After this research on the quantification of demonic coins, it happened to coincide with the recent explosion of demonic coins It's obvious that there's been more discussion pushed up, and Jian Zhongzhong's understanding of the demon coin alpha is far ahead of other languages And at the same time, I also received a lot of feedback from frontline traders. Here are a few inspirations that inspired me to participate in testing and made some interesting discoveries: 1. @ Aryaw_eb3 Yang Jie's OI brushing volume+cross exchange liquidation https://(((x.com)))/arya_web3/status/2042909187425103952? s=46&t=mBkGyMIcMFVYls_XirwEFw Most OIs are generated through brushing, which has a low cost, but there are almost no single large orders in the order book. The main force of liquidation may not necessarily be on Binance Following this approach, two data dimensions were extracted: inter institutional OI concentration and vol/OI Two conclusions can be drawn: -The lower the median OI ratio of Binance, the higher the degree of manipulation, and the higher the win rate of V4A, a strategy based on highly manipulated features (with a 170% increase in backtesting) -A high vol/OI means a high brush volume. V4A performs the worst among currencies with a brush volume of over 20x This also means that all strategies that simply monitor contract trading volume or OI are inappropriate, as it is likely that Zhuang deliberately stopped brushing to lure them into the car. This is a very good potential screening criterion In addition, on chain inflows and outflows from exchanges are also feasible indicators, which are basically synchronized with the V4A signal 2. @ wuk_Bitcoin Donato's "Price OI Deviation" https://(((x.com)))/wuk_bitcoin/status/2018017280685469758? s=46&t=mBkGyMIcMFVYls_XirwEFw The logic is that the price keeps rising, but the OI drops significantly and monotonically for several consecutive hours. This logic indicates that there is smart money or insider information quietly selling out I tested the data and found that this situation involves a large number of samples, and more importantly, it does not rely on manipulating currencies or events. It is an independent strategy that generates signals and can generate positive EVs (of course, the vast majority of backtesting data shows bear markets), which has a low degree of overlap with my previous V4A strategy Of course, there are also limitations: 1. Need a very tight rebound trail to close positions 2. After backtesting the sliding point cost, the single PNL is lower 3. Easily affected by OI manipulation 3. @ CryptoRounder teacher's rapid insertion+OI drop https://(((x.com)))/cryptorounder/status/2039651219782979634? s=46&t=mBkGyMIcMFVYls_XirwEFw This may be the best discovery for supplementing the V4A strategy - based on the manipulation of currencies, manipulation range filtering, and top signal filtering of pins in V4A, detect whether there is a sudden decrease in OI within 30 minutes The logic of this strategy is that the price at this position triggers a large amount of liquidation, leading to a sharp drop in OI. As the short positions disappear from the opponent's market, the market makers lose their willingness to maintain high prices and begin to decline, which is a more accurate top signal What surprised me greatly was that even though I established this strategy based on the top confirmation and manipulation cycle confirmation conditions of V4A, the backtesting almost did not overlap After backtesting, it was confirmed that Professor Rounder's strategy trigger was 2-4 hours faster than my previous strategy trigger based on pure K I added the strategies of Rounder and Donato to the observation as V7 and V8 To summarize, the current three strategies V4A (still under observation for old and new coins), V7, and V8 are shown in the following figure Finally, let me advertise again: To test your demon coin strategy, come to @ Hertzfflow_xyz Testing Network