Data Power Play: Our Take on the Stanley Cup Finals
This blog is meant to be a fun and unique take on predicting which team will win the Stanley Cup 2019.
The Stanley Cup finals begin tonight, and we at DataRobot are curious about the chances of our home team, the Bruins, to win over the St. Louis Blues. Following a similar methodology to Zach Mayer’s March Madness predictions, we built a model to predict who would win any given game in the playoffs. Using this model, we then simulated the seven-game Stanley Cup Finals between St. Louis and Boston.
In this simulation, the Boston Bruins beat the St. Louis Blues 62% of the time.
Bruins & St. Louis Blues (Shutterstock)
To get this prediction, we used historical data on every NHL season since 1980. From this data, we developed a couple of proprietary ranking systems using DataRobot to predict how a team would do during the regular season. In addition, we calculated an ELO rating for each team. This rating provides a way to estimate the relative quality of the team based on their results. Without any other information, a team who’s ELO rating is 100 points higher than another team should win 64% of the time. DataRobot’s models used this information and then effectively learned how to enhance this probability based on other factors.
Based on the performance of each team (shot, save, penalty kill, and power play percentages, along with these various rating systems), we trained DataRobot on playoff games from 1980 through 2018 to learn the probability that a given team would win a game based on its performance, the opponents performance, and who was at home.
Once we had this model to predict who would win a match, we simulated 100,000 Stanley Cup Playoffs for 2019 to see who would win most frequently. Finally, once we were comfortable with both the win probability model and the simulation, we simulated the Stanley Cup Finals between Boston, St. Louis, and San Jose.
To be honest, we were rooting for San Jose in the Western Conference Championship as this same simulation suggested Boston would have a slightly better chance against the Sharks (64% vs 62%).
We next considered the path both Boston and St. Louis took to get here. To start with, neither team was favored when the playoffs began. Based on the Tampa Bay Lightning’s historic season, they were the overwhelmingly favorite to win the Stanley Cup 54% of the time. Boston was a distant third, with only a 6% chance of claiming the Stanley Cup.
However, when not only Tampa Bay but Calgary (with a 13% chance of winning the Cup) and the other two division winners all lost to wild-cards (for the first time in the history of the Stanley Cup playoffs), Boston became the remaining favorite. To understand how big of an upset both of these results were, our model suggested Tampa Bay would win the opening series 86% of the time and Calgary 79% of the time.
Boston handled Toronto (a result our model predicted would happen 66% of the time), Columbus (likely to happen 61% of the time), and Carolina in the Conference Finals (63% of the time).
The Bruins opponents in the Stanley Cup Final navigated a tougher path to where they are today. St. Louis was a slight favorite (54%) in its victory over Winnipeg before becoming the favorite (63%) over Dallas that propelled it to the Western Conference Finals. In the Conference Finals, St. Louis was predicted to beat San Jose 51% of the time.
The Stanley Cup (Shutterstock)
It’s been an exciting season with the historic wild-card wins and with the Bruins starting with only a 6% chance of taking the Stanley Cup. These unexpected events are what makes sports analytics exciting and fun, because the course of an entire season can change drastically at any time. The Bruins are in a good position now and after crunching the numbers and sending out some good thoughts, we’re ready to see what unfolds. Best of luck to our home team, the Boston Bruins!
And just in case you’re thinking that DataRobot is headquartered in Boston so, of course, we picked the Bruins, all city bias has been removed from the models! We are not homers — we are data scientists.
Interested in more Sports Analytics? Visit our Sports Analytics solutions page for more content and insights.
Andrew Engel is General Manager for Sports and Gaming at DataRobot. He works with DataRobot customers across sports and casinos, including several Major League Baseball, National Basketball League and National Hockey League teams. He has been working as a data scientist and leading teams of data scientists for over ten years in a wide variety of domains from fraud prediction to marketing analytics. Andrew received his Ph.D. in Systems and Industrial Engineering with a focus on optimization and stochastic modeling. He has worked for Towson University, SAS Institute, the US Navy, Websense (now ForcePoint), Stics, and HP before joining DataRobot in February of 2016.
We will contact you shortly
We’re almost there! These are the next steps:
- Look out for an email from DataRobot with a subject line: Your Subscription Confirmation.
- Click the confirmation link to approve your consent.
- Done! You have now opted to receive communications about DataRobot’s products and services.
Didn’t receive the email? Please make sure to check your spam or junk folders.
Optimizing Large Language Model Performance with ONNX on DataRobot MLOpsJune 1, 2023· 11 min read
Belong @ DataRobot: AAPI Heritage Month with the ACTnow! CommunityMay 25, 2023· 3 min read
Deep Learning for Decision-Making Under UncertaintyMay 18, 2023· 5 min read
Discover insights on the specific conditions that make machine learning effective in certain financial applications, such as high-frequency trading. Read more.
In this article, we’ll first take a closer look at the concept of Real Estate Data Intelligence and the potential of AI to become a game changer in this niche.
In this blog post we’ll explore how Mindshare, a global media agency network, has leveraged data science tools to create a fast and reliable decision-making engine. Read more.