The Stonk Analyzer
Figure 1: Stonk Analyzer Workflow. Sections in blue are real-time market data.
Stonk Analyzer employs a machine learning model based on available variables and data points such as in-game stats, the market data, directional accuracy, magnitude and compare these outputs with the results using a standard trend line or dutch line equation.
Where,
Slope(b) = (NΣXY - (ΣX)(ΣY))(NΣX2 - (ΣX)2) Intercept(a)=(ΣY - b(ΣX))N
Using these comparisons, we will be able to “quantify” the finfluencer’s success as it relates to actual market performance. With a combination of factors, we can ascertain as to: how successful the finfluencer's are with their market predictions, determine the risk level of their predictions and their historical profitability. Some of the variables that will be modeled include their accuracy, directional sentiment, frequency of their change in positions among other factors to determine their probability of investment success. With these, we will be able to create the top-10, top-100 and top-300 call-makers as well as determine the constituents of the index itself.
Going forward, we will look to also implement new strategies based on historical results and devise a way of computing using Markov Chain (probability of making more money than previous year). We will also attempt to develop new methods for using stochastic volatility methods, in addition to well-known mathematical methods such as Black-Scholes model. All these methods and strategies will be implemented using IOSCO's principles of financial benchmarks. The integration with Uniswap will allow for seamless CLOB trading in the initial phase. Over time, we will develop the ability to enable easier cross-chain trading between different networks such as Polkadot, BSC, Solana and Ethereum.
Sentiment Ranking and Positional Magnitude
StonkLeague games have been constructed with 3 core tenets intended to offer an optimal gaming experience for casual players while also maintaining consideration for active and professional traders:
Cash-Based Gaming: contrary to many options on the market, stonk games are designed to offer real-cash competition. We believe this offers numerous values but most importantly incentivizes thoughtful player participation as each entry mandates a player put a financial contribution at stake in the form of an entry cost. While as an added benefit, this affords greater prize pool rewards as entry fees become the basis for the prize pool allocations.
Top of the Morning Style: Each stonkleague game is intended to be simple, and time efficient. Unlike full portfolio simulations which demand a player actively monitor a set amount of monopoly money assets, StonkLeague games require less than 5 minutes to play -- though players may engage more actively if they so choose. This is to maintain consideration for full-time traders who cannot afford, and would prefer not, to manage an entirely simulated portfolio through
Metric Capturing: All games have been designed to capture an essential snapshot of true market beliefs and behaviors. Core to any standard financial investment is directional sentiment (bullish or bearish) and magnitude of sentiment typically denoted by position size. Each StonkLeague game is designed to capture both of these elements as to best garner functional, valuable market insights and to create a more realistic market abstraction layer.
Last updated