JIP-25 - Expand the Validator Set and Modify Jito Stake Pool Eligibility and Ranking Criteria

[brian_smith]

Category: Protocol

Abstract

JIP-25 proposes a set of immediate actionable changes to StakeNet that are designed to mitigate emergent negative Stake Pool dynamics. Over time StakeNet has created a cryptoeconomic context that has converged towards biasing validators that utilise PvP tactics and engage in sybil activity. This is highly undesirable and following extensive validator feedback it is time for Jito DAO to act.

This proposal seeks to expand the validator set to 400, modify StakeNet eligibility and scoring systems and pause implementation of the priority fee requirement. It moves away from vote credits as a ranking mechanism towards a new ranking system that ranks based on commission, MEV commission, reverse validator age, and reverse vote credits to order validators in the pool. The goal is to move the Jito Stake Pool into a more productive economic equilibrium that biases high quality operators and lays the groundwork for a future JIP that introduces directed staking (link) and DeFi app directed staking.

Motivation

The current StakeNet system parameters were effective for most of its lifetime to date, creating a system for ordering validators by performance. However, validator vote credit performance for Solana has improved drastically since introduction of this mechanism and differentials are immaterial. The emphasis on vote credits as a mechanistic differentiator within the pool has led to a significant bias towards validators who hyperoptimize for this parameter. JitoSOL should reward validators pushing Solana towards its IBRL goals, not targeting inconsequential vote credit advantages.

Furthermore, permissionless validator delegation should not exceed standalone profitability by a wide margin. Otherwise sybil behavior is encouraged and anecdotal evidence suggests sybils are actively targeting JitoSOL delegations. JIP-25 would bring JitoSOL delegation below standalone breakeven given Solana’s current market environment. A broader validator set rewards more good Solana validators while discouraging sybil validators that do not contribute to Solana’s health.

JIP-25 proposes to modify StakeNet’s eligibility settings and scoring system to emphasise commission, MEV commission and validator age alongside a broadening of the stake pool validator set. A simple, transparent ranking of MEV commission (from lowest to highest), with validator age (from highest to lowest) as a tie breaking mechanism, will create a context that boosts JitoSOL yield, is harder to game and aligns incentives towards honest validators.

Finally, JIP-25 includes a further response to validator feedback and pauses the deployment of the priority fee sharing mechanism proposed in JIP-16. Engineering work has been completed for this action, but consensus has since drifted as the build was in progress due to changes in the broader Solana economy. At the time of the JIP-16 vote, priority fees were far higher than the current context (Q4 2024 and Q1 2025 priority fees were ~$350mm each whereas Q2 and projected Q3 priority fees are ~$96mm each Source: Blockworks). This decline of 72% in priority fees has a detrimental impact on validator earnings while reducing the benefit to JitoSOL holders of any sharing. The infrastructure is now ready and the DAO can propose implementation at a later date when the community decides its appropriate.

This proposal should be seen as part of a broader, phased evolution of Jito’s stake delegation governance. JIP-25 introduces incremental but important adjustments designed to reinforce Jito’s role as core Solana infrastructure. By aligning delegation incentives with positive-sum economic behavior and discouraging extractive practices, these changes strengthen validator quality, improve network health, and set the foundation for future upgrades.

Key Terms

• StakeNet Eligibility: The established StakeNet eligibility criteria remain.

• Delegation Set: The set of eligible validators that receive stake (up to 400).

• Rank: Ordering used by StakeNet when sequencing deposits/unstakes.

• Operator: Unique economic identity controlling one or more validators; used to enforce anti-sybil rules.

Specification

1) Delegation Set Size

• Increase the delegate set from 200 active validators, to 400.

• If more than 400 validators are eligible, the top 400 by Rank (defined below) compose the Delegation Set.

• All other StakeNet parameters (targets/churn) remain unchanged unless explicitly noted.

2) Modify Ranking Function

Replace the existing vote credit rank/score with:

1. Primary Ranking Score: Commission Rate (ascending; lower is better)

2. Secondary Ranking Score: Jito MEV commission (ascending; lower is better).

3. Third Ranking Score: Validator age (descending; older is better), computed as the number of epochs since the validator first became active.

4. Tiebreaker: Vote Credits act as final tiebreaker, to respond to the case where validator age extends beyond the lifetime of StakeNet (descending; higher is better)

Illustrative Example

• Validator Ranked 1st: 0% inflation | 3% Jito MEV | 36 mo history

• Validator Ranked 2nd:0% inflation | 3% Jito MEV | 5 mo history

• Validator Ranked 3rd: 1% inflation | 0% Jito MEV | 38 mo history

• Validator Ranked 4th: 2% inflation | 2% Jito MEV | 40 mo history | Vote Credits ++

• Validator Ranked 5th: 2% inflation | 2% Jito MEV | 40 mo history | Vote Credits

Interpretation: All five validators are eligible; ordering follows commission first, then MEV commission, then age, with vote credits as a final tie breaker.

3) Adjust MEV Commission Reference Period

Lengthen the MEV commission reference period from 10 epochs to 30 epochs. It specifically measures the highest MEV commission during that window. This is consistent with other reference periods and reduces possible manipulation associated with validators temporarily adjusting MEV commissions to earn stake more quickly due to the revised ordering criteria.

4) Pause Priority Fee Sharing

This delays the implementation of priority fee sharing as passed in JIP-16. No changes to existing parameters.

