Models

Advanced metrics are produced by versioned, interpretable models. Each card below states exactly what the model estimates, its inputs, how it was calibrated, and where it should not be used. Values from a model are always labeled model-derived and never presented as official results.

Expected Duel Probability (CS2)

xduel_cs2 · v0.1.0

CS2logistic regressionbaseline

Estimates

Difficulty-adjusting duel outcomes (Duel+, Opening Impact); relative comparisons within a game

Calibration

Hand-specified directional priors, sanity-calibrated on the synthetic fixture corpus v1 (fictional matches). NOT fitted on production esports data.

Reliability checked on synthetic corpus only; expect miscalibration on real data until re-fit. Confidence for downstream metrics is capped at 0.7 while status=baseline.

Feeds these metrics

Expected Duel Win ProbabilityDuel Outcomes Above ExpectedOpening Impact

Coefficients (v0.1.0)

intercept0.00

weapon_adv1.80

dist_close0.10

dist_long-0.05

health_diff1.40

armor_adv0.25

victim_flashed1.60

attacker_flashed-1.60

through_smoke-0.70

attacker_supported0.15

victim_isolated0.20

Not for

Betting or wagering decisions · Cross-game comparisons · Absolute skill claims

Expected Duel Probability (VALORANT)

xduel_valorant · v0.1.0

VALlogistic regressionbaseline

Estimates

Difficulty-adjusting duel outcomes (Duel+, Opening Impact); relative comparisons within a game

Calibration

Hand-specified directional priors, sanity-calibrated on the synthetic fixture corpus v1 (fictional matches). NOT fitted on production esports data.

Reliability checked on synthetic corpus only; expect miscalibration on real data until re-fit. Confidence for downstream metrics is capped at 0.7 while status=baseline.

Feeds these metrics

Duel Outcomes Above ExpectedEntry Value Above Expected

Coefficients (v0.1.0)

intercept0.00

weapon_adv1.60

dist_close0.10

dist_long-0.05

health_diff1.50

armor_adv0.20

victim_flashed1.70

attacker_flashed-1.70

through_smoke-0.60

attacker_supported0.15

victim_isolated0.20

Not for

Betting or wagering decisions · Cross-game comparisons · Absolute skill claims

Expected Duel Probability (R6)

xduel_r6 · v0.1.0

R6logistic regressionbaseline

Estimates

Difficulty-adjusting duel outcomes (Duel+, Opening Impact); relative comparisons within a game

Calibration

Hand-specified directional priors, sanity-calibrated on the synthetic fixture corpus v1 (fictional matches). NOT fitted on production esports data.

Reliability checked on synthetic corpus only; expect miscalibration on real data until re-fit. Confidence for downstream metrics is capped at 0.7 while status=baseline.

Feeds these metrics

Entry Value Above Expected

Coefficients (v0.1.0)

intercept0.00

weapon_adv1.00

dist_close0.15

dist_long-0.10

health_diff1.30

armor_adv0.10

victim_flashed1.20

attacker_flashed-1.20

through_smoke-0.50

attacker_supported0.20

victim_isolated0.25

Not for

Betting or wagering decisions · Cross-game comparisons · Absolute skill claims

Round Win Probability (CS2-style)

round_wp_cs2 · v0.1.0

CS2logistic regressionbaseline

Estimates

Win-probability timelines; Round Win Probability Added attribution

Calibration

Hand-specified priors, sanity-calibrated on synthetic fixture corpus v1. NOT fitted on production data.

Terminal states pinned to truth; mid-round states are directional priors pending real-data fit.

Feeds these metrics

Round Win Probability AddedPlant Probability Added

Not for

Live betting · Cross-game comparisons

Read the overall limitations →