Understanding Power Ratios, FQI and Evidence Quality

What power ratio means

Power ratio is your displayed power divided by enemy power. A value below 1.00 means the player is underpowered on paper. The ratio is useful because the same win is more informative when it happens at a larger disadvantage.

Conquest Lab does not publish exact lineup ratios. The values remain private and are used to compare evidence internally.

Outcome margins

Each recorded run uses the surviving side and remaining heroes to describe the result. A narrow win is different from a full-team instant win. Multiple attempts help separate a repeatable formation from battle variance.

Formation Quality Index

The workbook compares the observed outcome with a power-only expectation. When a team performs better than power predicts, the Formation Quality Index is positive. When it performs worse, the formation may be the problem rather than development alone.

What counts as trustworthy formation evidence

• The row contains a recorded win.
• The player-to-enemy power ratio is below the near-equal-power exclusion range.
• The result is not labeled overpowered, power advantage, or instant.
• The battle did not leave the full five-hero team alive.
• The formation and positions are complete enough to interpret.

Evidence labels

Guide conclusions use qualitative labels—anecdotal, limited, moderate, strong, or unverified—rather than publishing changing row counts. These labels reflect how much relevant evidence exists and how consistent it is.

• Anecdotal: an observation worth testing, not a rule.
• Limited: some supporting evidence, but lineup context matters heavily.
• Moderate: a repeatable pattern with meaningful support.
• Strong: a pattern seen often enough to guide decisions confidently.
• Unverified: projection based on skills, tags, stats, or earlier-generation trends.

Limits of the math

The dataset comes from real account progression, not randomized laboratory trials. Heroes appear at different stages and generations, account power changes over time, and some heroes are tested more aggressively than others. Rankings therefore combine battle evidence with role utility, position fit, stats, and generation ceiling rather than using raw win rate alone.