Medianization Algorithms

Medianization algorithms are computational methods used in decentralized finance and cryptocurrency price feeds to determine a representative asset price by selecting the middle value from a set of reported data points. By ordering all received price submissions from various sources, these algorithms effectively filter out extreme outliers caused by technical glitches, latency, or malicious manipulation attempts.

This approach ensures that the final price output is robust and resistant to the influence of a small number of corrupted or inaccurate nodes. In the context of oracle networks, medianization provides a reliable mechanism for settling derivative contracts that require accurate external data.

It is a fundamental technique for maintaining data integrity in trustless environments where no single source can be assumed to be honest. The algorithm functions by discarding the highest and lowest values until the central point is identified.

This process is essential for preventing price spikes that could trigger unwarranted liquidations in margin-based trading systems. Because the median is less sensitive to volatility than the mean, it serves as a more stable anchor for protocol operations.

Ultimately, these algorithms protect the solvency of decentralized protocols by ensuring that the reference prices used for collateral valuation remain accurate and difficult to manipulate.