EIP-1559 is said to be one of the biggest changes in the history of Ethereum. Previous to the London fork (EIP-1559 was deployed on August 5), the design of the new transaction fee mechanism was widely discussed for a few years by the core community, and meticulously modeled and simulated in the context of several academic-like initiatives.
Nearly three months have passed now since the deployment of these new features and we think that it’s time to start seeing analytics initiatives using 1559-specific on-chain data, regarding the potential effects of the new mechanism on MATIC total supply over time, changes on validators incentives, UX improvements for transaction creators, increasing incentives for exploiting MEV opportunities, among others.
We present a best, base, and worst-case analysis of the potential impact of EIP-1559 on MATIC total supply. After proving the stationarity of our time series, we take static histograms and compute quantiles related to worst, base, and best case analysis. Our results offer annual estimates for the impact of MATIC burning on the total supply.
We use ETH burnt / tx_fee ratio observed distribution as a proxy for the estimated distribution of expected ratios on Polygon PoS. From real data ETH ratio distribution, we take quantiles .05, .25, .5, .75, and .95 for best-best, best, median, worst, and worst-worst case analysis. In this context, the less MATIC we burn the better (in terms of total MATIC supply impact).
Then, we summarize and compute descriptive statistics over the distribution of daily tx fees from the Polygon dataset limited to the September time window. We use the following stats as the potential parameters for worst and best analysis, and intermediate cases: min, max, mean, median, .25, and .75 quantiles.
Finally, we get a matrix of the spectrum from best to worst cases combining burn/tx_fee ratio cases from Ethereum data and the stats set from Polygon PoS data.
Here you can find a short summary of our first report.