Part II -- Preseason WIN comparisons
A common theory among handicappers when it comes to the NFL preseason is that while teams in general don't care much about the outcome of a game, teams without a win try harder than teams that have already secured one or more preseason victories. The question is, does the theory hold up when historical results are assessed? We'll try to answer that here...
We ran through the results of all preseason games from 1997 to 2001 (eg a five year span) where there was a home team (in several games each season you have a "neutral site" and we will look these contests a little later). All we used initially in our query is how many preseason wins a team has coming into the game. Obviously then the first preseason game a team plays it has zero wins, if it's victorious then next week it comes in with one win, etc.
Since the sample size is already small, we elected to cap a team's wins at "2+" for the purposes of this analysis:
Home Team
Preseason Wins |
Away Team
Preseason Wins |
Won
(vs Spread) |
Lost
(vs Spread) |
Home
Win% |
| 0 |
0 |
34 |
52 |
39 % |
| 0 |
1 |
18 |
18 |
50 % |
| 0 |
2+ |
10 |
7 |
58 % |
| 1 |
0 |
16 |
12 |
57 % |
| 1 |
1 |
23 |
20 |
53 % |
| 1 |
2+ |
5 |
22 |
18 % |
| 2+ |
0 |
5 |
2 |
71 % |
| 2+ |
1 |
11 |
11 |
50 % |
| 2+ |
2+ |
15 |
9 |
62 % |
The preliminary conclusion to the question that started all this -- who is the better bet when a team with zero preseason wins plays a team which has already notched a W -- shows there's not a lot to the theory...in fact the winless side is just 14-21 against the spread on the road against teams with a win, and a modest 28-25 at home, for an overall 42-46 mark. We count that squarely in the "no predictive value" category.
However, what does jump out from the above are two very interesting points:
- Back an away team with two or more wins against a home team with exactly one win (22-5 for 82%)
- In a game of winless sides, play the away team (52-34 for 61%).
Well, point one makes some sense if you take the position that the team with two wins already has shown a stronger than usual desire to win in the early going and the home team with the win may be comfortable taking a casual attitude to the game (since they do at least have a win). The second angle (two winless sides, which obivously is the case in every week zero game and many week one games) though gets even better -- we ran the numbers looking at the lines on the game, and it turns out that playing the underdog in the matchup (and consider a pick'em game to have the home side as the underdog), yields a 59-27 record (68%)!
Finally, we alluded to the neutral site games, and these turn up nine meetings between winless sides, and the underdog again carried the day with a 6-3 record.
So, two more angles for the 'old arsenal heading into the preseason, but the wise guy in the audience is already carping that looking strictly at wins without regard to the stage of the preseason is foolish...hold on pal -- we're getting to that!
Rather than fill up this article with countless tables though, we'll give some of the more interesting data in summary:
Week 0 -- play the dog (10-4)
Week 1 -- a zero win home dog is 8-5, but more importantly, in a battle of winless sides an away dog has clicked with a 30-11 mark.
Week 2 -- zero win home dogs are 5-1
Week 3 -- nothing stands out...
Week 4 -- two or more win away sides are 11-1 against one win home sides, while in battles of 2+ win teams, the home squad is 14-6.
So to wrap it all up for this rendition of TwoMinuteWarning.com's preseason research we recommend the following:
- Play away teams with two or more wins against home teams with one win
- In a match-up of winless teams, play the underdog
- In the final week of preseason in a game between teams with multiple wins, lean to the home side.
|
Copyright © 2002 by TwoMinuteWarning.com, All Right Reserved
| 
