Duel Outcomes Above Expected

Duel+R6v0.1.0 · experimental · validated 2026-07-16

Computed · experimental. Values are computed, but the definition or model may still change.

What it measures

Duels won above what each situation predicted, per 100 duels. Positive means winning fights the situation said they should lose.

Formal definition

Σ(actual − xDuel) over every engagement the player took part in (as attacker: actual = 1; as victim: actual = 0 against the killer's expectation), scaled per 100 duels, using the R6 coefficient set of the shared duel model.

Inputs

duel events with pre-outcome context
xduel_r6

Exclusions

None beyond the standard confidence floor.

Sample & population

Minimum sample: 60 duels
Baseline population: Players in the same game/window with ≥ minimum duels
Direction: higher better
Unit: wins/100 duels

Percentile interpretation

Read within role: entry fraggers and anchors face systematically different duels.

Confidence = source confidence × model confidence. Values from baseline models are capped at 0.7. Below-minimum samples display the value grayed out with no percentile.

Example

Winning a 0.3-xDuel hold through a freshly opened hatch contributes +0.7.

Known limitations

v0 coefficients are synthetic-calibrated priors, not fitted on pro Siege
The model has no term for operator abilities or destructible geometry — a hole shot before the fight is invisible to it
Vertical advantage is approximated by floor delta only

Model card

xduel_r6 v0.1.0

MODELlogistic_regression · baseline

Training data — read this

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

Calibration

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.

Features

weapon_advWeapon-class matchup advantage (attacker perspective, -1..1)
dist_closeEngagement below close-range threshold
dist_longEngagement beyond long-range threshold
health_diff(attacker HP − victim HP) / 100 at engagement
armor_advAttacker has armor
victim_flashedVictim blind fraction (blind ms / 1500, capped)
attacker_flashedAttacker blind fraction
through_smokeEngagement through smoke
attacker_supportedAttacker teammate within trade range
victim_isolatedVictim's nearest teammate beyond trade range

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 intended for

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

Change log

v0.1.0 · 2026-07-16Initial baseline release

→ Rank players by Duel+