Preseason Tips part I -- "GAP" comparisons

The obvious place to begin is by simply looking at how teams have fared at various spread ranges.

Preseason Games from 1997 to 2007

Category	Won-Lost (ATS)	Win%
Home Teams	326-341	48.9%
Home Favorites	249-292	46.0%
Home Favorites of -1 to -2.5	54-55	49%
Home Favorites of -3	62-76	45%
Home Favorites of -3.5 to -6.5	119-139	46%
Home Favorites of -7 or more	14-22	39%
Home Underdogs	77-49	61.1%
Home Underdogs of +0 to +2.5	37-28	57%
Home Underdogs of +3	21-8	72%
Home Underdogs of +3.5 or more	19-13	59%

So the quick glance take is simply to play the underdogs, and especially home dogs which have been very successful overall.

Fair enough, but what also came to mind when we thought about this thorny issue -- since our approaches for the regular season, heavily dependent on statistical analysis, don't seem likely to do the job for the scrimmages -- is grouping team's by their prior season records. The theory here is that teams with lousy previous year performance might be more interested in posting some W's in pre-season, whereas the teams coming off strong seasons could really care less about the outcome of the exhibitions.

Pulling our quantifying motivation table structure out of the subscriber area, we elected to categorize teams as follows:

• 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

		All Games		w/Home Favorite		w/Home Underdog
Home Team	Away Team	Won-Lost (vs Spr)	Home W%	Won-Lost (vs Spr)	Home W%	Won-Lost (vs Spr)	Home W%
Good	Good	34-31	52%	28-26	51%	6-5	54%
Good	Average	34-34	50%	32-33	49%	2-1	66%
Good	Poor	39-43	47%	33-41	44%	6-2	75%
Average	Good	34-41	45%	23-31	42%	11-10	52%
Average	Average	44-36	55%	39-34	53%	5-2	71%
Average	Poor	28-38	42%	28-37	43%	0-1	0%
Poor	Good	31-37	45%	9-23	28%	22-14	61%
Poor	Average	40-38	51%	24-28	46%	16-10	61%
Poor	Poor	42-43	49%	33-39	45%	9-4	69%

Clearly we are not working with a ton of games where we split it up in so many ways. Some findings of note in include:

1. A battle of average prior year teams has favored the home squad
2. Average home teams have struggled against poor away teams
3. Home teams off less than 10 win years as favorites against prior year Good teams are just 32-54 (37%)
4. Home underdogs versus a "non prior year Good team" are 38-20 (66%)

The next step is to break it out by the specific weeks (conventionally labeled week "zero" to week four). The drawback to this is our already small sample size gets even smaller, but we went ahead and ran it anyway and rather than give you the full tables, we'll summarize the outliers:

Week 1 - POOR prior season teams are 11-20 against the spread at home against POOR or AVERAGE away teams

Week 2 - POOR prior season teams are an amazing 25-9 (74%) as away teams facing GOOD or AVERAGE prior year home teams, and POOR teams are also 31-23 as home teams.

Week 3 - In what some people consider the most serious week of play, there are several notable trends. Worse prior year home teams facing better prior year away teams are just 23-33 (41%) as the superior teams "come to play." Also in 'same versus same' contests, the home side has gone 34-25 so maybe the home field edge is not being suitably considered when teams put out their strongest preseason effort.

Week 4 - Home dogs are 19-8 (70%) but nothing else stands out as too strong in a week where some teams simply decide not to show up and avoid injury above anything else. Frankly the NFL could cut this fourth week of exhibition action and no one would mind.

We believe some of the GAP data conclusions are worth looking for in the 2008 action and we'll be posting a picks page listing all the plays for the "deemed worthy" angles...but this is just phase one of the preseason picks research, with more to come soon!

Also See:
Preseason Picks part II: WIN comparisons

Find out more:
What the press says about TwoMinuteWarning
Information on subscribing
Frequently Asked Questions