Dr. Bruce Reinig is one of two professors to teach multiple classes in the SDSU sports MBA program: Operations and Supply Chain Management in the summer and Business Analytics in the fall. Reinig, the 2012 sports MBA Faculty of the Year honoree, takes the baton from Dr. Lackritz to teach real-life applications of statistical concepts. In this professor interview, Reinig discusses the material taught in his classes, a few of his favorite sports-related studies and why people should never gamble on sports.
Bruce Reinig: The role that statistics plays in the scientific method; hypothesis testing in particular. I think that of all the subjects that are taught at a university, the scientific method is the most important. To advance as a civilization I think the road is through science, and statistics plays an important role in that process.
MS: What do you like about teaching in the sports MBA program?
BR: I genuinely like the students. The students are a diverse set from all over the country and the world really, and the fact that the students are part of an ongoing cohort is nice. Their groups appear quite close with each other, and it tends to be a dedicated group with a real common interest in entering the sports business. There are real synergies created by bringing this group of students together.
The program is also well-suited for San Diego State; it is an area that we have a competitive advantage. It is a program that SDSU can offer that our competitors cannot easily replicate because we have a lot of resources around us that can serve students’ interest in sports management. A high ranking university in a smaller community would have a tough time competing with us because of the resources available in San Diego County.
MS: Going back to our first class, what do you hope students get out of the Operations and Supply Chain Management course?
BR: I would like students to be able to make better business decisions and do so by modeling the tradeoffs that naturally occur when you confront real world decisions. I especially like forecasting because it’s critical to every business, and students are often able to incorporate the knowledge into their work when they graduate.
MS: What makes forecasting such a critical concept for MBA students to have in their toolbox coming out of school?
BR: It affects virtually every decision in the organization. If you have a business that you’re going to open tomorrow, then your staffing and your inventory all rest on your forecasting, what you believe demand’s going to be.
MS: We spent a day doing an SAP simulation. What do you like about that bringing together many of the concepts we learned in class?
BR: The SAP simulation gives students a sense of the various types of information that are needed to effectively manage inventory. Students need to make estimations of industry demand and figure out which of their products are selling and decide how much they should invest in marketing. Participants in the simulation have to bring multiple information resources to bear in a very short time period, and you get to see how technology can help enable that process.
MS: Business Analytics is a new class to SMBA. Why do you think it was an important addition to our curriculum?
BR: I think the skill set in the class is important to graduating MBA students. Employers are continually looking for more analytical skills, and among MBAs there’s a higher expectation placed on you entering the job market in terms of what your analytical skills are going to be. I see the class as really helping the students improve in the workplace.
MS: Why do you love logistic regression so much?
BR: I like logistic regression because it allows us to model the probability that some particular event will or will not occur. There’s a certain amount of beauty in how the model works. It goes back to the role of science in modeling in that you can see what factors actually influence whether or not some outcome occurs and you can separate that out. It’s the ability to address a whole class of problems that multiple regression is unable to address.
MS: What are some of the tools from this class you hope students take to their future jobs?
BR: I hope that students can break down a dataset in Excel, analyze the data, and communicate that knowledge effectively to others in a presentation or report. I think virtually every student will be tasked with making data-driven decisions and the concepts taught in the course can help improve that process more than the specific tools. And I think MBAs need strong analytical skills to be competitive in the job market. This includes classification, prediction, forecasting, cluster analysis and all of the topics covered in the class.
MS: What’s your sports background? What teams do you root for?
BR: I love college football. It is played at such a high level today and anything can happen in the college game. So I like college football, the NFL, and I am less of a fan of baseball. I kind of had a falling out with MLB when they went on strike in ‘94.
MS: Oh yeah? Most people have come back since then.
BR: I have somewhat, but I’m not as big a fan as I was before. It’s an embarrassment of riches with the Cardinals doing so well. (Editor’s Note: Dr. Reinig is a St. Louis Cardinals fan.) I also love coaching soccer. I’ve coached rec soccer about 10 times now. For the last three years I’ve coached for the school team that my kids play on. That’s been a real joy for me. And my favorite sporting event is the World Cup Finals.
MS: What did your learn from your study on seeding the NCAA Tournament?
BR: That was an interesting study. We addressed the question as to whether or not the NCAA selection committee has any biases in their seeding, not their selection. Once they’ve made their selections, do they seed the teams in a fair and equitable manner? They say publicly they do so without any consideration or bias toward conference affiliation. The challenge then is to see how you can test such a claim; how can one go about seeing whether or not bias exists? The answer is to include conference affiliation in the model and demonstrate that it is not predictive of the seed.
In terms of the seeding, we found that poorly seeded mid-majors tended to perform better than their seeds would predict but it was also the case that mid-majors that received high-level seeds tended not to live up to those lofty expectations. The story came out a bit different than what we initially anticipated.
MS: What have you learned about NFL home-field advantage?
BR: Home-field advantage is a fascinating open-ended research question. The question is, ‘What is the nature of home-field advantage in NFL games or NCAA football games?’ I suspect that advantage manifests itself differently from one sport to the next, and I suspect that certain teams enjoy more of a home-field advantage than other teams. What I find interesting about the modeling I’ve done with football games is that if you ask people what the nature of home-field advantage is, the first response tends to be that their performance metrics will be better, and when you control for the performance metrics, in the analyses I’ve done the home-field advantage only seems to increase. So what is the nature of this home-field advantage? Why is the home team more likely to win other than for reasons such as they are more likely to pass for more yardage or something like that? When I watch college football games, it seems so often the game just lies on a handful of plays that could have gone either way, and I think that home-field advantage sometimes has an influence on those small number of plays, and there’s where it could decide the game.
MS: Why do you feel betting lines are such an efficient market?
BR: The betting line is dynamic and closely represents the margin of victory that has approximately equal number of dollars on each side. I think the real test of it is not just does a particular team cover the line half the time but the line itself, does the betting line predict the outcome of the game? It does remarkably well in the sense that the linear regression line has a slope that’s not statistically significantly different than one and a y-intercept that does not statistically significantly differ from zero. It cuts right through at a 45 diagonal. To me that’s powerful evidence, it’s setting a high bar. It explains less than half the variability in the final outcome, which tells us that there’s a lot of natural variation in these football games, but the market seems to be able to model through all that noise and capture that portion of variance that is predictable.
MS: And that’s why we should never gamble.
BR: That is correct, you should not gamble. You shouldn’t gamble because you have to bet 110 to win 100. You have to be correct 52.38 percent of the time to break even. So if you want to make money in gambling on football games, your gains start after you’ve hit 52.38 percent. To me that looks like a big threshold to cross before you get to your profits.