SportsBook Breakers’ NFL Study

Beware of Low Winning Percentage Teams

It is easy to get sucked into betting on a bad team, but there are also many, many times when that makes sense.  In fact, when looking at all betting situations, in general you are better off betting on teams having losing seasons, as they cover in 50.8% games.  However, the key to betting on all teams, particularly those who are not of a playoff caliber, is value.  What we’ve uncovered is a situation where the value has been sucked completely dry.

We are curious about when non-elite, non-playoff caliber teams are given expectations they don’t usually face.  What happens when these teams are favored, and potentially as a significant favorite?

This is an easy subject to investigate with the power of the Sports Data Query Language (SDQL).  To explore the subject, we need to use just two parameters, “line” and “WP,” an easy shortcut for winning percentage.  For an easy and quick way to explore the subject, we will look at how teams perform in the SDQL using the grouping feature.  We defined the winning percentage to investigate as teams winning less than 62.5% of their games at the current time, the equivalent as a 10-win team, the number it generally takes to make the playoffs.  The SDQL text “WP<62.5 and line<0, -2, -3, -4, -6,-7, -9” produces the following result, grouped together by lines larger than the given number.  NOTE: Results date back to the beginning of the NFL database in 1989.

Line ATS SU # of Games
line < 0 1452-1578-88 (-0.12, 47.9%) 2023-1091-4 (4.77, 65.0%) 3118
line < -2 1212-1350-83 (-0.05, 47.3%) 1767-874-4 (5.45, 66.9%) 2645
line < -3 856-968-29 (0.06, 46.9%) 1317-533-3 (6.69, 71.2%) 1853
line < -4 644-755-26 (-0.04, 46.0%) 1037-385-3 (7.47, 72.9%) 1425
line < -6 392-522-21 (-0.38, 42.9%) 695-239-1 (8.30, 74.4%) 935
line < -7 240-343-11 (-0.69, 41.2%) 462-131-1 (9.07, 77.9%) 594
line < -9 131-174-6 (-0.80, 43.0%) 244-67-0 (10.44, 78.5%) 311

 

These results above are exactly what we like to see to back up such a hypothesis.  From the top, when looking at all favorites in this situation, they cover only 47.9% of time.  While that is not a beatable number in itself, considering it accounts for over 3,000 active instances, it is statistically significant. As the lines get larger, the results get steadily worse until reaching a play against point.  To find that exact point we use an open-ended parameter with the SDQL text “WP<62.5 and line

Line ATS SU # of Games
-5.0 51-54-3 (-0.69, 48.6%) 70-37-1 (4.31, 65.4%) 108
-5.5 65-58-0 (1.52, 52.8%) 89-33-1 (7.02, 73.0%) 123
-6.0 80-61-2 (1.58, 56.7%) 108-35-0 (7.58, 75.5%) 143
-6.5 65-87-0 (-1.28, 42.8%) 98-54-0 (5.22, 64.5%) 152
-7.0 87-92-10 (1.33, 48.6%) 135-54-0 (8.33, 71.4%) 189
-7.5 28-67-0 (-2.35, 29.5%) 67-28-0 (5.15, 70.5%) 95
-8.0 31-38-3 (1.22, 44.9%) 60-12-0 (9.22, 83.3%) 72
-8.5 19-31-0 (-0.60, 38.0%) 41-9-0 (7.90, 82.0%) 50
-9.0 31-33-2 (0.02, 48.4%) 50-15-1 (9.02, 76.9%) 66
-9.5 25-32-0 (-0.18, 43.9%) 41-16-0 (9.32, 71.9%) 57
-10.0 27-36-3 (-1.15, 42.9%) 51-15-0 (8.85, 77.3%) 66
 -10.5 17-29-0 (-1.28, 37.0%) 36-10-0 (9.22, 78.3%) 46

 

This data shows that the sweet spot for where this becomes a play against system is between 6.5 and 7.5 points.  We’ll play it conservatively here, taking the more than TD spreads.  Since 1989, teams that have won less than 62.5% of their games are more than TD-favorites are an underwhelming 240-343-11 ATS (SDQL: WP<62.5 and line<-7)

Other factors to consider:

Since winning percentage is a far more accurate measure later in the season than in the early weeks, it would seem this should make a different on the results.  When adding the SDQL parameter week, we find that this is not a major factor.  From weeks 2-4, when winning percentage is the least accurate, teams are 48-66-3 ATS (42.1%) in this spot.  From weeks 15-17, when winning percentage is best representation of a team’s ability, the system has gone 52-77-2 ATS (40.2%)

Looking at various winning percentages along this range, there is no significant different in the ATS result based on winning percentages below the 62.5% standard.

It is obviously quite rare for these teams to be road favorites of more than a TD, but when they are, the result is a brutal 21-43-2 ATS (32.2%) (SDQL: A and WP<62.5 and line<-7)

This system has not performed particularly well when facing a winning team.  While you might that that would be an even greater advantage to play against, these large favorites are actually 50-48-2 ATS against teams that have won at least half their games on the season.

This system is active most often when these average or worse teams are facing terrible teams and the results are quite juicy.  When facing a team that is winless, or has won no more than 10% of their games, these teams have gone 46-93-5 ATS (33.1%) (SDQL: line<-7 and WP<62.5 and o:WP<=10)

Summary:

When you are evaluating a team that does not normally play at an elite level, there is just too much that can go wrong to expect an elite performance, even when the matchup sets up well on paper.  Do not trust non-playoff caliber teams with big lines.