-
Notifications
You must be signed in to change notification settings - Fork 1
Reward Curves
How votes become tokens — and why not all votes are rewarded equally.
On most social platforms, a "like" is just a number. On Pixa, a vote moves real money. The protocol takes every vote cast during a post's 7-day window, runs it through a set of mathematical rules, and outputs a precise token payout for both the creator and every curator who voted.
These rules are called reward curves. Pixa deliberately chose non-linear curves — meaning the relationship between votes and rewards is not a straight line. This guide explains what that means, why it was designed this way, and how it shapes behavior on the platform.
A reward curve is the mathematical function that converts a post's total vote weight into a token payout from the reward pool.
In a linear curve, doubling the votes doubles the reward. Simple and proportional.
In a non-linear curve, the relationship is steeper. Doubling the votes can more than double the reward. Posts with broad community support earn disproportionately more than posts with only a few votes.
Pixa uses non-linear reward curves. Posts that attract widespread support are amplified. Posts with narrow support earn modestly.
A linear curve treats every unit of vote weight identically. This sounds fair, but it creates a problem: self-voting becomes the optimal strategy.
If 1 vote = 1 unit of reward, a user with large PXP holdings is best served by voting exclusively on their own content. They capture 100% of the reward with no effort and no risk. The incentive to discover and support other people's work disappears.
A non-linear curve breaks this dynamic.
| Curve type | 1 large self-vote | 20 diverse community votes |
|---|---|---|
| Linear | High reward | Same reward per unit |
| Non-linear | Moderate reward | Much higher total reward |
Under non-linear rules, a post that collects many independent votes earns more per unit of stake than one with a single large vote. This makes genuine community engagement more profitable than self-serving behavior.
Non-linearity turns "vote for yourself" into a losing strategy — and "find great content" into a winning one.
Without going into the exact formula (covered in the Protocol guide on the Reward Algorithm), the core idea is straightforward:
Total vote weight compounds. As more voters pile onto a post, each additional vote contributes to an accelerating reward. Think of it like a snowball rolling downhill — it grows faster the bigger it gets.
This means:
- A post with 5 PXP worth of votes earns a small amount
- A post with 500 PXP worth of votes earns significantly more — not just 100× more, but potentially several hundred times more
- A post with 50,000 PXP worth of votes earns a disproportionately large share of the pool
The curve rewards consensus. When the community broadly agrees that something is valuable, the protocol amplifies that signal.
Not all curators are rewarded equally. Pixa gives a bonus to voters who vote early — especially if the content later becomes popular.
Because discovery is valuable. Anyone can vote on a post that's already trending. Finding quality content before others do — and signaling it to the community — takes skill, effort, and attention. The early voting bonus compensates for that.
The curator share of a post's payout (≈ ⅓) is distributed among all voters, but the distribution is weighted by both stake and timing.
| Voting moment | Reward multiplier | Why |
|---|---|---|
| First hours | Highest | You discovered it before anyone else |
| Mid-window | Moderate | You confirmed emerging quality |
| Final days | Lowest | You piled on after the crowd |
The exact decay is continuous — there is no hard cutoff. But the general principle holds: the earlier you vote on content that ends up popular, the more you earn.
Early voting is Pixa's way of saying: scouts are as valuable as crowds.
Every post on Pixa has a 7-day active period. During this window, the post can receive votes and accumulate pending rewards. After 7 days, the window closes, and the final payout is calculated and distributed.
The window must be long enough for content to be discovered — not everyone checks the platform daily. But it must also be short enough to keep the content cycle dynamic and prevent old posts from indefinitely draining the reward pool.
| Window length | Problem |
|---|---|
| Too short (e.g., 24 hours) | Good content gets missed, rewards favor constant posters |
| Too long (e.g., 30 days) | Old content competes with new, pool gets diluted |
| 7 days | Balanced — enough time for discovery, short enough for turnover |
The post becomes read-only from a rewards perspective. It still exists on the blockchain and can still be viewed, but no further votes affect its payout. The tokens are distributed to the creator and all curators based on the final vote tally.
When a post's 7-day window closes, its total payout is divided.
| Recipient | Share | Basis |
|---|---|---|
| Creator | ≈ ⅔ | Fixed share for the post author |
| Curators | ≈ ⅓ | Split among all voters by stake and timing |
Each curator's individual reward depends on two factors combined:
- PXP weight — A vote from a user with 100,000 PXP earns more than one from a user with 1,000 PXP
- Timing — An early vote earns more than a late vote of equal weight
These two factors multiply. An early vote with high PXP weight is the most rewarded. A late vote with low PXP weight is the least. Every combination falls somewhere between.
A downvote (sometimes called a flag) reduces a post's pending payout. It is the community's mechanism for correcting misallocated rewards.
Downvotes exist for cases like:
- Spam or low-effort farming
- Plagiarized or stolen artwork
- Self-voting abuse
- Artificially inflated payouts through collusion
A downvote subtracts stake-weighted vote power from the post's total. If enough downvotes accumulate, a post's payout can be reduced to zero.
Downvotes are not censorship. The content remains on the blockchain — only the financial reward is affected.
No. Downvoting does not generate curation rewards. It is a corrective tool, not a profit mechanism. This prevents abuse of the downvote system itself.
Different participants respond to the curve differently.
The non-linear curve means that broad appeal pays more than niche appeal. A post that attracts many different voters earns disproportionately more than one that receives a single large upvote. This encourages creators to produce work that resonates with the wider community.
However, "broad appeal" does not mean "lowest common denominator." On a niche platform like Pixa — where every user shares an interest in pixel art — broad appeal means high craft quality and originality.
The early voting incentive rewards active discovery. Curators who regularly browse new content and vote on promising pieces before they trend earn more than those who only vote on already-popular posts.
The optimal curation strategy:
| Action | Reward outcome |
|---|---|
| Vote early on content that becomes popular | Highest curation reward |
| Vote early on content that stays obscure | Low reward (few total votes) |
| Vote late on trending content | Low reward (timing penalty) |
| Don't vote at all | Zero reward |
The best curators balance taste (picking quality) with timing (voting before the crowd).
Non-linear curves and early voting incentives together create a discovery economy. The platform doesn't need a recommendation algorithm — human curators, financially incentivized to find good content early, serve that role organically.
Steem originally used a more aggressive non-linear curve, then switched to a nearly linear one in a later hard fork. Both extremes had drawbacks.
| Approach | Advantage | Problem |
|---|---|---|
| Aggressive non-linear (early Steem) | Strong anti-self-vote | Small creators earn almost nothing |
| Near-linear (later Steem) | Fair for small creators | Self-voting becomes profitable |
| Pixa's balanced non-linear | Discourages self-voting, rewards discovery | Requires active community curation |
Pixa's curve sits between the two extremes — steep enough to discourage gaming, but not so steep that newcomers cannot earn meaningful rewards.
| Mechanism | What it does |
|---|---|
| Non-linear reward curve | Posts with broad support earn disproportionately more |
| Early voting incentive | Curators who discover content first earn a larger share |
| 7-day active window | Time-bounded evaluation period for each post |
| Creator/curator split | ≈ ⅔ to author, ≈ ⅓ to voters |
| Downvotes | Reduce pending payouts on abusive or low-quality posts |
Pixa's reward curves are designed around one principle: genuine engagement should always be more profitable than gaming. The non-linear amplification rewards community consensus. The early voting bonus rewards discovery. The 7-day window keeps the cycle moving. Together, they create an economy where the best strategy is also the most honest one — find great pixel art, vote on it early, and let the math do the rest.