We are well into the post-Moneyball era of deep sports analytics, and data nerds have become an integral part of professional sports. Broadly speaking, teams use analytics to improve their records and win championships, and fans use analytics to win fantasy sports leagues and win bets.
Despite the proven value of analytics in other sports, there are few analyses focused on professional surfing. There’s interest in predicting who wins each event (see the pre-event predictions offered by Stab magazine, Surfline, Surfer), but these tend to be based on a superficial look at recent performance and expert opinion. Others have gone the opposite direction and made attempts to use machine learning or algorithmic approaches to predict surfer performance, but these seem less well suited to surfing than they may be to other sports. Surfing is not a data heavy sport—there are only 10-11 events on the pro tour each year, with 36 surfers per event, and surfers are given a score for their two best waves per heat. Machine learning models and predictive algorithms perform best when they have enormous amounts of data to be trained on, and surfing just doesn’t have enough data to make for reliable predictive models. Furthermore, surfing is cursed with more unpredictability than other sports. The quality and size of waves change throughout the course of each event, favoring different surfers and styles of surfing. The way points are awarded is not completely objective; judges award surfers points based on their subjective assessment of each ride, and these opinions are often controversial.
However, the lack of data and inherent subjectivity that make predicting surfing a bad candidate for algorithmic solutions make it a good candidate for information visualization (or viz). Good viz enables the rapid exploration of large datasets, helping people find trends and relationships that might otherwise be unnoticed. Pro surfing is run by the World Surf League (WSL), which puts on an eleven event tour, with each event featuring 36 surfers surfing heats against each other and moving forward or being eliminated in a bracket-based format, sort of like March Madness. Surfer magazine, the bible of the sport, runs a fantasy league called Fantasy Surfer. Like other fantasy sports leagues, Fantasy Surfer gives you a budget to purchase athletes for your team. These players earn points for their performance at each event, and your goal is to earn as many points as possible within your budget. You can add or drop players between events, but can’t trade once the event is in progress.
I enjoy data analysis and developing visualizations, but I usually describe things that have happened in the past rather than trying to make predictions about what will happen in the future. Fantasy surfer gives me a bit of a playground to dive into data I’ve never worked with and see if I can develop viz that will aid myself in making predictions about the future. As it’s a league with public rankings, I can also compare my ability to pick teams against thousands of other players. It’s a good test case for me to develop visualizations and see if they’re “working”, ie helping me outperform other folks competing in Fantasy Surfer. I’ll be treating this blog like a sandbox where I can develop and refine visualizations, and discuss both the technical details and professional surfing.