Your extra 3% is actually pretty accurate, but only when you add the pushes in. Meaning the spread on the fave "insures" you a push about 3% of the time. If you factor in the pushes then fave does not lose right around 84-85% depending on what closing numbers you use.

But since you broke it off at 1000, I will simply look at the past 4 seasons, 2000-2003 inclusive, that is 1007 games. In those games the the SU winner was 823-163-21. So right around 82% win rate, and a 84.5% non loss rate. So we have no disagreements there.

The ATS rate for dogs alone in that time period is 505-465-21, with 16 games closing at a PK, and one game ending in an outright tie. So your numbers are also pretty close, a 50% cover rate, and 52% non loss rate. I am assuming you are using Stanford Wong's numbers.

Within those games the dog won SU 341 times and lost SU 649 times. That is 34.4% so again your percentages are in line.

Now with all that being said, you are basically looking at all this in a vacuum. The results you and I both cited are for ALL dogs all the time. Obviously gys aren't going to like the dog every single game. All I am saying is that when you break it down and you personally like the dog side, it is simply best to play them on the ML.

And oh, by the way. At 34% SU win rate, that means you need about a +190 to +193 ML to break even. The average line for the dogs that won SU was +4.6, depending on home and away and what book, that is generally going to be a +180/+200 ML. So even betting them in the blind on the ML would have appeared to be break even the past few seasons. With your percentage (37%) only a +170 dog is needed to break even, that is about a +3.5 dog at Pin, so not too much of a stretch there to get a SU winner.

But even I don't promote betting all dogs on the ML. In the very least if you like an NFL dog, split the bet up, half on the ML half with the points. The "worst" you will do is lose one way on a push. But basically you will either break even (less some vig collection) win both plays, or have a loss anyways.

It isn't about how many dogs win and cover in a certain amount of games, it is about how often they win and cover in the games a bettor plays them in. I think that is where you lose the distinction.

You could carry it out even further and bet both sides of each game. The fave minus the points, and the dog on the ML. That is why I know the cover rate and non loss rate. With line moves it is actually lower than the numbers above. But that percentage is your chances of getting "polish" middled".

With a 15% to 16% predetermined loss rate, you don't need a very large return on your ML dogs to absorb the times you do get polish middled and lose both ways, right around +135.

In the numbers I cited above had you bet every game both ways (without line differences) you would have won at least one way 823 times, lost BOTH ways 163 times, and lost ONE way 22 times(21 pushes and the tie game).

So depending on how you approach it. Faves won and covered 465 times betting 100 @ -107, that is $43458 in collections. Minus the 535 time they didn't cover (-54200) That leaves you a negative balance of -10742. Then take the 163 time you lost both sides That is -32600, then add/subtract the 22 times you lost one way (pushes) -2200. That is a grand total of 45542 you are in the negative. BUT you that gives you your break even number with the 341 dogs that won SU. That means to break even you need an average ML price on the dogs to be +133. (the numbers do not incude games that were lined at a PK)

Since we have already determined that the average ML price for dogs has been at least +185, this is actually a pretty profitable appraoch.

And I actually do it myself in some sitauations. Not every game like this diagrams, but here and there.

I am sure a lot of people are reading this and laughing or shaking their heads. But it is the most risk free way to bet the NFL, and in the games where the lines are 7 or less it is extremely profitable.

The only problem is that the NFL simply doesn't have enough games to make it a viable approach to take advantage of unless you are making some moderate sized bets. But it does show a profit.