To be honest, I’ve stepped into that trap in chain games about output versus inflation a few times, and once I did, I learned my lesson. When project teams design an output curve, they often just draw a nice-looking pie chart—but in reality, user behavior is alive: miners dump, big holders smash the market, and once the exit speed is faster than the number of new entrants, the pool’s liquidity collapses fast.

Like the game I looked at before: the design for crafting material consumption was pretty reasonable, but the output-to-input ratio was a bit too high. The extra tokens minted every day couldn’t even be absorbed by the burn pool, and in the end the APY dropped from the several-hundreds range straight down to single digits. The people who do the calculations had already left, and it was only the retail investors who were left holding the bag. Later, I changed to focusing on on-chain transaction data—watching not only the coin price, but also the balance between the issuance speed and the destination of fee flows.

Recently, I’ve been looking at a few privacy-sector projects. The community debate has been intense: one side keeps shouting “anonymity is the natural law,” while the other side is worried about crossing compliance red lines. I’m interested, but for now I’m only observing. After all, on-chain arbitrage is about logic, not emotion. As for those “I only look at the price chart” paranoids—I used to believe them too. But then I found that on-chain data can explain everything. The pricing mistakes caused by greed, at their core, are often hidden inside emotions—so now I pay attention to a whole stack of user behavior, and I trust a perfect model far less.