Sports BettingNFLData Analysis
Vegas vs Reality: How Accurate Are NFL Betting Lines?
Exploring 45 years of NFL betting data to see if Vegas really knows best.
I have 45 years of NFL betting data with spreads and over/under lines. How accurate is Vegas at predicting game totals?
Let me compare the average Vegas over/under line against actual scoring each season since 1979 when betting data begins.
Vegas is shockingly accurate. The predicted line (dashed) and actual scoring (solid) move almost in lockstep. They tracked the scoring explosion from 40 points per game in the 80s up to nearly 50 in 2020. The lines briefly diverged in the early 90s when defenses got aggressive, but Vegas quickly adjusted. Over 45 years, they're typically within 1-2 points of reality.
What about home field advantage? Has that been consistent over time?
Great question. Home field advantage is one of the most reliable phenomena in sports - or at least it used to be. Let me show you the full 60-year trend.
This is fascinating. Home teams won 55-57% of games throughout the 80s and 90s, but there's a clear downward trend starting around 2006. By 2019, it had dropped below 50%. And then look at 2020 - that massive crater is the COVID season.
The COVID year really stands out. Can you zoom in on recent years?
Here's the last decade with the 2020 COVID season highlighted.
In 2020, with empty or limited-capacity stadiums, home teams won just 47.3% of games - below a coin flip. The home margin went from +3 points to essentially zero. It's the strongest evidence we have that crowd noise actually matters. Interestingly, home advantage has partially recovered but remains below pre-COVID levels at around 51-52%.
What about the biggest upsets? When has Vegas been most wrong?
Let me find the games where massive underdogs won outright - teams favored by 15+ points who lost.
The 1993 Cincinnati Bengals hold the record - they beat the 49ers as 24-point underdogs. The Miami Dolphins appear three times: as 22.5-point dogs against the 2007 Patriots (ruining their perfect season bid), and as 21.5-point dogs against Dallas in 2019. Notice how many recent upsets involve the Jets - they beat the Chiefs as 19.5-point dogs in 2020 and the Rams as 17-point dogs that same year. Being bad enough to get huge spreads but still capable of winning? That's a rare skill.
So what's the takeaway - can you beat Vegas?
The data suggests Vegas is remarkably good at their job. Over/under lines track reality almost perfectly. Spreads are calibrated so that favorites cover roughly 50% of the time. The only consistent edge? Home field advantage is smaller than the 3 points Vegas typically assigns - especially post-COVID. If you're betting, look for road favorites where the spread might be inflated by an outdated assumption about home field.
4 visualizations generated with Dolex