5) Retire Vote Credits as Ranking and Raise Threshold

• The continuous vote credit ranking system is relegated to act as the final tiebreaker.

• A higher threshold for vote credit inclusion is introduced of 97%

Benefits / Risks

Benefits	Risks / Mitigations
Reduces PvP incentives: Binary gates remove micro-score arms races over vote credits	Novel Stake Pool dynamics: The new structure will introduce a new cryptoeconomic structure for the pool and monitoring will be required to ensure that the new set point is optimal
Broader set of 400: dilutes incentives to sybil and broadens accessibility to the Stake Pool	Short-term churn/yield drag: The proposal will take time to implement as churn caps create limits on implementation times.
Simple, intuitive ordering: low fees first; older first	Older Validator Advantage: The new scoring system introduces validator age for the first time. This disadvantages newer validators. Mitigated by using validator age as a tie breaker rather than primary ranking key.
Immediate implementation: No complex dynamics and new abuse surfaces.

Outcomes

Short Term

• At execution changes to StakeNet are immediate with new ranking criteria implemented.

Medium Term

• The validator set expands and stake is progressively rebalanced as churn occurs.

• Directed stake JIP is deliberated and planned.

Long Term

• A comprehensive new Stake Pool governance structure including an expanded validator set alongside a directed stake system that improves Solana ecosystem health and network performance.

[Simple]

I’m concerned that it may work against new validators. For example, how about including the onboarding eligibility for SFDP as a higher-priority evaluation criterion than validator age, so that validators with greater contributions to the ecosystem can be better recognized?

[huisky]

so i) and ii) basically will drive everyone to 0. Then it just come down to validator age basically?
sounds like new validator will never have a chance, unless old validator completely quit.

I think older validator could have few extra points, but new validators should also have chance to compete by having a better performance.

[Paws]

Once a validator has entered the pool, even if its performance is poor, it will not be removed except when an older validator lowers its commission rate and displaces it. As a result, the set of validators in the pool becomes fixed rather than shuffled, which I consider to be a problematic aspect.

[engineerhead]

This approach clearly favors long-standing validators. Some points should be given as suggested by @huisky
To address another issue, some weight could be given to geographic location, as previously suggested Geo-Distributed Jito Stake Delegation Process - #3 by drnick .
By combining these two ideas—adjusting for validator age points and incorporating location-based considerations—we can improve stake distribution and create more opportunities for new validators to participate

[meyerbro]

I believe this is the first time a pool is truly prioritizing long-term validators, and I want to commend you for that. Expanding from 200 to 400 validators shows that this isn’t about closing opportunities, but about rewarding real operators who have proven their commitment. The reality is that there isn’t enough stake to sustain such a large number of validators, and even Alpenglow has said that Solana should ideally operate with far fewer nodes. Remaining competitive is already extremely challenging, and long-standing validators bring genuine operational value, unlike the many white-label nodes simply running the same setup to capture Jito stake.

That said, I would personally prefer if validator history came before MEV commission in the priority order. This would emphasize the true benefit of sustained operation and help prevent a race to 0/0 commission that could enable a sybil factory of 400 validators running under the same incentives.

[Pumpkin]

Using validator age is a great way to shield against sybilling and ensure we’re delegating to experienced operators.

Fantastic idea. We might actually be able to get Jito stake again without having to run a mod and move to Amsterdam.

[Ian]

I am in favor of this proposal and will vote “Yes.”

The benefits of JIP-25 outweigh the drawbacks imo, the main drawback being its advantage for older validators, a concern others have rightly noted. In the short-to-medium term, this will likely drive many validators to lower inflation and MEV commission to zero, ultimately benefiting JitoSOL users and holders, and potentially increases geographic distribution if relevance of TVC decreases. While validator age will be the hardest metric to achieve, it both rewards longevity and incentivizes proper operator behavior over time.

This shift should reduce the MEV blacklisting burden on Jito, the Solana Foundation, and other contributors, while also providing clearer visibility into validator history, whether that involves sandwiching, timing games, or other malicious/anti-IBRL behaviors. Tying incentives to validator history also makes it harder for bad actors to repeatedly re-enter the set under new addresses. Lastly, a complete validator history could support future proposals such as JIP-XX on direct staking. I have no strong opinion on TVC serving as a tiebreaker for validators of equal age.

[kuma]

It is excellent that Jito has changed its policy of determining delegates based solely on TVC values in order to achieve a 0.1% increase in APY. This change significantly alters the current situation, where delegates could only be obtained by using mods that did not necessarily have a positive impact on the network. Additionally, those who have been committed to Solana validators for many years are the individuals who supported Solana during its downturn, and I believe it is appropriate to have stake pools that reward such individuals.

[ily]

I support the proposal to shift Jito’s delegation program from prioritizing TVC to prioritizing the number of epochs a validator has been actively running. I believe it makes more sense to reward validators who have supported Solana for many years than to encourage competition to modify MODs solely to improve TVC rankings.

Additionally, I believe that taking geographic diversification into account—particularly across regions such as South America, Africa, and Asia—would further strengthen this delegation strategy.

[mike]

Will validators’ “seniority” reset if they’ve gone delinquent for more than an epoch? I feel like they should.

[vika]

I do not support the proposal.
If you would like to enforce anti-sybil, then increase minimum validator age requirement from 30 epoch to 1 year (better two years). Unlikely anyone would run a sybil for 1 year with top performance to get stake afterwards.