|
|
Preseason Picks, part III
Wagering on NFL preseason games may be seen as risky by many, but perhaps there are some bona fide ways to tackle the challenge.
Also see NFL Preseason Picks, part I where we looked at whether prior year regular season records had any bearing on preseason results, and NFL Preseason Research, part II which looked at the preseason records of teams in the current exbhition season coming into the game.
|
Preseason Picks research -- Two Year Team Trends
We've looked at preseason matchups using a number of different prior season stats, and found some useful tips. Perhaps though there is some benefit to looking at a slighly longer window of a team's prior regular season performances?
First off we will revisit the notion of the "G-A-P" grading system, which for reference we classified as:
- Good -- teams with a 10 or more wins in the prior regular season
- Average -- teams with 7 to 9 wins
- Poor -- teams with less than 7 wins
Now, rather than look at one team's status against another, we will simply focus on the two year trend for a given side:
Preseason results against the spread (1998 to 2005)
| Last Yr |
2 Yrs Ago |
Home
Favorite |
Home
Underdog |
Away
Underdog |
Away
Favorite |
As
Favorite |
As
Underdog |
| Good |
Good |
29-35 (45%) |
4-2 (66%) |
27-16 (63%) |
11-16 (41%) |
40-51 (44%) |
31-18 (63%) |
| Good |
Average |
23-23 (50%) |
4-2 (66%) |
15-14 (52%) |
8-10 (44%) |
31-33 (48%) |
19-16 (54%) |
| Good |
Poor |
14-14 (50%) |
2-2 (50%) |
16-10 (62%) |
4-4 (50%) |
18-18 (50%) |
18-12 (60%) |
| Average |
Good |
19-27 (41%) |
4-3 (57%) |
22-26 (46%) |
3-7 (30%) |
22-34 (39%) |
26-29 (47%) |
| Average |
Average |
19-32 (37%) |
4-2 (66%) |
25-22 (53%) |
3-5 (38%) |
22-37 (37%) |
29-24 (55%) |
| Average |
Poor |
21-20 (51%) |
3-4 (42%) |
22-22 (50%) |
2-1 (67%) |
23-21 (52%) |
25-26 (49%) |
| Poor |
Good |
11-15 (42%) |
1-3 (25%) |
11-12 (48%) |
2-4 (33%) |
13-19 (41%) |
12-15 (44%) |
| Poor |
Average |
18-18 (50%) |
13-10 (56%) |
35-22 (61%) |
1-0 (100%) |
19-18 (51%) |
48-32 (60%) |
| Poor |
Poor |
21-34 (38%) |
19-7 (73%) |
45-31 (59%) |
1-7 (13%) |
22-41 (35%) |
64-38 (63%) |
We didn't expect to see very strong divisions in the data, but in fact we do, as a number of "Past year record - record two years ago" scans show fairly strong results both as "bet on" and "bet against" patterns. Indeed if we elect to sort the results by the "Z-score" probability rating, it winds up like this for most statistical significance:
- Back Poor-Poor teams as UNDERDOGS
- Go against Poor-Poor teams as FAVORITES
- Go against Average-Average teams as FAVORITES
- Back Good-Good teams as UNDERDOGS
- Back Poor-Average teams as UNDERDOGS
None of these have great Z-numbers since the samples are small (the four above range from 2.57 to 1.79), but in preseason you are always dealing with small samples! So with our five-pack of trends in hand, let's look at what this means for the NFL 2006 Preseason:
POOR-POOR teams
Bet on these teams as underdogs, against as favorites:
- Arizona
- Cleveland
- Detroit
- Oakland
- Tennessee
- San Francisco
GOOD-GOOD teams
Bet on these teams as underdogs:
- Denver
- Indianapolis
- New England
- Pittsburgh
AVERAGE-AVERAGE teams
Bet against these teams as favorites:
- Minnesota
POOR-AVERAGE teams
Bet on these teams as underdogs:
- New Orleans
One note is that we didn't include games played on a neutral field in the analysis, so don't go looking for a play in the American Bowl or Hall of Fame game for instance!
In case you missed the earlier parts of our preseason research, you may want to read:
- NFL Preseason Gap Comparisons
- NFL Preseason WIN comparisons
Since all of our research involves games with a home team, there are no recommendations for the opening weekend's games, but we will be back with weekly predictions for the rest of the preseason schedule!
Find out more:
What the press says about TwoMinuteWarning
Comparing NFL Handicapping Services
|
