they have spent Kin >= 1 time in app i in the last 30 days.

Generally, the MAU should be an analog of real users in an app, where the assumption is that one spending account correllates to an actual spending user in the app. This is expected to correlate with the number of users the app brings into the Kin Ecosystem, and more users leads to larger apps getting rewarded more for their contributions.

However, we have noticed that some apps may have trouble controlling this number, due to bots or user churn after a few spends that leads to abandoned accounts which inflate the MAU. For example, an app with only 5,000 downloads may show a MAU in excess of 30,000 due to abandoned accounts. When an app has a higher AUB and excess accounts, it may be rewarded disproportionately more than apps with actual users.

This leads to some developers getting less kin than they should due to this imbalance.

While many of the apps have been proactive in controlling this, a minor change in the definition of MAU will also help to address this universally.

Summary of Improvements

We propose to alter the definition of MAU from:

A user is an MAU in app i if they have spent Kin >= 1 time in app i in the last 30 days.

to:

A user is an MAU in app i if they have spent Kin >= 3 times in app i in the last 30 days.

The data we have shows that this change, while minor, will keep bot & churn accounts from from being counted as MAU. For the payout calculated on April 11 2021, the MAU and AUB cap would be as follows:

AppMAU = 1 spendAUB cap (100,000 * MAU)MAU >= 3 spendsAUB cap (100,000 * MAU)
QG3294894,800,00054154,100,000
pgbv82582,500,00043243,200,000
lsff12212,200,000989,800,000
l83h501,61050,161,000,000132,41613,241,600,000
lipz55,9815,598,100,0007,302730,200,000
p365512,23651,223,600,000180,50218,050,200,000
t1B578,9467,894,600,00072,2827,228,200,000
xnXb233,50723,350,700,00055,7275,572,700,000

This method is more fair and will ensure rewards count towards users who are engaged in the ecosystem, and will help minimize the effect of bots on rewards.

This will also only affect apps with both large kin deposits and churn or bot accounts. The effect will be to ensure only actual active users are counted towards the reward.

Implementation

We propose this is included in the KRE as soon as it is approved by the Kin Foundation so it can be part of KRE 3 by April 30 2021


We invite the community to review and provide input on this proposal before April 25th, 2021.

To review and comment via GitHub, you can find the full proposal here:
KRE 3.0.1 Proposal on GitHub

You can also discuss this proposal on Reddit here:
KRE 3.0.1 Discussion on Reddit