Conclusion

• The diversity between choosing blocks with lower transactions count compared to higher ones was higher among the inactive miners;
• The border has become blurry since then and the new miners mine almost every block they can;
• While the data shows a tendency toward blocks with higher transaction count for active miners, it has been due to the increased average of transactions in each block compared to blocks mined previously;

Analysis

• During the 2600 days lifespan of the network, more than 15M blocks have been mined and 1.7B transactions have been confirmed by slightly above 5.6K miners;
• On average, each miner mined 2.7K blocks with an average transactions count of 110;
• Out of two miners groups, the active miners have been responsible for nearly two-thirds of the overall mining activity;
• While the minimum, the maximum, and the average number of transactions per block were in a small range for inactive miners, the numbers for active miners have covered a wide range with a considerable margin;
• The average size of mined blocks for active miners has been much higher compared to inactive miners, indirectly indicating their tendency to mine blocks with higher transactions count;
• The higher lifespan of the active miners compared to a month and a half for the inactive miners resulted in putting more focus on the analysis of this group.

Introduction

The Merge is coming to put an end to the current Proof-of-Work (PoW) consensus mechanism on Ethereum, which also means an end to the mining of new blocks. Mining is the lifeblood of PoW and Ethereum miners utilize their time and computation power to process and confirm transactions, and ultimately produce blocks. With the miners being phased out in favor of Proof-of-Stake (PoS) with the upcoming Merge, an overview analysis of the mining ecosystem of Ethereum in the course of more than seven years has been presented in this dashboard.

Methodology

The required data for this analysis were collected from the fact_blocks table of the core schema of the ethereum database. During its lifecycle, Ethereum has seen numerous miners helping the network to become decentralized. That means many of them might not be currently active or operating. As a result, the miners were categorized into two different groups of active and inactive miners. The active miners were considered those who mined at least one block in the past 30 days. For the analysis of the blocks-transactions distributions, only miners with a lifespan of more than 1 day were considered to remove unreliable data points and outliers. For the analysis of transactions-count distribution, all of the miners as well as the active miners were evaluated.

Active Miners

• Among the active miners, ethermine, f2pool, and nanopool have been the three largest mining pools with more than 1M mined blocks;
• These three pools have had a mined blocks share of nearly 75% compared to other active miners, combined;
• They have also had a confirmed transactions share of more than 70% of the total, among the active miners, with more than 1M confirmed transactions each;
• The majority of the active miners have had a high average transaction count in a similar range, with only a few exceptions;
• Only 12 miners have had a minimum transaction count of more than 10. Except for these, almost all the other active miners mined blocks with transactions counting as low as 0 and as high as more than 1400 transactions per block;
• All the miners in the top half of the spectrum have had a high maximum transactions count, showing their ability/tendency to mine blocks with a high number of transactions;

Inactive Miners

As Ethereum mining is coming to an end, inactive miners have been separately investigated to showcase their legacy. Among the top inactive miners, sparkpool and dwarfpool were the two most dominant mining pools. The measurement of average transactions count and the maximum transactions count of the mined blocks by these miners have indicated that not all of them went all in to mine a block. For example, while sparkpool had a high average and a high maximum, dwarfpool focused on blocks with lower transactions-count and both its average and maximum were much lower compared to sparkpool. This comparison applies to all other notable inactive miners and the data shows a good mix of both approaches.

Miner Groups

• Miners were segmented into incremental logarithmic groups based on the number of blocks they mined;
• While groups with higher block ranges have had fewer miners, their share of mined blocks and confirmed transactions has been higher compared to groups of miners with fewer blocks;
• For all of the miners, as the number of mined blocks increased, their average transactions count, as well as their maximum transactions count have also increased;
• For the active miners, while the maximum has been lower for smaller groups, it reached a steady high for the top groups.
• All the active miners' groups have had a high average for the number of transactions per each block;
• While the blocks-transactions distribution clustered in lower values for all miners, it has steadily increased for active miners;
• The distribution compared with the maximum transactions count of mined blocks has been similar for both the active miners and all of the miners;
• In general, the average transactions count has been much higher for almost all of the active miners compared to all, except the top miners with a slightly lower average due to the high number of blocks they mined, which has flattened their average;

The average 110 transactions count per block was considered as the threshold separating low and high transactions count. While the number of miners for high and low has been almost identical, the number of blocks mined by higher transactions count has been higher among the active miners. Also, the total number of transactions confirmed by these miners from the blocks with high transactions count has been much higher compared to the low ones. The monthly analysis of the mined blocks over the past 7 years has also indicated that in recent years, the tendency to mine blocks has moved from blocks with a low transactions count to blocks with higher transactions count. One major reason is the fact that the network activity has immensely increased compared to the first few years and the average transactions count of each block has gradually increased.