Points Systems: The Good, The Bad, and How To Improve Them

This is part two in our series on points systems. In this article, we will suggest how to improve points mechanics moving forward. If you want to better understand the reasoning, we recommend reading part one first, covering the basics, the pros, and the cons of points.

We believe the industry has to change its current approach to make points an effective attention tool, more sustainable in regards to token prices and user retention, and create satisfied users.

The first to create a new meta will capture most of the upside. They will bring back optimism, stand out from the other copycat points programs in the market, and capture the attention and imagination of users.

Below are some ideas worth exploring in our opinion.

Points allocation math isn’t that difficult. It really isn’t. However, once boosts and other extras get added on top, the complexity of the system can get very explosive very quick. This is in part what happened with the mini-points valuation debacle this past summer when LRTs were promising boosts left and right, especially in OTC deals.

This has very rapidly diluted previously existing positions. Combined with the opacity around these deals, it was very difficult for depositors to reverse calculate what their “fair” points-to-token conversion should be. Enter deterministic, or linear points systems!

One protocol that built their points architecture based on a predictable and transparent infrastructure called Point Guard. We weren’t just virtue signaling about transparency earlier.

TAU Labs is working with YieldNest as the in-house Tokenomics, Pointomics, Research and Data arm. While we back what we say with facts, math and in-depth research, please keep a healthy level of skepticism, as this is crypto after all.

Ok, now back to Point Guard. It is a protocol points system that is validated by an EigenLayer-powered AVS. The AVS effectively scans ERC20 or ERC721 contracts and calculates reward points for users. These contracts interface with EigenLayer to ensure that the points awarded to users have been deterministically calculated.

We have worked together with YieldNest to make the documentation around Seeds as simple to understand as possible, while being as detailed as we can. Any user should, therefore, be able to reverse-calculate their own number of Seeds, simply by knowing their applicable boosts and deposits. From the docs: 

“In order to determine our user’s Seed allocation (Di), we take the base rate (R) of two Seeds, multiplied by the amount of ERC20 token (Ei) deposited and again multiplied by the amount of hours participated (hi). To this base deposit, the user may have various boosts applied (Bi), for a certain amount of hours. The boosts apply only to the base rate and therefore stack but do not compound. The user also is able to receive their referral bonus (f), calculated as being 10% of the referred wallet’s balance.”

Knowing how the calculation applies to one user, we can then deterministically reverse-calculate the total number of Seeds (Ptotal) across all deposits and staked positions, provided all boosts are known.

While not without its limitations, we believe points systems should be as transparent and open as possible not just in public posts on X but also through the published documentation and business logic that they use.

As alluded to earlier, teams know about the issues of the current points meta which corners them into a defensive situation where they feel like they have to withhold information from the public to minimize getting farmed by mercenary capital.

Our vision for points programs is to find a middle ground between flexibility (for teams) and transparency (for users) - to swing the pendulum back towards a more user-centric model.

So far, the reaction to mercenary users has been to put a supposed band-aid on the symptoms rather than building a better system from first principles.  We argue that we can give users more transparency while combating extractive behavior through mechanisms, such as multi-point systems and in-app rewards.

Multi-point systems

Rather than awarding the same points “currency” to all rewarded actions, we can utilize different types of points for distinct activities. For example, award gems for in-app usage and shards for social engagement. That allows us to separate reward systems from one another which prevents negative spillover effects and provides more targeted incentivization.

Inspiration can come from games which need to manage complex economies. They do so by isolating different economic pillars. That way, social sybil attackers only dilute each other, but not in-app users.

Source: https://medium.com/@0xKepler/a-road-forward-for-web3-gaming-e70d82dd19e3 

Similarly, web3 games learned that it’s better to have a separate utility token and governance token as they serve different purposes which are almost impossible to bring together. For instance, governance tokens should be fixed supply to make them more investible while in-game tokens have to respond more dynamically to supply and demand.

In-app rewards

Do not award points that translate into tokens directly for all user actions. Some lower-value interactions can be awarded offchain credits instead which users have to use in-app to unlock rewards.

In conjunction with the above-mentioned multi-point system, this can make for a powerful combo that boosts user retention and decreases ineffective token dilution.

As mentioned earlier, protocols have to consider the individual user holistically for their points systems. If we just take single actions and reward them without this holistic view, we fail to capture their interconnectedness and what information they give us about a user.

Step 1: Model the conversion funnel

• Decide on your north star metric (e.g. revenue, retention rate)

• Sketch the user journey and identify what you want users to do to drive the north star metric

• Assign a value to each user step. In the best case, you have hard data on that. Otherwise, make an educated guess.

For example, if data shows that a user retaining for 7 days has a 30% higher chance to spend money, you can quantify that. That can inform how many points to award for daily logins. But be aware that incentives change user behavior and therefore the calculations

Now, we have quantified the value of user actions based on our business objectives and can award points accordingly.

Step 2: Create a holistic progression system

Connect user actions together to account for the complexity of user behavior. For example, unlock referrals only after users show a minimum threshold of in-app activity. Or give multipliers on in-app token earnings for social engagement.

In the multi-point system we discussed earlier, we begin by separating various actions into distinct categories. Each category is treated differently with its own points “currency” to minimize the risk of negative spillover effects between categories. Later, we strategically combine some of these categories (e.g. via multipliers) to encourage the desired behaviors or incentives. This method allows for more targeted and effective rewards within the system..

We’re still optimistic on points. They solve the time gap dilemma between when token rewards are needed and when tokens become valuable. They give protocols time to develop network effects which then can be captured via a token. Early users are then rewarded for their contributions through an airdrop based on their accumulated points.

To make this a reality, we have to overcome the stigma of current points designs that lack transparency. As with other longer-term developments in crypto, there are solutions to these issues but they will need a certain level of pushback from users to occur. Fortunately, we’re already seeing that!

Future point metas will need to be far more transparent about how points are distributed and converted into tokens in the future. They will also need to better account for the complexity of user behavior and risks of exploitation. This will make points systems more resilient, lead to a more effective allocation of rewards, and result in better retention.