In part one of this series we looked at how a team's won-lost record coming into a match-up had historically effected the outcome. A reader astutely pointed out that we might want to look at the same approach, only focusing on the spread winning percentage of each team coming into a game.

We'll stick with the same class categories:

• Poor -- teams with a spread percentage of under 40%
• Mediocre -- teams with a spread percentage of 40% to 49%
• Good -- teams with a spread percentage of 50% to 59%
• Excellent -- teams with a spread percentage of 60% or higher

ANALYSIS: Some fairly strong patterns emerge here -- home teams with poor spread records have been solid picks against teams with good/excellent spread records (146-104 for 58.4%). Then you have the oddities of the Mediocre spread record teams: at home against poor spread record sides, the poor record team has covered 63% of the time, but then when a Mediocre spread team is on the road against a good one, the mediocre club has got the cover almost 60% of the time. When you see things like that it's easy to believe it's just a data aberration and not much value. The first observation though since it crosses two consecutive classes seems to hold more promise.

Let's look a little further then into the "home team with poor spread record" by changing the standards for Poor to 0 to 33%

So with a more stringent requirement it gets even a little stronger. This all seems to harken back to some of the fundamentals of such things as Turnover Difference and Contrarian Bridgejumping, namely the teams with bad records are often under-rated. Before we get too worked up about it, let's do the sanity check against the 2001 results:

Not a lot of success then in 2001 with this approach, a modest 18-19 record between the two angles. Requiring the more stringent (33% or worse) standard would have helped move the numbers to 14-11 and 4-4 respectively, which at 18-15 (54.5%) is no great shakes but at least ahead of the vig. It's a healthy 9-4 if you drop it to 25% or worse...but now we're really reaching!

The third possible angle from the first chart was the Good-Mediocre combo and that did play more true to form, with the mediocre spread record team covering 7 out of 10. Our attitude then at this point is there may be some modest value, but overall we're a little more sold on using the straight W-L record.

One last piece of interest revolves around the extreme teams -- the ones who are winless against the spread, and those that are undefeated against it coming into a contest (week 5 on).

Those are some strong numbers on tiny samples. The "angles" would be the following, with the 2001 records in parentheses:

• Play against a team with no spread wins that's on the road (2-3)
• Play on a team with no spread wins at home (1-0)
• Play against a team with no spread losses (1-0)