In the highly volatile and promising world of Ethereum, informed investment decisions require expert knowledge and comprehensive data analysis. Goldmine Alerts serves this exact purpose. We deploy an exhaustive, data-driven process to uncover promising crypto opportunities.

## How Does It Work?

Our system builds on a three-pronged approach: transaction parsing, wallet analysis, and coin parameter simulation.


## Transaction Parsing

The process begins with an ETH Node, our main source of raw Ethereum data. This data undergoes an extensive cleansing process in the Swap Parser, where we remove noise and extract relevant information. We constantly analyze the percentage of transactions that we parse with our Parsed Data Deed Dive and continually work on its improvement. Currently, we do not parse 0.5% of all DeX swaps that exhibit custom logic, including custom routers and coins with overly complex logic. This means our calculation accuracy stands at 99.5%.

As of this writing, we possess around 5 months of data and our analysis is based on the following data volume:


## Coin Analytics

To perform accurate coin analytics, we need clean data. Even correctly parsed swaps may occasionally need exclusion from the analysis:
- Sandwiched swaps where price can significantly deviate from the standard coin price.
- Swaps from special whitelisted or dev wallets that can trade coins at starkly different prices.

The objective of cleaning coin prices data is to set a robust foundation for further analysis.

The subsequent step is to populate essential coin metrics with real-time transactions and keep them updated. These metrics include total invested in coins, the number of traders, liquidity, and other parameters derived from months of research. These parameters correlate with a coin's future growth potential.

We present these analytics in the following format, showcasing most parameters.


## Wallet Analysis

Next, we analyze wallet profits, losses, and other parameters to categorize wallets. This step is key to discerning investor behaviors and the performance of different coins. It's crucial to ascertain whether a wallet belongs to the dev team or a legitimate trader who cares about their investment. We also forecast wallet future statistics based on past performance.

Dev teams usually use bots and something we term a 'single-coin-wallet' created to purchase only one coin, without any intention of selling it.

There are also wallets that profit from very low cap coins with a small number of total traders. We consider such trades as losses.

Wallet analysis provides us insights into all wallets in the blockchain. Given a sufficient amount of transactions, we can forecast the quality of their investment decisions.


## Coin Parameter Simulation

Coin and wallet analysis are interconnected. Over the past 2 years, we've strived to select the right parameters that can, with a good level of accuracy, predict whether a coin's price will increase or decrease.

We can't definitively predict the fate of a single coin, as the market is in constant flux and countless factors can influence outcomes.

However, statistical analysis over the past few months yields solid results.

We consider simple milestones that coins may reach in their lifetime if they're successful: reaching a certain investment level (250k, 500k, and 1m).

Here you can see the conversion funnel, e.g., the percentage of coins that reached 500k from those that previously reached 250k:


Our goal is to identify these coins and apply our analytical parameters to exclude those with low likelihood of growth. We aim to find the 'gold' among the 'rocks'.

Next, we need to identify which coins turned out to be profitable.

For each alert we generate, we select an average price for the next 5 minutes following the alert. This simulates a manual purchase of the coin, considering the price can increase within a few blocks.

Then we apply a straightforward selling strategy: selling 50% on 2x and the rest on 4x. We always use accumulated average price data to calculate the selling price, ensuring it is realistic. After that, we apply accumulated statistics, calculating total X. If coins reach 20x according to this strategy, we will still sell at a maximum of 4x (or any first x of accumulated data that is higher than 4x).

Here I'll present statistics with and without our filters for coins that reached 250k investments over the past 3 months.



The total number of coins is 1903, which is more than sufficient to build statistics. The resulting X is around 1, implying neither profit nor loss. However, when factoring in gas price and unpredictable events like blacklisting wallets or other possible tricks, this X becomes a loss.

Now, let's examine how the statistics change when we apply our filters.


Observe that the number of coins reduced from 1903 to 111. These are the remaining coins, but the resulting x is 1.46. This figure is quite high as a cumulative return, especially considering the simplicity of our selling strategy, which can be optimized.

This is how Goldmine Alerts are created. We're working on numerous other features, which I'll be excited to share in the next article.

Goldmine Alerts isn't just another crypto alert system. We are a platform built on rigorous statistical analysis and market research. We delve deep into data to present our users with the golden opportunities of the crypto world.

Join us today and become a part of this thrilling journey.

https://twitter.com/crypto_rick_gm
https://t.me/Goldminealerts_bot