It's incredibly difficult to make money picking NFL games against the spread. As basketball handicapper Haralabos Voulgaris stated at the 2013 Sloan Sports Analytics Conference: "The ball's not even round."
As a betting proposition, the NFL presents many challenges, including the fact there simply aren't enough games. Compare a full NFL season (256 games) to a full Major League Baseball season (2,430) or NBA season (1,230). An advantage gambler will prefer a sport with more games for a few reasons.
First, you have a lot of trials and data with which to build a model. Second, you can be more selective in your plays, as there won't be value in every game. And finally, if you have a small advantage, you want as large a sample size as possible so that you can avoid short-term variance.
There is so much public information available about the NFL that it's almost impossible to have an information edge. Injury news becomes public information instantly via Twitter, and advanced analytics are widely written about and disseminated; they are often figured into the line. Even line moves in the NFL are confusing because so much money is bet on each game.
But as I've written before, everyone wants picks. So when ESPN asked me to do some "SportsCenter" appearances, I knew they would want me to give picks.
I suggested creating a narrative around the picks that would give the viewer more information than simply a pick on a game. These narratives would be based on real stats and analysis, and hopefully would be useful to those trying to pick NFL games against the spread.
Because my TV appearances are pretty quick, I don't get a chance to explain too much of what goes behind my picks. Here's what I look for when making them and where I believe the advantage lies in the NFL:
1. Find teams that are overrated
In the NFL, like other sports, teams tend to get overrated based on public opinion, formed on readily available numbers like won-loss record or power rankings. But often those types of numbers don't tell the whole truth. Advanced analytical systems like Massey-Peabody and Advanced Football Analytics capture more meaningful measures and provide a more predictive view of a team's real ability. For example, a team can have a reputation of giving up a lot of passing yards, but that may be because teams that are trailing are attempting a lot of passes against them.
For most of the early season, the Arizona Cardinals had the best record in the league and were at the top of many power rankings, even though most of the advanced analytics crew rated them as a mediocre team whose success was predicated on nonsustainable elements such as defensive touchdowns and turnover margin.
I had believed the Cardinals to be overrated all season, and for the most part they had proven me wrong. But in Week 12, I stood by my guns and predicted a Seattle win by more than a touchdown, and in Week 13 I thought they would fall to Atlanta. I don't think there was genius in these picks, but looking at their yards per play offensively and defensively showed a disconnect from their reputation as an elite team.
I see quarterback Drew Stanton having some big shoes to fill in terms of ball security. Carson Palmer's interception rate was among the lowest in the league and it's hard to see Stanton in his first real season as a starter being able to replicate that feat. I'll likely be looking for opportunities in the future to pick against the Cardinals.
This highlights another interesting and somewhat counterintuitive factor to consider: Turnover margin, more than any statistic (except obviously points), plays a large role in whether a team wins or loses, but past studies have shown that it is hard to predict turnovers based on past turnover rate. Interceptions have some predictive value, but fumbles have very little; they are a pretty random occurrence.
When a team's success is largely predicated on not turning the ball over or causing turnovers -- as the Cardinals' has been -- it is likely to become overrated because a lot of the reasons for its success are not sustainable.
2. Look for teams that are undervalued
The poster child this season for being undervalued is the Baltimore Ravens. At this point in the season most advanced metrics have them as an elite team, while their 8-5 record would lead you to believe they are simply above average. When they played New Orleans in Week 12, they opened as 4-point underdogs. I remarked on "SportsCenter" that this was a bad line because by advanced metrics, Baltimore was a better team than New Orleans. The line eventually went down to 3 points, but even at that line there was value on the Ravens.
In Week 13, I stuck by my assessment and the market agreed, as the Ravens' line versus the Chargers opened at minus-5.5 and rose to 6.5. I liked the Ravens in that game and while they played well for most of the game, they lost outright to the Chargers on a last-minute touchdown. This past weekend, they overcame a 10-0 deficit to cover as 3-point road underdogs to the Miami Dolphins. Injuries in the secondary have definitely made the Ravens vulnerable defensively, though. Couple that with the market's intelligence about their "value" and the underrated Ravens are no longer that. That's how quickly an advantage can disappear in the NFL.
