I'm sure there will be a lot of questions on the Margin Difference system Formulawiz has developed since it is doing very well, so I thought I would document how to apply the system and share it with the thread. Formulawiz has already reviewed and blessed this. I'm only including a couple of examples... this post will be long enough as is. If you want more examples, send me an e-mail at ptanko@hotmail.com and I'll send you some more.


Margin Difference System
You need the College Football Prophet Zone 2 program for this (same for NFL).

Make sure you've downloaded the latest stats update from Sportrends and placed it in the correct folder on your computer. Sportrends usually updates their data by Tuesday for both NFL and College. It doesn't change until the following week.

From the Prophet Zone menu, select "Matchup" then “Team Matchups”. Select a date and game from the drop down lists. Leave the games back at 10 in the top right hand corner. Leave the drop down menus for Away/Home as is (they are located to the right of the team name). Do not use Overall for this system. If you leave the system as they default, you will be fine. Read the Margin Difference data for each team at the bottom right hand side of their respective areas.

Filters:
1. 3 pt overlay for college and NFL
2. No spreads > 15
3. No small colleges
4. Wait until 15-20 minutes prior to start of game play to make your play.

How to use:
Example 1:
• Hypothetical data for example purposes.
• Margin Diff data: UCLA -9.4 @ USC 20.3. I typically round them to the nearest 0.5 pt (e.g. -9.5 for UCLA and 20.5 for USC).
• Vegas line UCLA -3 which is the same as USC +3

1. Calc the difference between the team’s Marg Diff. From -9.5 to 20.5 its 30 pts. This would be the estimated typical delta in the score between these two teams using their historical data. Again this is all hypothetical. In this example we are not using the real data on Sportrends but you apply the same process to the Sportrends Marg Diff data to make a pick.
2. Divide the result by 2. 30/2 = 15
3. Which team should win according to Marg Diff >> USC. They have a better Marg Diff than UCLA. So Take the expected winner (USC) and apply the result from step 2 to their Vegas line. Also, note you do not add HFA. It is already a part of USC’s Home Margin Difference value.

(Step 2 results) + (Line for the Marg Diff favorite) = Result
USC: 15 + 3 = 18

4. Compare this result to an overlay of 3 pts (NFL is 3 pt overlay as well). 18 is >>>> 3 pts so USC is the safe pick (at least for now).
5. Apply additional filters to make sure USC is the pick. Do this by comparing each team's Marg Difference to the Vegas line.
a). Does UCLA typically lose by more than the -3 line? Yes, not only does UCLA not cover by the 3 pts they are giving USC, they typically lose by 9.5 pts (according to the Marg Diff). Great! This helps confirm our USC pick.
b). Does USC typically win by more than the 3 pt line? Yes, USC typically wins by 20.5 pts (according to the Marg Diff). Great! This also supports our USC pick.

Things are looking good for a bet on USC. But we still need to apply the other filters.

Special Note: Step 5 can be tricky to apply. If both a) and b) criteria are not met, you can still bet the game by comparing how close each team’s Marg Diff is to the Vegas line. See additional examples below.

Apply the Filters:
1. 3 pt overlay – good to go on this one. 18 >>>> 3
2. No small colleges. Not a problem here.
3. No >15 pt spreads. Also not a problem.
4. Wait until 15-20 minutes prior to start of game play to make your play.

For the following examples, the filters (no small colleges, no spreads > 15 pts) are met so they will not be checked. All of these examples are hypothetical. The data is not what is shown on Sportrends. Each example follows steps 1-5 above.


Example 2:
• Marg Diff data: Conn -10.5 @ Iowa -9.0
• Vegas line Conn -3.5 which is the same as Iowa +3.5

1. Calc Δ on Marg Differences: -10.5 to -9.0 = 1.5
2. Divide by 2: 1.5/2 = 0.75
3. Calc expected winners value using Vegas line: Iowa has higher Marg Diff (-9.5 is higher than -10.5 for Conn) so use the Vegas line for Iowa to do the calculations.

0.75 + 3.5 = 4.25 Iowa is the pick. If, when going through these steps, this calculation result comes out negative, then you pick the other team and confirm that the overlay is < -3 instead of > 3.

4. Compare to 3 pt Overlay: 4.25 > 3 overlay so we are good to continue.
5. Apply Step 5 x-checks:
a). Iowa typically loses by 9.0, as seen by their Marg Diff of -9.0, which means they typically lose by more than the 3.5 points they are receiving in this game, so this DOES NOT support our Iowa pick.
b). Conn typically loses by 10.5, as seen by the Marg diff of -10.5, and the Vegas line expects them to win by 3.5. So clearly they typically do not cover a spread of 3.5 as a favorite, so this supports our Iowa pick.

Since both a) and b) did not agree with our pick, compare the Marg Diff for each team to the Line to help make a decision. Clearly it will be much harder for Conn to win by more than 3.5 (-10.5 to 3.5 = 14 a large value) than it will be for Iowa to lose by more 3.5 pts to Conn (-9.0 to -3.5 = 5.5 a much smaller value) so it would still be safe to bet on Iowa. Or put another way; since Conn usually loses by 10.5 pts, it will be very hard for them to win this game by 3.5 pt as determined by the delta between their Marg Diff and the line (-10.5 to 3.5 = 14). Since Iowa typically losses by 9, it will not be as hard to only lose by 3.5 or less as seen by the delta between their Marg Diff and the line (-9.0 to -3.5 = 5.5).

Pick: Iowa +3.5