[Proposal] SIP-00011: Reducing the Active Validator Set from 80 to 21

Overview

SIP-00011 proposes reducing Story’s active validator set from 80 to 21. Story originally launched with 64 validators and expanded to 80 as part of the Polybius upgrade. With the network now established, maintaining 80 active validators incurs consensus and operational overhead that outweighs the marginal benefit of the additional validators.

At a coordinated hardfork height, only the top 21 validators by total stake will continue producing blocks and earning rewards. The remaining 59 validators will become inactive.

This post explains the rationale, what changes for validators and delegators, and how to prepare.

Why Reduce the Validator Set?

Consensus performance. CometBFT consensus messaging scales quadratically with validator count. With 80 validators, each consensus round requires up to 6,320 messages (80 x 79). With 21, this drops to 420 messages (21 x 20) — a ~93% reduction in consensus messaging overhead, resulting in faster block finality and lower bandwidth requirements.

Validator commitment. A larger set dilutes the economic incentive per validator, making it harder for operators to justify the infrastructure investment required for high-quality nodes. A smaller set concentrates rewards among committed, well-resourced operators who can deliver higher uptime and better performance.

Operational focus. Coordination, governance communication, and upgrade rollouts become more manageable with fewer active participants — especially important as Story moves toward more frequent protocol upgrades.

Economic sustainability. With staking rewards distributed across fewer validators, each validator earns a larger share, reducing the risk of validator churn due to insufficient returns.

Industry alignment. Many high-performance L1 networks operate successfully with similar or smaller validator sets. A set of 21 is well-established as a practical number that balances decentralization with performance (BNB Smart Chain, EOS/Antelope).

Transition Plan

To protect delegators and give exiting validators time to wind down operations, a structured exit process will be followed:

  • Locked validators must provide 3 weeks’ notice before exiting. Affected delegators will be notified so they can unstake during the notice period.
  • Unlocked validators must provide 1 week’s notice before exiting and must also publish a public redelegation notice for retail/public delegators.

In both cases, the validator should keep its node running until all delegations are removed. The Foundation may provide additional transition support to eligible exiting validators who complete the structured exit process and maintain node operations through the applicable notice period.

What the Validator-Set Reduction Means for Delegators

Nothing happens automatically. The protocol only changes which 21 validators sign blocks. Your delegation stays where it is until you move it.

  • If your validator stays in the top 21 → nothing changes for you.
  • If your validator falls outside the top 21 → your stake stays attached to that validator, but you stop earning rewards. You have to act.

Your Three Options if Your Validator Becomes Inactive

1. Do nothing

  • You earn zero rewards.
  • Your tokens remain on an inactive validator.
  • They are not automatically returned. The validator can in theory re-enter the top 21 later if its relative ranking improves, in which case rewards resume.

2. Undelegate

  • Starts a fresh 14-day unbonding period from the moment you undelegate.
  • After 14 days, tokens are credited to your withdrawal address on the EVM side.
  • During unbonding, the stake earns nothing and cannot be moved.

3. Redelegate to a top-21 validator (recommended for most delegators)

  • Your tokens immediately start earning rewards on the new validator.
  • There is a redelegation entry in the background that takes time to fully clear, but you keep earning the whole time.

If Your Stake is Locked (Period Delegation)

Some delegations have a lock period during which undelegate is blocked until the lock expires.

  • You cannot withdraw early. Even if your validator becomes inactive, you must wait for the lock to expire before you can undelegate.

  • You can still redelegate. Redelegation is allowed even during a lock, and the lock metadata moves with you to the new validator. This is the only way to escape an inactive validator before your lock ends.

  • If you stay put, you earn no rewards for the entire remaining lock period.

Summary Table

You are Your validator is What happens
Delegator Top-21 (active) No change. Keep earning rewards.
Delegator (flexible) Outside top 21 (inactive) You earn nothing until you act. Best move: redelegate to a top-21 validator.
Delegator (locked) Outside top 21 (inactive) Cannot undelegate until lock expires. Can redelegate to a top-21 validator and keep your lock.
Validator Top-21 (active) Continues normally. Higher relative voting weight than before.
Validator Outside top 21 (inactive) No rewards. Must manually undelegate self-stake via CLI to begin the 14-day unbonding period. Tokens are not returned automatically.

Slashing — Is Your Stake at Risk?

Slashing is when the protocol burns part of a validator’s (and its delegators’) stake as a penalty for misbehavior:

  • Double-signing (signing two conflicting blocks) — penalty of 5% of stake.

  • Downtime (validator offline too long) — penalty of 0.02% of stake.

Top-21 validators: Full slashing exposure, ongoing. Because there are now only 21 validators instead of 80, each validator has a larger share of the network. The protocol’s tolerance for any one validator going offline is lower, so liveness risk increases.

Inactive validators (outside top 21): For 14 days after the switch, the validator and all its delegators are still at risk of slashing — but only for misbehavior committed before it became inactive. The protocol has up to 48 hours to detect and punish such misbehavior. After the 14-day unbonding period, slashing is no longer possible.

