Part 1: Crossing Data with Snowboarding: An Investigation of Snowboard Cross

Time Trials and Head-to-Head Racing

This is the first in a multi-part series using analytics to bring some order and understanding to the marvelous chaos of snowboard cross.

INTRODUCTION

One thing that makes Snowboard Cross special is that it is the same activity on the same course done in two different formats. Competitions typically start with a time trial qualification round, in which every competitor makes solo runs down the course. In the Olympics, everyone qualifies for the next round, but on the World Cup circuit, the time trial also serves to cut athletes from the field. Inside the cut line, that first round also determines seeding for the heats, in which groups of 4 (occasionally 6) athletes make a run together and the top two make it through to the next round. The seeds determined by qualifying set the matchups in the heats, and start gates are selected in seed order. I will get to seed order in a later article. Ultimately, just one set of medals is awarded for the two disciplines that make up the qualifying round and heats round. But are the qualifying round’s time trialing and the heats’ head-to-head racing significantly different disciplines from each other? And if so, in what ways? To answer this, I analyzed FIS World Cup and World Championship results from events with both a qualifying round and a heats round from the start of the 2017 season through March of 2026.

DIFFERENTIATING BETWEEN THE DISCIPLINES

            If time trialing and head-to-head racing were not significantly different disciplines, then any athlete who is good at one discipline should be equally good at the other. So, to show that they are different, all we need to do is prove that one athlete is better at one discipline than the other. To find one, we only needed to look as far as the flag bearer for a 119-athlete delegation at the Opening Ceremonies of the 2022 Beijing Olympics.

Italy’s Michela Moioli is one of the best qualifiers in the world. Her average qualifying ranking of 3.97 ranks second among women with at least 10 World Cup starts since the start of the 2017 season. But when the time comes for the heats, Moioli manages to find an even higher gear.

I built a model that combines biographical factors for each athlete and situational factors to augment the predictability of heat outcomes from qualifying position and used it to determine how athletes perform relative to expectation.

In this graph, the men are on the left side and the women on the right. A blue bar indicates that the athlete is significantly better in heats than in qualifying, and an orange bar indicates that the athlete is significantly better in qualifying than in heats. Michela Moioli is one of the best qualifiers in the world, yet, over her 59 events, her bar is blue. It follows that she is one of the most decorated snowboarders of any discipline in Olympic history, with 4 Olympic medals from the last 3 Games, including one of each color, on top of 7 World Championship medals and 3 season-long World Cup titles.

            Of the 107 athletes with at least 10 competitions since the start of the 2017 season, just 14 of them, 7 men and 7 women, have positive Z-scores, indicating that their performance in heats has been better than expected. Interestingly, one of those athletes, Eva Adamczykova, is the only woman with a better mean qualifying rank than Moioli’s. Of the 5 athletes with the statistically significant blue bars, 4 have won Olympic gold, and the other has won silver and bronze at the Olympics and a World Championship medal of each color. These right-tailed distributions suggest that not only are time trialing and head-to-head racing different skills, but head-to-head racing is also the more difficult of the two. This is a particularly good sign for the futures of the three athletes with positive Z-Scores who are under the age of 23: 20-year-old Lea Casta of France, 22-year-old Leon Ulbricht of Germany, and 21-year-old Nathan Pare of the USA.

DESCRIBING THE RELATIONSHIP

            To augment the predictability of heat outcomes from qualifying position, I built an ordinal logistic regression to determine the probability that an athlete finishes in 1^st place, 2^nd place, 3^rd place, or below 3^rd (including any disqualification) in a given heat, with athletes who did not start their heat excluded.

While time trialing and head-to-head racing are significantly different from each other, qualifying rank is still the most impactful predictor of heat outcome. The relationship, however, is non-linear and varies by gender. Other statistically significant factors include rider age, the qualifying ranks of the mean heat opponent and the one at the cutoff between advancement and elimination (3^rd best in a heat in which 2 advance, for example), and the overall size of the field.

This graph shows the final rankings points earned in the overall event by qualifying rank and gender. There are two factors determining the nature of this graph. The first is the size of the field. Over 80% of the men’s events in this dataset have 32 athletes qualify into the heats with the rest qualifying 48. By contrast, over 80% of women’s events have just 16 athletes qualify with the rest qualifying 32. This is why the red line representing the women’s field ends at 32 and the grey standard error bar around it explodes starting in the 20s. And second, the difference in heat performance predicted by a gap of 1 place in qualifying shrinks quickly, leading to a relatively even middle of the pack, especially for the men. The fact that the women’s line does not flatten out to the same extent also suggests that there might be fewer women in the middle of the pack in Snowboard Cross than men. For a National Governing Body (NGB), this represents an opportunity to develop more female snowboard racers to draw into a less competitive field. 3 Olympic medals are on the line in Snowboard Cross. 2 more go out in Parallel Giant Slalom, and there are even more than that available in the Snowboard Cross and the Banked Slalom at the Paralympics. For NGBs, there is a significant opportunity to produce a return on that investment.

CONCLUSION

            We return to our initial question: are time trialing, as in qualifying, and head-to-head racing, as in heats, significantly different disciplines from each other? Yes, they are different. And not only are they different, head-to-head racing is meaningfully harder than time trialing. The fact that some athletes can be better at one discipline than they are at the other shows that the two disciplines cannot be the same, and the fact that so many more athletes are better time trialers than better head-to-head racers indicates that head-to-head racing is a far rarer skill. But the relationship between the two skills is still very strong. There are large gaps in head-to-head ability at the top of the qualifying field, followed by an evening out in the middle. That evening out is less significant among women, suggesting the possibility of a weaker field on the women’s side.

            In part 2, we will look at race conditions and how they impact different types of riders. Over the rest of this series, we will be looking at lane selection, scheduling, team-building strategy, and more attributes of Snowboard Cross, culminating in a real-time prediction model.