How would you design a database schema to handle high write throughput while still supporting analytical queries?

Explain how MVCC (Multi-Version Concurrency Control) allows readers and writers to avoid blocking each other in databases like PostgreSQL.