If you undelegate or redelegate: Pulling out does not retroactively protect you from misbehavior the validator committed while you were still delegated. If your old validator double-signs and gets caught within the slashing window, your unbonding or redelegated tokens can still be slashed. After your undelegation or redelegation has fully matured (~14 days), the stake is permanently safe.

What This Means in Practice

If your validator is in the top 21: Do nothing.

If your validator is outside the top 21:

  • Flexible stake: Redelegate to a top-21 validator. You will keep earning rewards through the transition.

  • Locked stake: Also redelegate. It’s the only way to keep earning during your lock.

  • Doing nothing is the worst outcome for your returns — your stake earns nothing until you act, even though slashing risk fades after 14 days.

Quick FAQ

Q: Will my tokens be automatically returned to me?

No. The protocol only changes who signs blocks. Your delegation stays where it is until you move it.

Q: How fast can I get my tokens out?

Undelegating takes 14 days. Redelegating to a top-21 validator is effectively immediate from a rewards standpoint.

Q: Can my validator come back into the active set?

Yes. If a top-21 validator gets jailed or loses stake and your validator’s relative ranking improves enough, it can re-enter the active set. If that happens, your delegation resumes earning rewards automatically.

Q: Will I lose my tokens if my validator misbehaved before becoming inactive?

Possibly, but only if there’s a double-sign offense detected within the protocol’s evidence window (48 hours). The penalty is 5% of stake. After 14 days post-switch, this risk is over.

Q: Should I undelegate or redelegate?

For most people, redelegate. You skip the 14-day waiting period in practice (your new delegation earns from day one), and you don’t abandon your locked period if you have one.

Acknowledgement

The Foundation recognizes and appreciates the work of all validators who have supported the Story network to date. Many operators have contributed time, infrastructure, operational support, and ecosystem participation during an important phase of the network’s development. This change should not be read as a dismissal of those contributions. The transition plan is intended to help affected validators and delegators move through the change in an orderly way, with minimal disruption and no loss of delegated funds.

Links

Hey team, appreciate the transparency and detailed transition plan. We wanted to share some perspective as operators who’ve spent time across multiple ecosystems and care about where Story lands long-term.

We get the “why” just think 21 is too aggressive a jump.

The math makes sense. Going from 6320 messages down to 420 is a huge improvement, no argument. But a set of 50 gets you most of that benefit, down to 2450 messages, which is still a 61% reduction. It’s meaningful and it keeps a much broader validator set intact.

For what it’s worth, here’s what we’ve seen across other networks navigating similar decisions…

Osmosis has reduced gradually over time from 150 to 120 to 100, and currently sits at 90. Each step was measured before taking the next one.

Band Protocol landed at 53 validators when they launched v3, specifically citing the need for tight coordination with 1 second block times. Similar performance goals to Story and roughly similar solution, but to 53, not 21.

Sei is an interesting comparison with 400ms finality, super fast. Their active set is 39 validators and every one of those runs on a foundation delegation. Even with that level of central coordination and control, they still maintain nearly double what this proposal targets.

The SIP references Binance Smart Chain and EOS as examples of successful 21 validator networks. Its worth noting though that BSC actually expanded beyond 21. They added 24 candidate validators in the 2024 Feynman upgrade, bringing their total participating set to 45. Even Binance decided 21 wasn’t quite enough.

Nakamoto Coefficient…

With 21 validators and CometBFT’s 2/3 threshold, you need just 15 validators to finalize blocks. Depending on how stake concentrates, the number of independent entities needed to halt or censor the network could drop into single digits. That’s a meaningful shift in Story’s network security and the kind of thing that’s hard to walk back once institutions start evaluating the chain.

Brookings published a piece in April last year specifically about recentralization risk in blockchains. They called out that smaller validator sets create concentration risks and noted that Cosmos ecosystem chains have generally promoted decentralization by expanding validator participation, not contracting it. So it can be argued when mainstream policy institutions are watching this stuff, the optics of going to 21 matter beyond just the technical tradeoffs.

What we’d like to see…

We’re not saying don’t reduce the set, but that a gradual and measured reduction is worth considering.

Start at 45-55. Measure block finality, consensus times, bandwidth and the actual bottlenecks. If the numbers say you need to go smaller, go smaller with evidence behind it.

Consider a validator call before moving to on-chain vote. The operators running this infrastructure might reveal things that aren’t obvious from the protocol side.

Where we stand…

Full transparency, Atlas Staking has been selected for a foundation delegation and would hopefully land in the top 21. So this isn’t about protecting our seat.

We’re raising this because we think Story has a real shot at being a category-defining network and the validator set is part of that. We’ve built content, participated in governance, and invested in infrastructure here because we believe in what this team is building.

Happy to jump on a call, write up more detailed analysis, or help however is useful. We’re here for the long haul.

Matt & Eric

Atlas Staking

1 Like

Agree with the points raised above from @Atlas-Staking
This is a pretty big change , and would be great to get more clarity and discuss it properly on a validator call.