Tracking All NFL Over/Under Bets The Last 5 Years

One of the most popular gambling tips I have heard during my time in the industry is “when in doubt, bet the under.”

• The general public doesn’t like to bet under.
• The human brain is hardwired to bet the over.
• Sportsbooks shade the lines in favor of the under.

If I know that sportsbooks shade the under, everyone else must know that sportsbooks shade the under.

Therefore it’s fair to wonder: is there actually any advantage in betting NFL unders?

Our boy @brenttsky delivers with a solid graphic and a 40-year-old band reference that has me flashing back to my record album collection of the early 1980s.

Data_Josh is back.

We dumped five years of NFL data onto his desk and asked him specifically:

Can we make a living betting NFL unders?

Below is what we found out.

Let’s go!

Breaking It Down By Team

The NFL is made up of eight different divisions, with four teams in each division. There are 32 total teams.

If we analyzed the O/U record for all 32 teams over the last five years, I would expect about half the teams (16) to finish with more overs and half the teams to finish with more unders.

But that is not what we found…

Our graph above shows us that over the last five years of regular season games, 22 teams have had at least 51% of their games (or more) finish under the total.

That means only ten teams (31%) have finished with more overs than unders during the last five years.

Instead of a 50/50 split, it’s actually a 70/30 split in favor of the under.

70/30 sounds like kinda a big deal.

Shout-out to Denver and Pittsburgh, who both had 60% of their games finish under the total since 2017.

If you were a $100 bettor and had blindly backed the under for all Steelers and Broncos games during that time, you would be up $2,509.

When To Bet Over / When To Bet Under

We collected closing line totals for 1,288 games over the last five years. Of those 1,288 games:

• 618 games finished over the total (48%)
• 670 games finished under the total (52%)

But what if we break down those numbers even more?

Over the last five years, only looking at games played from Week 1 through Week 9, we learned that 50.5% of the games finished under.

If we analyze only the games played from Week 10 until Week 17, we see that 54.3% of the games finished under.

In fact, the under outperformed the over in the second half of the season four of the five years we analyzed.

This tells us that unders have been more likely to occur as the season goes on and that our greatest edge as a bettor was in the second-half of the schedule.

Do Winning Teams Cash More Overs?

So far, we have learned that almost 70% of NFL teams have more unders than overs during the last five years.

We have also learned that from Week 10 onwards, unders have cashed at a rate of 54.3%.

However, teams who win the most games must score a lot of points, and therefore, cash a lot of over tickets.

Or at least that is our working hypothesis anyway.

Our table above highlights the five best teams based on regular season win percentages over the last five years.

Those five teams have a combined 68.8% winning percentage but we can see that three of those five teams still had more unders than overs during those five years.

The numbers would be much worse if it weren’t for the New England Patriots and their 59% over rate.

It turns out that even the teams who have won the most games over the last five years didn’t payout on over bets.

Finding Week 1 Value

Whether you want to bet an over or an under, it always makes sense to play the Best Odds available.

When it comes to totals betting, you can often shop around and find an extra half point.

We have researched the Regular Season Week 1 lines currently posted and identified the few games where a whole point opportunity still exists.

Getting an extra point is always helpful, whether we are betting a total straight up, or teasing a total up or down.

By choosing the Best Odds available and taking advantage of extra points when they exist, we give ourselves the very best chance to beat the sportsbook.

And beating the sportsbook is why we spend so much time in the data-science lab.