3. Look for the effect of recency bias and other biases
I've previously written on ESPN Chalk about the biases that cloud our judgment, but I'll try to explain recency bias again here. In short, we all get over-influenced by recent events because they are hard to get out of our heads. When recent events are especially spectacular or horrific, they tend to influence public opinion a bit too much, and this at times can present opportunities.
Take the 24-3 beatdown the Cleveland Browns put on the Cincinnati Bengals in Week 10. Bengals quarterback Andy Dalton was the laughingstock of the league and Cleveland looked incredibly impressive, thoroughly dominating the Bengals in every phase of the game.
In Week 11, Cincinnati faced a tough road game versus the Saints, and that line, which opened at minus-7, ran up to minus-10 in some locations. Nobody believed the Bengals had a chance. They won 27-10.
Cleveland, a team that was seen as struggling before its game with Cincinnati, faced the Houston Texans in Week 11 and was installed as a 4.5-point favorite, when just two weeks before the Browns had barely beaten the hapless Buccaneers. The Browns lost 23-7.
Hindsight bias helps us look back on these games as a clear case of Cincinnati being underrated by recency bias and Cleveland being overrated. This is oversimplifying the situation, but as you look at games, one adage to live by is "you are never as good or as bad as your last game." Or even more simply put: Don't judge a team by one game.
4. Identify why spreads seem 'too good to be true'
This is somewhat a TV narrative, but we all see those lines that jump out at us because they seem "too good to be true." Normally, it's a small road favorite that seems sure to win the game. A perfect recent example was the Cowboys in Week 12 playing on the road against the New York Giants. That line opened at Cowboys minus-3 and was eventually bet up to Cowboys minus-4.5.
ESPN Chalk
The new ESPN Chalk section is now live. Check out betting coverage for all your favorite sports. Home page »
In these situations, there is always talk that "Vegas knows something" or that "Vegas is begging you to bet this team." I have never been an oddsmaker, so I can't speak to their motivations, but what I surmise from conversations and anecdotal stories is that oddsmakers try to balance the action they get from the public with the action they get from the professional bettors. While they don't mind being sided in a game (having uneven action), they don't want things to be too extreme.
There are times that the oddsmakers know they are going to get a lot of action on a certain side regardless of what their numbers, like this past Monday night with Green Bay facing Atlanta. When a team like the Packers is on a run like they have been (5-0-1 ATS at home, 7-of-9 ATS wins before Monday night), oddsmakers know that people are naturally inclined to bet on Green Bay. When they made that line, they knew they had to give up value on whomever Green Bay is playing to get bettors to bet the other side. In my eyes, 13.5 points was too many for Green Bay to be getting against a capable Atlanta offense (even if the Falcons' defense is atrocious).
In general, any time you see a line that you don't understand, my advice is to put yourself in the place of the oddsmakers and attempt to understand why they would make that line. In the Vegas world, if something looks too good to be true, it probably is -- and there is something that you are missing. They're not in the business of giving away free money.
5. You might have to kiss some frogs
There's an adage that you have to kiss a lot of frogs before you get a prince, and the same can be said for picking NFL games. When teams are really bad, like Oakland and Jacksonville have been this season, the oddsmakers know that they have to give up some value to get people to bet on these teams. This can often present some value to the bettor.
It's hard to reason betting on a really bad team, but that is figured into the line, and in the NFL the bad teams are usually not as bad as the narrative of their ineptitude. While the Raiders are an abysmal 2-11 this season straight-up, they are a winning 7-6 ATS. The same can't be said for the Jaguars, who are also 2-11 SU but a poor 3-9-1 ATS.
These are two teams that will present value to the gambler because linesmakers will not want to be overexposed on these terrible teams.
Picking NFL games against the spread is incredibly difficult, but hopefully the framework above can give you a hint to create your own methodologies. Assuming you aren't a quant jock yourself, the first thing I would do is educate yourself by reading some of the advanced analytical blogs like Advanced Football Analytics or Massey-Peabody. These are a good starting point to understanding the game outside of the common narrative.
Then I would try and look for opportunities like the ones mentioned above -- and yes, you will have to be contrarian at times. I would avoid piggybacking on big line moves, as a lot of the value in a game will likely be gone if the line has moved more than a point.
The NFL is the hardest sport to beat over time, but there are opportunities from time to time. A good starting framework and discipline will go a long way to improving your chances.