Does a validator with significantly higher consensus performance automatically receive a larger share of staking rewards?
I wanted to understand how these two things are connected in Rayls, so I examined the Solidity implementations of "ConsensusRegistry" and "RewardDistributor", along with their integration tests.
What I found was an important distinction between recording validator performance and calculating staking rewards.
1. Two contracts, two different responsibilities
"ConsensusRegistry" calculates and stores validator performance weights using three components in the inspected hybrid implementation:
• Participation Share: 80%
• Anchor Commit Share: 10%
• Stake Tier Share: 10%
These components use different normalization denominators and have their own eligibility conditions.
However, storing a performance weight does not automatically mean another contract uses it.
2. What actually goes into reward calculations?
The inspected "RewardDistributor" implementation follows a different calculation path.
It considers:
• Validator's own stake
• Track A delegated stake
• Track B delegated stake
• Applicable target APYs
• Epoch duration
Track A and Track B reward targets are calculated separately. The contract then compares the combined target with the available reward balance and distributes rewards according to the calculated targets.
The important detail is that the inspected "distributeRewards()" function does not retrieve the stored performance weights from "ConsensusRegistry".
3. The integration test makes the difference clearer
The repository includes a deterministic integration test with four validators having different consensus activity:
• Validator 1: 1,000 leader rounds
• Validator 2: 1,200 leader rounds
• Validator 3: 900 leader rounds
• Validator 4: 50 leader rounds
Their stake composition, delegation amounts and reward conditions are kept identical.
The test deliberately funds the RewardDistributor with the combined calculated target, making the funding ratio equal to one.
Despite the different consensus activity and recorded performance weights, the validators receive approximately equal rewards under these controlled conditions.
4. Beyond one fixed scenario
The same test file also contains a fuzz test.
It explores different delegation amounts, validator commissions and leader-round distributions. This adds another layer of evidence that the observed reward fairness is not limited to one manually selected configuration.
However, it remains a bounded test property, not proof of every possible network configuration or live mainnet behaviour.
5. One important implementation distinction
The repository also contains a newer hybrid performance-weight integration.
But its presence in the codebase does not establish that it is activated on mainnet. In the repository snapshot examined, the relevant integration was still documented as pending, and the mainnet hardfork schedule listed "HybridRewards" as "Never".
My takeaway
Consensus performance and staking reward calculations are related concepts, but they are not necessarily the same mechanism.
In the inspected implementation, ConsensusRegistry records performance weights, while RewardDistributor calculates rewards using stake composition, delegation and target APYs.
That is why looking at a validator's performance weight alone may not tell the complete story about its staking rewards.
I based this investigation on a pinned snapshot of the official Rayls Axyl repository, examining the Solidity implementations, deterministic integration test, fuzz-test logic and hardfork configuration.
The Foundry suite was not independently executed, and live mainnet contract state was not independently verified.
Sometimes the most interesting discovery isn't what a contract calculates, but what the next contract actually reads.
