- Joined
- Jan 31, 2019
Can you run it for pairs as well?For funsies, I asked ChatGPT to analyze the RD and FD score sheet along with the panel of judges, and here's what it gave me:
SUMMARY MEMO
RE: Analysis of Judging Bias — 2026 Olympic Winter Games, Ice Dance
DATE: February 19, 2026
PURPOSE
This memo summarizes findings from a statistical analysis of all nine judges' scores across both the Rhythm Dance (RD) and Free Dance (FD) segments of the 2026 Olympic ice dance competition in Milan, Italy. The analysis sought to determine whether any judge exhibited systematic bias — for or against specific teams — and whether such bias materially affected final standings.
KEY FINDING
One judge — Jézabel Dabouis (France) — exhibited a pattern of nationalistic bias across both competition segments that, with high probability, altered the gold medal outcome.The final combined margin between gold medalists Fournier Beaudry/Cizeron (FRA) and silver medalists Chock/Bates (USA) was 1.43 points. Dabouis's scoring pattern produced cumulative deviations well in excess of this margin.
BACKGROUND
- Nine judges scored each segment. The ISU system trims the highest and lowest marks at the individual element/component level before computing scores.
- Five judges served in both segments; four were unique to each segment.
- Dabouis was the only judge from a gold-medal-contending nation who served in both segments, compounding the opportunity for bias to accumulate.
DABOUIS (FRA) — DETAILED FINDINGS
Scoring Differential: FRA vs. USA Top Teams
Export as CSV
Segment Dabouis FRA–USA Gap Panel Median FRA–USA Gap Deviation from Median Rhythm Dance +5.74 pts +0.29 pts +5.45 pts Free Dance +7.71 pts −0.03 pts +7.74 pts Combined +13.45 pts +0.26 pts +13.19 pts
Dabouis's combined two-segment gap between the French and American top teams was the largest of any judge on the panel.
Pattern of Bias
The bias operated in both directions simultaneously:
- Pro-French: Dabouis gave Fournier Beaudry/Cizeron the highest score on the panel in the RD (93.34 vs. panel mean ~89.96, deviation +3.38) and among the highest in the FD (137.45 vs. panel mean ~135.09, deviation +2.36). She also boosted the second French team (Lopareva/Brissaud) by approximately +1.98 points above panel mean in the FD.
- Anti-American: Dabouis scored Chock/Bates 4.64 points below the panel mean in the FD (129.74 vs. ~134.38). She was at or tied for the lowest GOE on 7 of 9 elements for Chock/Bates in the FD, while simultaneously being at or tied for the highest GOE on 6 of 9 elements for Fournier Beaudry/Cizeron. She also scored all three USA teams below their final placements in the FD (average deviation: −2.3 ranks).
Impact on Gold Medal
The cumulative effect of consistent small-to-moderate biases across dozens of individual element and component marks — many of which would not individually be extreme enough to trigger the ISU's trimming mechanism — plausibly exceeded the 1.43-point final margin. Multiple recalculations confirm that excluding Dabouis's scores would have reversed the gold and silver medal positions.
RECIPROCAL USA JUDGE BIAS — CONTEXT
The FD panel included a USA judge (Janis Engel, Judge 5) who showed a reciprocal pro-American tendency:
Export as CSV
Metric Dabouis (FRA) Engel (USA) Home–rival differential +7.71 (FD) +4.10 (FD) Direction Boosted FRA and suppressed USA Primarily boosted USA; scored FRA reasonably Elements at/tied for extreme GOE on rival 7 of 9 (lowest) 1 of 9 (lowest)
Engel's bias was approximately half the magnitude and unidirectional (boosting without suppressing), making Dabouis's pattern the dominant influence on the final result.
OTHER NATIONALISTIC BIASES IDENTIFIED
Export as CSV
Judge Nat. Segment Home Team Boost Outcome Impact Christian Baumann RD only van Rensburg/Steffan: +6 rank deviation, +4.92 pts above mean None (team finished 23rd) Leslie Keen RD + FD Lajoie/Lagha: +5 rank deviation in FD, +7.84 pts above mean None (large point gaps at those positions) Nicholas Russell RD only Fear/Gibson +2, Bekker/Hernandez +4 rank deviation; highest-variance judge in RD None Patricia Moritz RD only Harris/Chan: +4 rank deviation, +6.22 pts above mean None Virpi Kunnas-Helminen RD + FD Turkkila/Versluis: +3 (RD), +2 (FD) rank deviation None Richard Kosina RD + FD CZE teams: +3.5 avg rank deviation (RD); mixed in FD None Marta Olozagarre FD only Smart/Dieck +2, Val/Kazimov +3 rank deviation; highest scores on panel for both None Isabella Micheli FD only Guignard/Fabbri: +1 rank deviation, +5.43 pts above mean None
All judges with home-country teams in the competition showed some degree of positive scoring bias toward those teams.
However, none of these biases was consequential to final standings due to larger point separations at those ranking positions.
STRUCTURAL CONCERN
The ISU assigned Dabouis — the judge from one of the two primary gold-medal-contending nations — to both segments of the competition.This structural decision allowed an individual's bias to compound across the full event. By contrast, no USA judge served in the RD; Engel's influence was limited to one segment.This asymmetry in panel composition amplified the effect of Dabouis's bias relative to any countervailing American bias.
CONCLUSIONS
- Nationalistic bias was universal. Every judge with a compatriot team showed measurable pro-home-country tendencies, consistent with decades of academic research on figure skating judging.
- Only one case was consequential. Jézabel Dabouis's systematic pattern of inflating French teams and suppressing American teams across both segments — producing a combined 13.19-point deviation from the panel median on the FRA–USA differential — is the only bias that plausibly altered a final ranking.
- The gold medal outcome is in question. With a final margin of just 1.43 points and Dabouis's deviations far exceeding that threshold, the evidence strongly suggests the gold medal would have been awarded to Chock/Bates absent Dabouis's influence.
RECOMMENDATION
This analysis supports a formal review of:
- The ISU's panel assignment protocols, particularly the practice of assigning judges from medal-contending nations to both segments
- The adequacy of the current trimming system in mitigating cumulative, moderate-but-consistent bias
- Whether supplementary statistical monitoring (e.g., real-time deviation tracking) should be implemented at future Olympic competitions
What that means, of course, is that this kind of thing will just keep happening again and again. 
, but oh well -- talking is what NVC pays them for, after all.
