Should the Olympic sprint skaters run the 500 meter twice?

The Olympic 500 meter sprint competition is the Formula One event' of speed skating, and is watched by millions of television viewers. A draw decides who should start in inner lane and who in outer lane. Many skaters dread the last inner lane, where they need to tackle heavier centrifugal forces than their companions in the last outer lane, at maximum speed around 55 km/hour, at a time when fatigue may set in. The aim of this article is to investigate this potential difference between last inner and last outer lane. For this purpose data from eleven Sprint World Championships 1984--1994 are exploited. A bivariate mixed effects model is used that in addition to the inner-outer lane information takes account of different ice and weather conditions on different days, unequal levels for different skaters, and the passing times for the first 100 meter. The underlying unfairness parameter’, estimated with optimal precision, is about 0.05 seconds, and is indeed significantly different from zero; it is about three times as large as its estimated standard deviation. This is enough for medals to change necks. Results from the work reported on here played a decisive role in leading the International Skating Union and the International Olympic Committee to change the rules for the 500 meter sprint event; as of the Nagano 1998 Olympic Games, the sprinters are to skate twice, with one start in inner lane and one in outer lane. The best average result determines the final list, and the best skaters from the first run are paired to skate last in the second run. It has also been decided that the same rules shall apply for the single distance 500 meter World Championships;these are arranged yearly from 1996 onwards.

💡 Research Summary

The paper addresses a long‑standing concern in Olympic speed skating: the 500 m sprint, a marquee event watched by millions, is run only once and the lane assignment (inner versus outer) is decided by a draw. Because the inner lane has a smaller radius (≈25–26 m) than the outer lane (≈29–30 m), a skater who finishes the race on the inner lane must negotiate a larger centripetal force (mv²/r) at peak speed (≈55 km/h). This extra force, combined with the onset of fatigue, can make the inner‑lane finish more difficult and potentially slower. The author set out to quantify this “unfairness” using data from eleven Sprint World Championships (SWC) held between 1984 and 1994, each of which provides two 500 m runs per skater (one on each lane) together with a 100 m split time.

Data and Model
For each skater i on day 1 and day 2 the observed variables are:
‑ x₁ᵢ, x₂ᵢ – 100 m split times,
‑ Y₁ᵢ, Y₂ᵢ – final 500 m times,
‑ z₁ᵢ, z₂ᵢ – lane indicator (‑1 for inner, +1 for outer).
The core statistical model is a bivariate mixed‑effects regression:

Y₁ᵢ = a₁ + b·x₁ᵢ + cᵢ + ½·d·z₁ᵢ + ε₁ᵢ
Y₂ᵢ = a₂ + b·x₂ᵢ + cᵢ – ½·d·z₁ᵢ + ε₂ᵢ

where:
‑ d is the “unfairness parameter” (average time difference between inner‑ and outer‑lane finishes),
‑ cᵢ represents each skater’s latent ability, assumed N(0, κ²),
‑ ε₁ᵢ, ε₂ᵢ are day‑specific random errors, assumed N(0, σ²) and independent.

The model captures day‑to‑day variations through separate intercepts a₁, a₂ (and a common slope b linking the 100 m split to the 500 m result). The random‑effects structure induces a bivariate normal distribution for the pair (Y₁ᵢ, Y₂ᵢ) with a covariance matrix that reflects both within‑skater ability (κ²) and measurement error (σ²). The intraclass correlation ρ = κ²/(κ²+σ²) quantifies how stable a skater’s performance is across the two days.

Parameter estimation proceeds via restricted maximum likelihood (REML). The author examined whether the slope b should differ between days (b₁ ≠ b₂) and whether the error variances should differ (σ₁ ≠ σ₂). Empirical tests showed that imposing b₁ = b₂ and σ₁ = σ₂ does not materially affect the estimate of d, while simplifying the model and improving precision.

Results
Table 1 in the paper lists the estimated d̂ and its standard error (SE) for each championship:

• 1984 Trondheim: d̂ = 0.131 s, SE = 0.038 s
• 1985 Heerenveen (outdoor): d̂ = 0.090 s, SE = 0.058 s
• 1986 Karuizawa: d̂ = 0.035 s, SE = 0.066 s
• 1987 Sainte‑Foy: d̂ = –0.151 s, SE = 0.080 s
• 1988 West Allis: d̂ = –0.147 s, SE = 0.090 s
• 1989 Heerenveen (indoor): d̂ = 0.128 s, SE = 0.047 s
• 1990 Tromsø: d̂ = 0.096 s, SE = 0.087 s
• 1991 Inzell: d̂ = 0.023 s, SE = 0.040 s
• 1992 Oslo: d̂ = –0.019 s, SE = 0.086 s
• 1993 Ika​ho: d̂ = 0.032 s, SE = 0.041 s
• 1994 Calgary: d̂ = 0.010 s, SE = 0.043 s

The weighted grand average across all eleven events is d̂ = 0.048 s with SE = 0.016 s, yielding a t‑ratio of about 3 (p ≈ 0.001). Even after discarding four “noisy” championships (those with SE > 0.08 s), the remaining seven give d̂ = 0.065 s, SE = 0.017 s (t ≈ 3.8). Thus the inner‑lane disadvantage is statistically significant and of a magnitude that can alter medal positions, given that Olympic 500 m victories are often decided by hundredths of a second.

The paper also discusses the physical meaning of a 0.05 s difference: at a top speed of 26 s for 400 m, a skater covers roughly 15.4 m per second, so 0.05 s corresponds to about 0.75 m, i.e., 0.15 % of the race distance. In longer events this translates to 15 m over 10 km or 65 m in a marathon, underscoring the practical relevance of even a few hundredths of a second.

Policy Implications
Armed with this quantitative evidence, the author submitted a proposal to the International Skating Union (ISU) in 1992, arguing that the 500 m should be run twice—once from each lane—and that the final ranking be based on the average of the two times. The proposal also suggested a “reverse start list” for the second run, mirroring formats used in alpine skiing and ski jumping, to keep the competition spectator‑friendly. The ISU Technical Committee endorsed the idea, and after a vote at the 1994 Boston Congress the rule change was adopted unanimously. Consequently, starting with the 1998 Nagano Winter Olympics, both men’s and women’s 500 m events have been contested twice, with the average time determining the medals. The same rule was applied to the single‑distance World Championships from 1996 onward.

Additional Observations
The author notes that women’s 500 m appears less sensitive to lane effects; data show smaller d̂ values and fewer accidents, which aligns with anecdotal reports that female skaters cope better with the inner‑lane curvature. Nonetheless, the male data provide compelling statistical proof that the single‑run format was unfair.

Conclusion
Through a rigorously constructed mixed‑effects model, careful handling of day‑to‑day environmental variation, and a thorough outlier analysis, the paper demonstrates that the inner lane imposes a measurable disadvantage of roughly 0.05 seconds on average. This difference is large enough to affect Olympic outcomes, justifying the rule change that now requires two 500 m runs with opposite lane assignments. The study stands as a textbook example of how statistical methodology can directly influence sport governance and improve fairness at the highest competitive level.