the tech lab · list analysis
test your sixty.
find the tech worth testing in your sixty. paste your list, check the evidence, then take a hypothesis to your next match.
open a result card to inspect match counts and copy choices.
online tournament lists and matches inform this estimate.
experimental: historical comparisons may include incomplete decklists and unresolved match results. use these estimates to plan tests, not predict tournament performance. card choice, player skill, and the rest of the list can all affect a result. a positive delta does not prove that adding a card will improve your win rate.
how the estimate works
the model compares matches with and without a card. Bayesian smoothing reduces the effect of small samples. opponent frequency sets the field weights.
core cards (≥75% inclusion) are excluded from the list adjustment. their individual comparisons remain visible. tech cards have <40% inclusion. copy-specific estimates appear when enough data exists.
the approximate 90% model interval sums card variance. it excludes baseline uncertainty and dependence between cards. inspect match counts before choosing a tech to test.
both pools match counts and averages source inclusion rates. online match history can extend beyond the period used for current field weights.