Bayesian Average

$\mathrm{Score}=\frac{C\times M+R\times W}{C+R}$


C = confidence constant
M = global average win rate
R = runs with card
W = card win rate

Example 1: The "Lucky" Card (1 run, 1 win)

$\mathrm{Score}=\frac{5\times 0.20+1\times 1.0}{5+1}=\frac{2}{6}=\mathrm{33.3\%}$

Example 2: The "Proven" Card (10 runs, 6 wins)

$\mathrm{Score}=\frac{5\times 0.20+10\times 0.60}{5+10}=\frac{7}{15}=\mathrm{46.7\%}$

Result: The proven 60% card (46.7%) outranks the lucky 100% card (33.3%).