Why is the system bullish while the market continues to decline? Recently, someone asked: CoinKarma's system has been oversold, but the market is still continuing to decline. Is it uncertain? There is a common misconception hidden behind this question. The trading system calculates the expected value, not the predicted single result Many people think that the trading system should be like fortune telling: if you say it's up, it's up; if you say it's down, it's down. But a good trading system doesn't operate like this. What it tells you is that based on past data patterns, entering this position is cost-effective in the long run. Will it earn money in a single transaction? I don't know But after doing it a hundred times, the overall result is positive. This matter can be understood using Blackjack: You got 20 points, but the banker happened to get Blackjack, and you lost this bet. But the next time you get above 19 points, you should still place a bet because the chances of your opponent having a larger card face than you are already very low. You're not betting on a single outcome, you're betting that this decision is right in the long run. Transactions are the same. Taking our quantification as an example, the previous multi position entry position was at 90K, and later dropped to around 85K. After 57 rounds of adjustment, it finally exited at 96K, with a profit of over 8%. If you pursue 'absolute precision', you can adjust the indicators more strictly, and they will basically become more accurate. But the problem is that if the tuning is too strict, you may not have many trading opportunities in a year. Adjusting too loosely may lead to premature entry. There must be a trade-off in between. Although I started to be bullish at 90K this time, due to the weak market, I lowered the minimum long position opening point to 85K. I don't expect the market to rebound immediately, but based on the retrospective analysis, the expectation of building long orders here is relatively high, but at the same time, some space needs to be reserved to avoid entering the market too early. The same logic applies to appearances. In the previous wave, we sold all our positions at 96K, but later it rose to 98K, an additional increase. The reason why I didn't plan to do that part is because the program calculated that the probability of a subsequent increase is low, so I can go if I need to. Leave a buffer when entering and also leave a buffer when exiting. Both sides are expected value judgments, not guessing the head and bottom. Having expectations alone is not enough Here is a key concept to be discussed: A good trading system helps you calculate the 'expected value', but good position management helps you reduce the 'variance of the net asset value curve'. Expectations tell you whether this position should be entered or not. But a single transaction can either win or lose, which can cause your net worth curve to fluctuate dramatically up and down. The role of position management is to make this curve smoother. The so-called position management is to give up the opportunity to make a quick profit once you see it right, in exchange for a better long-term holding experience and the chance to survive in the long term. Ultron's strategy is basically based on the Market Pulse Index to rebase positions. Sometimes the entry is very accurate, and the order is immediately dropped out; Sometimes I take a little longer to place an order. But in the end, almost all of them were making money. We have now placed all quantified transaction records on Hyperliquid, which is completely open and transparent. If you have studied our past trading records, you will find that although our winning rate is high and the curve is beautiful, we have not actually achieved perfect bottom fishing and top flight. Because it was never possible to do this in the first place. Ultron uses programs to calculate expected values and also uses programs to manage positions, making the net asset value curve as smooth as possible and reducing net asset value variance. Over 600 days of actual trading performance, both long and short positions have a winning rate of over 80%, outperforming BTC performance by 5 times and achieving a pullback lower than spot trading. That's why we need to do automated trading Because it is difficult for humans to do both "calculating expected values" and "position management" at the same time. People are easily prone to getting drunk. When you see the signal, you want to All in and open the lever 20-30 times in one go. Or if you carry the order halfway, your mentality will collapse, and you will cut the position at the worst. Our trading system can achieve a winning rate of around 80% in a wide range of volatile market conditions. But if you don't manage your position well, the remaining market conditions that will make the trading system suffer may cause you to lose everything you earned before in one wave. But Ultron's past historical net worth records have proven that headwinds can be managed through good position management, and having headwinds does not hinder controlling drawdown and making long-term profits. Position management itself is against human nature and really requires a bit of talent. Countless smart traders ultimately fail in position management, which is not related to effort, but rather to personality. It is difficult to learn solely through effort. So the focus of CoinKarma's development after that is to use automated trading tools to complement this system. The tool will automatically help you adjust your position, avoiding human issues. You are responsible for understanding logic, while machines are responsible for enforcing discipline. So back to the original question: entering the market is not because we think it must be at the bottom, but in the long run, it is cost-effective to enter this place, and the probability of making money is relatively high.