King of the North and the Bruins: The Sport of Analytics

KingofNorth

What a wonderful week (June 10 2019) in sports and more importantly sports analytics. The Toronto Raptors basketball team which recently have been branding themselves as the ‘North’ were victorious in winning their first-ever NBA championship. Meanwhile in a game played a day earlier the St Louis Blues also won their first-ever NHL Stanley Cup defeating the Boston Bruins. Both series of games gives us some great data points to explore and perhaps see how analytics might have changed the outcomes.

Empty Net

I have written before about coaches’ reluctance to follow optimization strategies in gameplay decision making. Like the business world, it is not always about winning or having the best outcomes. Even when the statistics say so, sometimes it can be difficult to change peoples ‘tried and true’ approaches. Take, for example, hockey’s empty-net strategy.

The empty net strategy is when a team is behind in score, usually late in a game, they remove their goalie, adding an additional player on offence. This improves their chances to score and is an aggressive attempt to tie or win a game. The data shows historically teams do this within the last minute when trailing by one goal and within the last 90 seconds when down by two. However, analytics shows this is wrong. It is often far too late to make this gambit work. Further, often when the opposing team scores on the open net they immediately drop their strategy and put the goalie back in the game. Strategies like this are based on probabilities and require multiple trials aka ‘staying in the game’. Abandoning the strategy at the first negative event is like cashing out your retirement plan because of one bad day in the stock market. It is almost impossible to reach your expected return.

Which brings me back to the Bruins. In the final game, they found themselves down two points with eight minutes to go. Analytics suggest they should have used the strategy much earlier in the game. Although they did eventually pull the goalie with 3 minutes remaining it was not enough time. Clifford Asness and Aaron Brown’s work suggests pulling the goalie much earlier between six and four minutes remaining. They didn’t and the rest, as they say, is history. Ironically, the Bruins have been credited with inventing the empty net strategy back in 1931!

OptimalGoaliePull_ArnessBrown

Basketball’s Player Line Up Analysis

In Wayne Winston’s book Mathletics, he explores the analysis of different groupings of players and how they contribute statistically to game outcomes. And of course, he does this all in Excel, a tool nearly everyone can use and understand. In the NBA finals, the Warriors were faced with a similar challenge when one of their key players, Kevin Durant, was injured rupturing his Achilles. It is not clear how much analytics was used in re-shuffling their line-ups; however, the outcome was a close final game in which they lost to Toronto. What is interesting is how much the NBA is gathering data and how (potentially) it could be used for game-time data-driven decision making. The data is available on NBA.com and includes the combination of players in their Player Line Up statistics.

No matter if you are a sports fan or just a curious data scientist, sports analytics is a great place to explore and still a relatively new frontier for developing data-driven strategy. It is been 10 years since Moneyball began changing baseball, and surprisingly those lessons have been slow to be adopted by other sports. But the Analytics Revolution is happening, although slower in sports. Earlier this year the NFL began releasing more detailed football play by play data and beginning next year the NHL will do the same for hockey. More advanced data collection, IoT sensors, data scientists, and perhaps some growing interest from the sports managers can change game outcomes. Maybe the same will be said for business managers staring at their latest business metrics. Maybe they want to win too and they will ask “how can we be the king of the north”? And a data scientist will answer…

Note: Views are my own, not those of my employer.

#analyticsrevolution #datastrategy #AI #innovation #digital #technology #bigdata #tech#machinelearning #analytics #artificialintelligence #predictiveanalytics #Datascience #digitalrevolution #disruption #Moneyball #4thindustrialrevolution #sportsanalytics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s