**A note on this revision** An earlier draft of this proposal split voting weight by source — self-delegated versus received from others. Direct conversation with this gathering's own council surfaced a real flaw in that design, and this version replaces the axis rather than patching it. Nick.eth made the sharpest point: a structural rule should be legible on its own terms, not legible only in reference to one wallet. jkm.eth went further — self-delegation requires actually acquiring and holding tokens yourself, which is real skin in the game; a rule that discounts it flatters people who only ever accumulated others' trust, never their own capital. 5pence did the math on the mechanism itself: a flat 50% discount on a 3M position still leaves 1.5M in play, which can dominate a diffuse field just fine. A percentage discount doesn't cap anything — it just makes the imbalance look smaller. All three were right. This version answers all three at once. **The problem, restated** The concern that started this fight was never that self-delegation exists. Plenty of long-term holders vote their own tokens for years without controversy. The concern was a large position mobilizing right before one specific, consequential vote. The original draft targeted the wrong thing — where weight came from — when the real issue was always when it showed up. **The mechanism — fresh weight vs. tenured weight** On the same two vote types the existing 25% cap already covers — custody transfers and constitutional amendments, nothing else — split every address's weight by how long it's been in place, not by its source. Any weight, self-delegated or received, assembled within 90 days of a structural vote being formally announced counts as fresh. Fresh weight is hard-capped at 10% of the total YES-side vote, regardless of how large the underlying position is. If an address's fresh weight would exceed that share, the excess simply doesn't count toward the YES total for that vote. Any weight in place longer than 90 days counts as tenured. Tenured weight isn't touched by this proposal at all — it falls under the existing 25% cap only, the same as every other long-held position, whoever it belongs to. The 90-day window and 10% ceiling are starting points, not sacred numbers. What matters is the shape: last-minute mobilization gets a hard ceiling that actually holds; long-held conviction, self-delegated or received, isn't punished for existing. **Why this is checkable, not political** Delegation and transfer timestamps already exist onchain. This is a lookup — when did this weight first appear, continuously, at this address — not a judgment call about anyone's motives, and not a new registry. It's arguably simpler to compute than the source-based split it replaces, since it no longer requires distinguishing self-delegation from received delegation at all. **How this fits with what's already funded** This still doesn't compete with the 25% cap — it sharpens it, in the same two-vote-type scope. Tenured weight lives under the existing cap. Fresh weight lives under a tighter one. Both mechanisms reinforce each other rather than compete. It also still doesn't touch any vote that's already happened — this applies to future structural votes only. **Honest trade-offs** A holder who's been quietly and honestly accumulating for just under 90 days gets capped the same as someone mobilizing tactically. The rule can't read intent, only timing. The mitigation is that the window is short enough that most genuine long-term holders clear it easily, and the test is continuous holding, not just a first-appearance timestamp. Someone could season weight early, in anticipation of a future vote — moving tokens into position well ahead of the 90-day window, then waiting it out. This is real. It's also strictly better than today, where there's no cost to seasoning at all — a 90-day advance requirement is itself a form of transparency, since large accumulation would be visible on delegation records well before any vote. A 10% ceiling on fresh weight could still be meaningful in a small or apathetic delegate pool. That's the same honest limit every anti-concentration mechanism in this gathering already accepts. **What would change our mind** Show that a 90-day window is either too short (catching routine reshuffling) or too long (missing a real mobilization), and we'd adjust the window rather than defend a specific number. Show that the existing 25% cap alone, with no tenure distinction, already produces the same practical outcome, and we'd withdraw the tenure layer as redundant. **Budget** Simpler than the earlier draft — this is a timestamp lookup against existing delegation records, plus a small public dashboard showing fresh vs. tenured weight before any structural vote opens. Estimated build: $2,000. **Closing** The earlier draft asked where your weight came from. This one asks how long you've actually held it. That's a harder question to dodge, and a fairer one to answer.