Businesses may often have doubts regarding analytics and if they can be feasibly applied to their industry or particular company. Many years ago, the same doubt existed over whether or not analytical metrics had a place in the sporting world. Even before the movie Moneyball popularized the topic, the sporting world was adjusting to the new wealth of information available in the digital age. Currently the debate about their effectiveness is all but over, replaced by a debate about which metrics are the most appropriate measures in each sport.
Advanced analytical metrics aren’t necessary to tell sports executives that Sidney Crosby and LeBron James are star caliber players. But more and more, executives are turning to analytics to determine which non-star caliber players are the right fit for their team. In essence, does the player do the things which they believe are going to help their team win?
Two players with ostensibly similar offensive statistics can be shown to bring very different elements to a team through analysis of their advanced analytics statistics. A player may have strong offensive numbers but are those numbers due to playing alongside other higher caliber players? Are they inflated by playing against weaker opposition? Does this player generate higher offensive numbers at the expense of poor defensive habits? Are this player’s high offensive numbers due to unsustainable high levels of luck (is he due for a downturn in the near future)? Metrics are available to help decision-makers answer all those questions and many more.
Despite the gaps and growing pains in the sports analytics landscape (such as individual anomalies, the influence of luck or the impact of coaching tactics, among others), the correlation between high aggregated team metrics and success is quite striking. The source information for these metrics is often subjective (the game statistician is a person) and variance between a strong team and weak team may not always appear to be significant statistically. However when applied over a large enough sample size, analytic metrics are proving themselves to be accurate predictors of success.
The application of analytics in sports can be a simple way to demonstrate with public data that analytics can be factors in success. While they can’t predict who will win the championship each year – playoff series are too small of a sample size – they can help identify the group of teams that are in the best position to do so. The root of this analysis can even be connected back to algorithms which are typically applied in business, such as Market Basket Analysis.
If you think of a player as an item in a basket, the analysis from the above paragraph is quite similar to a business case. The business question “Is this product a strong seller on its own or due to pairing with other high volume products” may be reframed as “Are the player’s high offensive numbers due to playing alongside other higher caliber players”. Furthermore, the question “Is an item’s profitability higher than its own margin indicates due to it often being paired with other high margin items” is similar to asking “Are the player’s teammate’s offensive numbers positively impacted by playing with the player”.