Super Bowl Spreads – The Power of Actionable InsightsFebruary 05, 2016
Over the course of this week, our team has been scrutinizing the Super Bowl betting lines, craving an answer to the question: “Can social media help predict when betting lines are about to change?”
After all, the Super Bowl is THE most bet-on sporting event in America – and this year’s Big Game is projected to rake in $4.2 billion in wagers.
We seized this opportunity to put predictive social media analytics to the test. Though many of today’s sports betters already salivate over a good algorithm for using advanced analytics to make smart wagers, it’s rumored that football gamblers are lagging behind in this arena. We opted to change that.
Because whether you be a gambler, a small business owner, or a big-time corporate guru, we believe that predictive analytics can help you strike gold. Here’s why.
You Can Find Best Times to Market.
Our initial Super Bowl experiment was created to find a relationship between social media activity and Westgate SuperBook line movements. The “big picture” idea was that by honing in a trend, you could better expect that trend to happen again and be the first to take action on it.
This isn’t just a key theme for sports betters – if you’re marketing for a team, you could potentially find a relationship between certain in-game moments and the level of exposure that your team receives online.
For instance, perhaps more people get chatty on Twitter whenever a tight end, versus a running back, makes a big play – resulting in a bigger appetite for your team’s tweets in these moments. Or maybe people are most active on Facebook within the first ten minutes, versus an hour, after a game ends.
Test this theory across all your social channels to build a precise, responsive, and multichannel strategy for special events. By understanding trends, and even nuances like this, you can make sure that no marketing effort goes wasted.
You Can Track New Customer Behaviors.
Whether they be in sports or retail, insurance or hospitality, it’s every brand manager’s dream to know the behavioral patterns of their consumers. Advanced analytics gets us one step closer to this. (Read about how we recently used analytics to predict politics here.)
In our Super Bowl experiment, we made sure to find a “benchmark” number representing the volume of social activity that each team normally receives. By having this number, we were able to tell whenever there was an unusual spike or dip in either team’s social performance.
Likewise, your brand could benefit from defining “baseline” behavioral patterns for your different customer segments—and comparing daily, real-time data against these to pick up on unusual behaviors (both negative and positive) the moment they happen.
Imagine this: you’re a retailer and one day you log onto our platform to track all the social media posts referring to fashion, only to find that your audience in New York City is suddenly talking much more about fall fashion, a certain trend, or a certain designer than usual.
Do these spikes in social activity correlate to higher sales? If so, you now not only know when to sell, but what to sell.
With such actionable insights, your brand can reach new heights—something that became very apparent as we witnessed the power of analytics in our Super Bowl experiment.
So. What Happened in the Super Bowl Experiment?
Ah, yes. In case you needed a refresher, our experiment sought to find a relationship between Westgate SuperBook line movements and the Panthers’ and Broncos’ social media performances. We were inspired by this theory:
“The more that people are talking about a team on social media, the more that people are placing bets on that team. This should then forecast spread movement against that team up until a certain point, since bookies are [most interested in hedging their risk and maximizing their profits.]”
Though we do not encourage betting, we wanted to see if we could crack the code behind the way some bookies manage their spreads.
1. In general, the Westgate line tends to move up when the Panthers experience more social activity than the Broncos and an increase in activity from their benchmark score.
a. On Jan. 25th, Panthers had 30% more overall activity (232% increase from benchmark). Line went up .5.
b. On Jan. 27th, Panthers had 9% more activity (7% increase from benchmark). Line moved up 0.5 twice for a total of + 1 point.
c. On Feb. 2nd, Panthers had 1% more activity (18% increase from benchmark). Line went up .5.
2. There seems to be a stronger correlation between Panthers social activity and upward line movements vs. Broncos social activity and line movements.
3. On Feb. 3rd we saw downward line movement. We suspect that bookies lowered the line to compensate for an inflated gap, caused by the different trajectories in overall social activity around each team(up for Panthers, down for Broncos). Previously, if one team saw a decrease (-) in activity from their benchmark score, the other team usually did too.
4. On all days except for two, the Panthers had overall higher total activity than the Broncos; in aggregate, Panthers generated much more social activity around their team.
Analytics can help you “read between the lines” in something as unpredictable as gambling lines. Though audience behaviors aren’t exact science, and you can’t ever be 100% accurate with your predictions, you can certainly get closer than ever before with advanced analytics.