2019 Super Bowl Prop Bets: a data driven approach

IT IS SUPERBOWL TIME!

In less than 4 days, the two best teams (not to be controversial but Saints should be playing) in the NFL will face off in Atlanta for Super Bowl 53.  If you don’t already know this about me from some of my previous blogs and podcasts, I am from Boston and I am a huge Pats fan.  Each year, Cory, our fearless CEO tells me not to write about the Pats but as a loyal fan, I have to defy him.  This year will be no different but I will put a little twist on it. 

Since I haven’t said it yet… Go Pats, Tom Brady is the GOAT.

The Super Bowl is one of the biggest betting days of the year.  I am not really into betting hundreds or even thousands of dollars on a sports game but I do like to put a few dollars down on some of the prop bets.  For those of you who haven’t experienced the joyous world of prop betting, this is when you place bets on some of the more ancillary aspects of the game.  Things like: who is going to win the coin toss, which coach will the camera pan to first, what song will be sung first at halftime.  There can be some crazy ones like which color liquid will be poured over the head of the winning head coach and some can be pretty mundane like will the coin toss be heads or tails.   I’ve decided to choose a few of the more common prop bets, collect what I think is the right data and provide an analysis driven prediction for what you should bet on.

1.Who wins the coin toss?

This is probably the most popular prop bet out there and in all reality its based on a flip of coin. To make my prediction I used historical coin toss data from the past 52 years. The data source I found had a couple holes that I thought would be impactful for predicting the outcome. The data source had: Super Bowl number, teams playing, stadium, coin toss results and coin toss winner. What it didn’t have was which team actually called the toss. The home team, determined by year, (NFC for off, AFC for even) is the one that calls the toss. Obviously this is important as you will see from my analysis below. I had to manually add this to the data set to make it complete.

Analysis:

  • Rams are the home team this year and will be calling the coin toss
  • Rams are 3-0 in winning the coin toss but all three of those times, they won because the home team called it wrong.  This will be the first time the Rams have called the coin toss. 
  • The last time New England played the Rams, NE called the coin toss (heads) and lost
  • 46% of the time the team that calls the coin toss wins it
  • New England is 4-6 in winning the coin toss but 1-3 when not calling the coin toss

Prediction – Rams win the toss

2. Will the coin land on heads or tails?

This was an easy one since I used the same data source as before. Only addition I had to do was determine which stadiums had domes or retractable roofs

Analysis:

  • Over the 52 Super Bowls played, the coin has landed on heads 25 times and tails 27 times
  • When the Super Bowl is played in a domed or retractable roof stadium, the coin has landed on heads 9 times and tales 7 times
  • Honestly, this is a 50/50 decision

Prediction – The coin will land on tails… (common, tails never fails)

3. How long with the national anthem be? (Over/Under)

This year, Gladys Knight, the “empress of soul” will be singing the National Anthem the over/under in Vegas has been set for 1:47.  Covers had a great article that presented a clean data source I used to generate my analysis.

Analysis:

  • Over the last 20 years, the average time for the National Anthem has been 1:57
  • Only 15% of performances done over the past 20 years have been under 1:47
  • Gladys Knight is a Soul and R&B singer.  There have been 3 other Soul/R&B singers who have sung the National Anthem over the past 20 years (Alicia Keys, Aretha Franklin & Jennifer Hudson) and their average is 2:17.  This is by far the highest average by genre
  • You might be asking, why did Vegas set the O/U so low?  Well in 1991, Gladys Knight sang the National Anthem on a national stage and it clocked in at 1:40. 

Prediction – Take the over on this!  I would argue that the 1991 data is an outlier and not indicative of the Super Bowl.  She will be singing in her home town of Atlanta and its one of the biggest stages. 

4. Will the team that wins the kick off receive the ball or defer to the second half?

This was a hard one to get the data for. I had to go into the play-by-play for each of the last 15 Super Bowls and see which team kicked the ball and correlate that with who won the coin toss. It was extremely time consuming!

Analysis:

  • Looking at the last 10 Super Bowls, the team that has won the kick off has deferred to the second half (kicked off to start the game) 9 of 10 times
  • In the last 15 years, 60% of the teams that have won the coin toss have deferred to the second half
  • There has been a fundamental shift if coaching over the last decade or so that revolves around getting you defense on the field.  Belichick (the greatest coach of all time) is well known to always defer to the second half

Prediction – The team that wins the coin toss will defer to the second half and kick off to start the game

5. Will Tom Brady complete his first pass?

Another tough one to get the data for since there isn’t anything out that that captures this data in a logical format. Once again, I had to go to the play-by-play for each game and find Brady’s first possession and pass play and see if it was complete or not.

Analysis:

  • For the regular 2018/2019 season, Tom Brady’s pass completion was 66.5%
  • In the regular season, Brady completed his first pass 10 of 16 games which is 63%
  • In Brady’s past 8 Super Bowls (yeah that’s right 8 Super Bowls… #GOAT) 50% of the time he has complete his first pass of the game
  • Brady has thrown the ball to start the game all 8 Super Bowls
  • In his past 4 Super Bowls, he is 1 of 4 in completing his first pass
  • The Rams have given up a complete first pass in all three of their Super Bowls

Prediction – Brady will complete his first pass of the game

6.How many total points will there be for the Super Bowl?

This one wasnt too tricky to do since there is a lot of data sources out that that have points information for the Super Bowl. I downloaded a table from ESPN and added the points for each team.

Analysis:

  • The average total amount of points in all 52 of the previous Super Bowls is 46
  • In the last 10 years, the average total score is 50
  • In Super Bowls where the Rams and Patriots have participated in the average is 50
  • In the Tom Brady era, the Patriots have averaged 25 points per a game
  • Rams have never been to a Super Bowl under Goff but they have averaged 31 points per a game against playoff teams in the regular season
  • The Patriots have averaged 36 points per a game against playoff teams in the regular season

Prediction – 53 Points.  This will be a higher scoring game than the 10-year average but not by much

7. Who will win the game?

Easiest one of them all!

Analysis:

  • I will never go against my beloved Patriots.  Tom Brady is the GOAT, Bill Belichick is the greatest QB of all time…

Prediction – No Brainer…PATRIOTS WIN!

This is my first annual Super Bowl prop bet prediction blog.  Let’s see how accurate I am and what new data sources I need to incorporate to make more accurate decisions next year.   My biggest takeaway is one that we all know.  Knowing what data to use, collecting the data, cleaning the data is CHALLENGING and TIME CONSUMING.  To give an example, to figure out Brady’s first pass completion, I had to go to the play-by-play for every game.  For this little project, it was manageable but a pain but at scale it becomes impossible. 

Check back on Monday and let me know how you did?  Disclaimer, if you bet with me and lose, Big Data Beard is not responsible but you can call me out. 

Enjoy the Super Bowl, Go Patriots!

Your Bearded Friend, Brett

** Disclaimer: We at Big Data Beard Inc. do not condone nor endorse illegal gambling. This is strictly for entertainment and data purposes. Big Data Beard Inc. is not responsible for any losses of any nature.