This week I want to dig into a signal I’ve been observing over this season and how it could play out in Week 14. If you’ve been following me, you know I watch every snap of every game and grade every performance. The goal is to come up with a number that reflects true performance better than the final score or box score can. Other handicappers use a variety of metrics, such as yards per play, EPA/play, success rate, etc. All these are useful in their own way but have flaws which I try to address in my Rating system.
In theory, betting NFL sides can only be profitable if you have a better read than the market. Lines are hyper-efficient and reflect true market value by the end of the week. Bettors are results-driven, particularly by wins and losses. The reaction to last Monday’s Bills/Patriots game is a perfect example. That game truly could have gone either way, but books and handicappers upgraded the Patriots and downgraded the Bills because the Patriots won the game. Recency bias plays into the lines, so misleading results relative to true performance can create value for those paying attention.
So the simplified potential signals are:
- A team that wins a game with a lower Rating than its opponent should be overvalued by the market in its next game.
- A team that loses a game with a higher Rating than its opponent should be undervalued by the market in its next game.
Backtesting this, if you had blindly bet teams in these situations over the course of the season, here are the results:
- Teams coming off a deceiving win: 10-22 ATS in their following game.
- Teams coming off a deceiving loss: 19-11 ATS in their following game.
But digging a little deeper, the trend becomes extremely strong when you look only at games where a team that played significantly better than their opponent won the game outright. Team Ratings range from 0 to 10 but usually fall in the 3.0-7.0 range, clustered around 4.0-6.0. If you limit the sample size to games where the losing team had a Rating at least 0.3 points higher than the winning team, it becomes a stellar signal:
- Teams coming off a very deceiving win: 2-12 ATS in their following game.
- Teams coming off a very deceiving loss: 10-4 ATS in their following game.
Now, trends are not worth anything when not grounded in a legitimate angle. This one is. Teams that suffer a frustrating or undeserved loss not only come out of the game undervalued by the market, but they tend to play hard, with a chip on their shoulder. They also tend not to take situations for granted. The opposite is true for teams that get a fluky win.
So, what does that mean for Week 14? Let’s take a look at the most misleading results from Week 13.
DET 29 (4.83) – MIN 27 (5.36)
The delta of 0.57 between the Vikings and Lions created the most misleading win of the weekend. The Vikings moved the ball more effectively, converted third downs, and even won the turnover battle. But they failed too frequently in the red zone without their two best red zone weapons (Thielen and Cook). The epic choke on defense at the end made losers out of more than a handful of teaser bets on Minnesota. The signal says bet on MIN (which I already did) and fade DET against Denver.
NE 14 (4.37) – BUF 10 (4.8)
This was a weird game, and you never want to read too much into a weather game like this. But it’s worth noting how starkly my Ratings (delta of 0.43 in favor of Buffalo) contrast with the public narrative around this game. The Bills had multiple chances to win this game, driving down the field twice into the wind. But they were ultimately unable to finish. The Patriots got ahead on a breakaway run on 3rd and long, but the story probably would have been very different if they were unable to play with the lead all day. The signal says bet on BUF against Tampa Bay and fade NE after their bye against Indy next week.
SEA 30 (4.0) – SF 23 (4.36)
The 0.36 delta puts this result in the “very misleading” category by a small margin. Box score scouts might be tempted to say the Seahawks deserved the win because they had two lost fumbles and a missed field goal. Those are high variance events that went against Seattle. But the 49ers outplayed Seattle throughout the game, averaging 6.5 yards per play to Seattle’s 4.8. And when you take out Travis Homer’s fake punt run, that number drops to 3.8. Those types of plays are not predictable or repeatable. Calls went against the 49ers most of the day. More often than not, San Francisco wins this game. The signal says bet on SF against Cincinnati (although, there is a contrasting signal in that game) and fade SEA against Houston.
LAC 41 (5.0) – CIN 22 (5.14)
Although this one technically falls into the “minor misleading” category with a delta of 0.14, the disparity in score (19 points) and the public perception of what happened in this game make it a stronger signal than some tighter games. Two big plays defined this one: Ja’marr Chase bobbling what should have been a long TD and turning it into an interception, and the Chargers recovering a Mixon fumble for a TD. Outside of that, it was a fairly even matchup. Considering the Bengals were playing down two offensive linemen and Burrow hurt his finger, this was a strong showing by Cincinnati. The signal offsets the SF signal and suggests fading LAC against the Giants, if you can stomach it.
PIT 20 (4.78) – BAL 19 (4.89)
Not a huge signal here as this game was fairly even, but I did think Baltimore outplayed Pittsburgh by 0.11 and came away empty. The Baltimore angle is offset because Cleveland is coming off the same situation before its bye (where they outplayed the Ravens and lost). But it’s worth nothing that the Thursday Night tilt between Minnesota and Pittsburgh featured both trends, which is one reason I felt strongly about MIN -3. Fading PIT was the right move.
Of course, you always have to factor in more than one trend when making a bet. But if you find other angles you like, this trend may prove helpful in supplementing those angles. If this is helpful, I’ll update this weekly with ongoing signals.