1/ Based Token Farming: While everyone else was fixated on trading, the real "honey pot" was in the Prediction Market. To start with the conclusion, the cost of trading was around 100M at best, but I was able to lower it to the 30M level in the Prediction Market. How was this possible? I am sharing the strategy. 2/ The Key is the Correlation Between 'Builder Fee' and Probability In the Prediction Market, 2M Gold is distributed as a fixed amount per Epoch. Calculations showed that the structure was such that the builder fee increased as the probability approached 0%. In other words, this meant you could efficiently burn through fees (costs) while reducing the number of trades. 3/ Setting Up a Perfect Wash Trading Environment The key was selecting tokens with thin order books and 0% probability. Since mixing in other people's money complicates wash trading, I selected only tokens with no volume and boosted the trading volume using multiple accounts on PolyMarket. At the time, the Prediction Market achieved 10k daily revenue, and 8k of that was my share. 4/ 'The Aesthetics of Moderation' - Since the pool is fixed at $5,000 per account per epoch, mindlessly burning money actually results in a loss. That is why I approached it with multiple accounts. After analyzing the data, I found that investing $5,000 per account was the peak of gold mining efficiency. As expected, farming is not about luck, but a battle of thorough calculation. So, when I had farmed quite a bit using the prediction market, I really saw the point where I should graduate. Because the profit was significant even if the base reached just 200 million. But I am sad because I am left with nothing but regret.