Sichuan pepper plate learning with 8 characters<<Pinch the head and remove the tail, leave water in the middle> The definition of a good disk design must be elegant, harmonious from top to bottom, with controlled splitting speed and a reasonable "tail removal" mode First, let me explain what learning 8 characters is 'Pinch the head'=forced/high incentive locking 'Tail removal'=delayed gratification+penalty mechanism Intermediate flow "=arbitrary flywheel/investment nesting/compound interest reward ->Take the example of Arb DAO 🧵 Castle Capital's long article (to be posted in the comments section) has a critique of Arb DAO itself. Although I feel that the entire Arbitrum ecosystem is in a subtle awkward state - wanting and wanting, but in the end, they didn't ask for anything. Arb is already the most powerful ETH L2 player, but there are still structural issues From the perspective of dividend distribution, investing 153 million ARB can indeed boost TVL in the short term, but the problem is that this money is essentially just buying time. It's like giving adrenaline to an ICU patient, if the medication goes too cold, it still needs to be cold. The 300% TVL growth during STIP and the subsequent 80% capital outflow are the best proof. In theory, a multi round incentive plan (STIP → LTIPP → Backfund) should form a positive cycle, but in practice it completely turns into passing the buck. The new agreement came in for the purpose of generating subsidies, with no intention of long-term construction. This is like opening a restaurant and not providing meals to guests, relying solely on issuing coupons to maintain customer flow - the ones who come in the end are all those who specialize in eating coupons. A good economic model should be like a water pump, capable of extracting value from various links and feeding back the ecology. But ARB DAO is now completely doing charity work, and once the money is distributed, there will be no follow-up (Sui used to do the same before, spreading grants with big hands, but later found out that it was not right, they were all here to make a profit, so it is difficult to see projects that receive Sui grants now) I think the key is to solve three mismatches: 1. Mismatch of token allocation: DAO has too few chips in hand and cannot control the market 2. Mismatch of incentive cycles: Construction in bear markets, harvest in bull markets 3. Mismatch of value capture: Light subsidies do not extract water, purely for public welfare purposes And in my personal opinion, using VE as a model to control the outflow rate is not wrong, but it can be considered as controlling the splitting rate of semi-finished products, and this set has been validated by the market. The actual conversion rate is pitifully weak Self consistent income+controllable split rate structure+nested multiple platforms/products The Arb DAo example doesn't seem to have much to talk about, mainly because I've seen it. By the way, I also need to write something formal that can deconstruct the disk Finally, review a lesson on learning 8 characters and formulas Fund accumulation=forced/high incentive lock up (cutting the head)+delayed fulfillment/punishment (removing the tail)+flywheel&reinvestment/additional income accumulation funds (intermediate flow)