Once the lights went out on the Conference Championships, books hustled to the line to try to be the first to get their prop menus in order. The Super Bowl game props were some of the first ones posted. But is there value here?
Maybe you’ve kicked the tires on “race to X points” or “margin of victory” plays and found them lacking. You’re not crazy.
Game props used to be a great source of value. It’s a market that has tightened up in recent years, though, Rufus Peabody said.
“It really shifted big time with legalization. Sportsbooks like DraftKings and FanDuel started putting out props every week,” Peabody said. “That catalyzed market efficiency. It sped it up.”
The reason, Peabody said, is that so many of these props can be solved algorithmically.
Analyzing Super Bowl Game Props with Basic Math
Here’s what we mean. Margin of victory is a fairly common game prop. Will the Niners win by 1-6 points? Will the Chiefs win by 7-12?
We break down how to calculate a fair price of exact margin of victory props in this article. The same method can be used to calculate these wider margins.
The short version is: use the Alternate Lines Calculator to find the fair prices on San Francisco -2.5 and San Francisco -3.5. Convert them to implied probabilities (you can use the Odds Converter) and subtract them. In this case Niners -2.5 (49%) minus Niners -3.5 (40.8%) equals a fair probability of 8.2%.
Repeat that process for every other number from 1-6, add up the probabilities and convert them to a price. Fair on the Niners by 1-6 is around +386. We’re only getting +305, so this is a pass.
As with many game props, once we have the spread and total, we can work out a whole lot more. But so can the books. What’s easy for the books is usually bad for us.
What to Avoid in Super Bowl Game Props
A clue that we weren’t going to find much value in those margin props was that it was a multi-way market. Multi-way game prop markets hold more, which makes them harder to beat. Consider this one, which you can find at several books: Which team will score first, and what will the final result be?
To solve this, you could start by looking at several years’ worth of games lined -2 with a total of 47.5 to see how often the favorite scored first and won the game.
But before you go to that trouble, you might want to check the hold in this market using our Hold Calculator to see the 11.33 percent house edge makes this a tough nut to crack unless there’s a wild misprice on any one of those lines.
There are plenty of these multi-way markets among the game props. Before you bet into them, check to make sure you’re not trying to overcome a significant hold.
Find New Frontiers
Your best chance of finding a misprice is to look for bets that books don’t normally offer. Margin of victory has become a common offering for books every week of the NFL season. But there are plenty of bets on the menu that only come out once a year.
On “Will this thing happen?” type props, the value tends to be on the “no.”
The public likes betting on things to happen. Your advantage lies in gobbling up the better price when the weight is on the other side of the ledger.
If you’re going to find value in these, it’s likely going to come from no safety, no overtime, no both teams won’t score a touchdown and field goal in the first half, and so on.
Using Logistic Regression to Beat Super Bowl Game Props
Peabody’s main tool for analyzing game props is to run regressions on the data.
A logistic regression is a tool for looking at a dependent variable and using independent variables to inform the answer.
A dependent variable is a binary outcome we’re trying to solve for. Whether or not the game will go to overtime is a dependent variable. Independent variables are all the things that affect the yes or the no. In this case, the spread and the total are two key independent variables. They’re a big part of solving the question of whether the game is headed to OT.
“‘Will the game go to overtime’ is a function of the spread and then our mean expectation and the variance around that, which is a function of the total. ‘Will there be a score in the last three minutes of the fourth quarter?’ is a function of the total, and then the absolute spread.
“If it’s a blowout, it’s less likely that you have a score in the last three minutes of the fourth quarter than if it’s a really close game. So, the more likely it is to be a blowout, the more likely you are to not have a score in the last three minutes of the fourth quarter. I priced the ‘yes’ at -107. If it was a 12-point spread, I’d price the ‘no’ at -120.”
You’ll need to do a bit of work to source the data (NFLfastR is one spot if you’re comfortable using the R language) and figure out which independent variables are important to each prop.
But what about if you don’t have the time or inclination to wade that far into the data?
Figure Out What’s Not Priced In
“There are opportunities,” Peabody said, “if you think there’s something in these algorithmic props that, is maybe team-specific that wouldn’t be accounted for in the spread in total.”
For example, you may want to know whether a team will come back from a double-digit deficit to win. A regression using the spread and total might tell you a baseline number.
But if you think the Chiefs are better poised to do it because they have an explosive offense and Patrick Mahomes, you might make your own adjustments to the price based on that.
“I don’t get into that level of detail, but the point is that the level of detail in the pricing of these props is not all that sophisticated,” Peabody said. “It’s the rule, but if you can find the exceptions to it, that’s where someone can find some value, potentially, without doing any math modeling.”
Got more questions heading into the home stretch? Join the #super-bowl channel in our Discord and fire away.
