College football bowl season is upon us! And to make the multitude of odd bowl games involving teams we haven’t watched all season more entertaining, I’m sure you’ve filled out a Bowl Pool or two. There are 35 Bowl Games, so that means you have to pick the winner of 35 bowl games. That’s quite a few decisions to make!
When making decisions, we usually perform some sort of data analysis to help us make the choice. A farmer might use design of experiments to determine which fertilizer yields the most growth for his crops. A dietician might use hypothesis testing to determine whether a particular diet aids in weight loss. But when it comes to something as simple as football games, do we use data to make our picks, or do we use...something else?
I propose that in situations when we’re choosing between teams we know almost nothing about, we tend to lean towards the name that is cooler. Specifically, Toledo! After all, we’re just trying to make the bowl season more entertaining and what’s more fun than yelling “Holy Toledo!” each time the Rockets score a touchdown? When Toledo played in the Little Caesars Bowl two years ago, I tested my theory quickly in a text conversation with my friend Erik.
Me: Who do you like in that Toledo-Florida International game?
Erik: Holy Toledo! FIU stinks!
Me: Can you name one player on either team?
Erik: Gus Richardson
Erik: I made that up.
The point being, he just went with the cooler name, Toledo! So this year I’ve decided to put my theory to the test. I’m in 3 different bowl pools, and after removing myself I was able to get data for 64 picks in the Toledo-Utah St bowl game. Here are the results
- 18 people chose Toledo (yes, my friend Erik was one of them)
- 46 people chose Utah St
Toledo vs. Nevada
This destroys my theory right? Well, not quite. You see Utah St was a 10 point favorite over Toledo. So it’s no surprise that more people chose Utah St. These results show that people were using data to make their selections. But I’m not ready to give up. The Toledo-Utah St game occurred the same day as the Arizona-Nevada game. In that game, Arizona was a 9 point favorite. So if people were using the spreads to make their decisions, it stands to reason that about the same proportion of people that picked Toledo should also have picked Nevada (with even a slight lean towards Nevada). If a significantly greater proportion of people picked Toledo, then there might be something to my theory. I’ll use Minitab’s 2 Proportions analysis to analyze the data.
We see that 28.1% of the entries picked Toledo, while only 12.5% picked Nevada. The p-value is 0.013, which is less than 0.05, so we can conclude that this difference is statistically significant.
Toledo vs. SMU
Southern Methodist is a 12.5-point underdog in a game where nobody knows much about the two teams. How does Toledo compare to them?
NOTE: For the one pool I’m in, I can’t view everybody’s selections until after the game has been played. Since SMU hadn't played their bowl game at the time of this writing, their sample size will be smaller.
More people picked Toledo again! The p-value is right at 0.05, but since we’re only dealing with football picks here, we’ll say this difference is statistically significant too.
Toledo vs. Notre Dame
Now how about a game that everybody should know something about, the BCS championship game. Notre Dame is a 10-point underdog, so let’s compare Toledo’s proportion to Notre Dame.
Yet again, the proportion for Toledo is higher. This time we can’t conclude that the difference is statistically significant (at least at the alpha = 0.05 level), but we have shown that in similar games, more people are picking Toledo than any other underdog.
So did the same thing happen in your bowl pool? Let me know! It’s possible my family, friends, and co-workers just like the name Toledo more than the rest of the country. After all, even after Toledo was crushed by Utah St. in their bowl game and he read about my theory, Erik still didn’t regret his decision.
Erik: I'm sticking by my choice to pick Toledo, for now and forever.