On a fall Saturday afternoon, a group of kids are wrapping up their summer soccer league. As a treat, one of the parents brings cookies for the group.
Suddenly, one of the children begins struggling to breathe.
There were peanuts in the cookies, and the child is having an allergic reaction. The coach yells for an EpiPen, but nobody has one, so she dials 911.
Three minutes later, a fire engine roars into the park’s parking lot, and a firefighter sprints to the child. As she squeaks and reaches out for help, the firefighter uses her auto-injector to deliver a life-saving dose of epinephrine.
Five minutes after that, with the child stabilizing and breathing again, an ambulance arrives to take him to the hospital for monitoring.
After the shock wore off, the coach wonders with gratitude why a fire engine was sent in the first place.
Dr. Laura Albert’s research
One of the reasons a fire engine would be sent is the statistical modeling research done at the University of Wisconsin - Madison by Mechanical Engineering faculty Dr. Laura Albert. According to the UW-Madison College of Engineering website, Albert researches “modeling and solving real-world discrete optimization problems with application to homeland security, disasters, emergency response, public services, and healthcare.”
The research on emergency response, for example, focuses on how to match the right resources with the right needs at the right time. In one aspect of this research, Albert looks at how to get the right mix of vehicles to an emergency.
Albert explained to B5Q in an interview last week, “Why would you send a fire engine to a health emergency? Well, It makes a lot of sense. They can be first responders and deliver the right care. They can treat people.”
For those of you who want a taste of how complicated this is (there is much more to it than my hypothetical situation in the opening section), check out the slides from Dr. Albert’s presentation at the University of Oklahoma last spring for a primer on her work.
Albert recently earned a National Science Foundation grant to study how to best use police resources to route people to recovery instead of the criminal justice system as well as disrupt opioid supply chains.
“The challenging thread going through that research is, wow, there is all this extra work for police to do and there are no extra resources,” Albert said.
In addition to Albert’s primary work in the School of Engineering, she also uses her skills in statistical modeling towards sports analytics.
Albert got into sports analytics because she wanted to learn more about statistical methods and apply them to cyber-security. “It’s hard to get real data, and I wanted to learn a little more about statistics but also bring it into the classroom and talk about application.”
Albert uses approaches such as Markov chains to not only rank teams where they are but also predict where they could be ranked at the end of the season. If you are wondering, she believes that the playoff committee got it right last season. In fact, she noted how impressive it is that that human rankings are often very similar to the most reliable computer rankings, such as her model.
New #BadgerBracketology Collge Football rankings:— Badger Bracketology (@BadgerBrackets) September 30, 2019
1 Ohio St
7 Penn St
10 Notre Dame#NCAAF #CFBplayoff #Analytics https://t.co/2XxmA1tKhb
While she prefers to not specify who, she consults with UW coaches to support their use of analytics.
Particularly, Albert’s analytics help coaches make decisions like what lineup combinations produce the best results and when to rest players. Not only can the mathematical models use lineups and rest time to optimize performance, but it can also be used to minimize the risk of injury.
Happy Selection Sunday, Badgers! Engineering prof and bracketology expert Laura Albert McLay puts her mathematical spin on March Madness.Posted by University of Wisconsin-Madison on Sunday, March 13, 2016
Sports and society’s problems
One of Albert’s most fascinating perspectives is how sports analytics can help solve society’s important problems. Sports are accessible for learners, and sports analytics are great tools to help teach the underlying concepts that generalize to addressing issues like police responses and the opioid crisis.
“If somebody learns a method and it all started because they wanted to win their fantasy football league, I think that’s the whole point of all of this,” Albert says.
“I say that as a professor who’s trying to educate the next generation to solve the problems we don’t even know about yet. If you look at all of these huge societal challenges, data engineering and analytics have a big role to play.”
The role of analytics in sports
Albert believes that analytics can help cut through human bias and get to what truly matters in sports.
“You can think about analytics as offering a bunch of different tools that can help your understanding (or not),” Albert said.
“It’s not like there’s one analytical model to rule them all. It’s interesting for me. I like looking at the different models. Even with all these different approaches, certain themes emerge.”
Then, when people start diving deeper into these models, it brings up questions like, “why does this model favor Alabama and this one favors Clemson?” These questions, Albert argues, help motivate learners to gain a deeper understanding of the analytics.
In the world of sports media, there are some who scoff at analytics. “I don’t understand the skepticism at all,” Albert said, “I look at the world, and I use math in my life every day.”
However, Albert, who plays sports, appreciates the ability to not over-quantify athletics and live in the moment as an athlete. “When I’m running, sometimes [I go] by feel and not just obsess about what my watch says all the time about my workout.”