EV Lab
Bayesian meta-weighted expected value · paste any decklist · see which techs move the needle
— paste a list to detect —
▾
How it works
Each card is split into matches where you ran it vs. didn't. Win rates are estimated with a Beta prior (empirical Bayes), then weighted by how often each opponent archetype appears in the meta.
CORE cards (≥75% inclusion) contribute zero delta by definition. TECH slots (<40%) are where the signal lives.
When enough data exists, EV is computed at your exact copy count vs. zero copies, and the copy curve shows win rate at each count in the matchup panel. The ×N avg column shows the typical count players use when they include the card.
Each card is split into matches where you ran it vs. didn't. Win rates are estimated with a Beta prior (empirical Bayes), then weighted by how often each opponent archetype appears in the meta.
CORE cards (≥75% inclusion) contribute zero delta by definition. TECH slots (<40%) are where the signal lives.
When enough data exists, EV is computed at your exact copy count vs. zero copies, and the copy curve shows win rate at each count in the matchup panel. The ×N avg column shows the typical count players use when they include the card.