I'll take your best argument, THE NUMBERS DONT LIE

First off....  Thanks for taking the time to make this kind of post.  There is very little of this these days here - and I for one value it and think it is way more important and has more potential than picks posts.

I like what you say about ROI vs %.  Profit is the only thing that matters.  And in this game, any place you can squeeze pennies to make more profit is worth the effort.  It is always easier to save a dollar than to make a dollar.

One thing to add to your work....  It is often easier to find a lower house hold (spread in price between two options) on spreads than MLs.  For example - at lowvig they deal a 10 cent line on spreads, but often a 22 cent line on MLs.  This certainly doesnt disprove anything you say above.  It is just something to be aware of.

Any thread that gets vanzack up off his yacht to post in is fine by me

And I completely agree with this post...not that there's anything wrong with that

As stated, this is the type of thinking that is or has been lacking from this site for some time...very few understand or even think this way any longer...it's missed

HD and line < 3.5 and season > 2013

(202-190-19) 51.5%  Last 10 years of HOME DOGS of less than 3.5.

Supports the ml wager versus spread assuming juice is OK

I respect you trying to seek a way to save profits or earn them, and then share them with the forum. Great!
But in your example there are a few flaws ..... This is NOT an attack/flame/insult ....it is a friendly response to make corrections....

First, in your math example it should NOT be $3400 @ +130 then subtracting the 5 ML dogs that covered ATS but lost ML
What you are in effect doing is counting those 5 (twice!) first with the 29 winners, then subtracting them. That's a false net of 0.
         It should be 29 wins x +130 = +3770  (not 34 wins x +130)
        subtracting those 5 ML losers =  -500
         .................for a net profit of +3270   (not $3920)

Yes, this is still better than $3090 betting ATS, but far from the $3920 profit you quoted.

Second, you've assumed a mean avg of +130 for the small DOGS. But from my review,
the mean avg for small DOGS getting pts between +0.5 and +3 (per your example) but then playing them on the ML is nowhere near +130....it's actually between +121 and +122 mean avg, but I am rounding up to +122. (More accurate would be using +121.5)
      So that changes the numbers even more!
      29 wins X +122 = +3538
      minus 5 losers  =    - 500
      for a net profit of   +3038    (not $3920)

The more realistic $3038 betting ML dogs compares with $3090 betting ATS dogs.  Not much diff but ATS is the clear winner (not the ML dogs)

Third, and this is a biggie caveat, this is a small sample size. Variance in numbers over an extended period can be quite significant. So big "grain of salt" with any conclusions from these small numbers either way 

**Additionally, the small dogs from the first few weeks of any season are shaky guesses on limited information.  Small dog lines that occur after week 4 are reflective of each team's level of play evaluation thru 3 - 4 games.  I would not trust numbers from the first few weeks and how those stats might influence decision making when the lines are much more solid in later weeks, and MANY cappers avoid them altogether for this reason.

@unplucked_gem

 Thanks. These kind of posts have popped up here n there over the years. Definitely a rare occurrence. 

Even after reading your post, which makes perfect sense, there are those who still can't resist buying the hook. It pains me seeing -120, -125, or -130 on spreads.

To each their own

Thats why I keep a units record with my win %. My bottom line ( Profit/ ROI) is what is important to me at the end of the day / season

Let me clarify , the SU data is 994-709 or 58.4%, DOG money line very much the PLAY!

  Ooops....rechecking my own work I see I made errors too, including I forgot to include those 3 pushes, which makes it even more in favor of going ATS rather than ML.... 

Of your 78 "coin flip" games:
•  41 times the fav covers the number
•  29 times the dog wins outright
•  5 times the dog loses ML, but covers the number ATS
•  3 times the number pushes.

The 3 times the fav line pushes(1, 2, or 3 pt spreads) would mean the DOG player betting ATS would also push, for a net of zero ......but the ML dog would have LOST all 3

So, disregarding the 41 times the fav covers the numbers, which is an equal number of losses for either the ML DOG bettor or the ATS DOG bettor, we get this math:

ATS DOG player: (ie, bet $100 to win $90.91)
29 outright winners X $90.91  = +$3538.00
5 lose but cover ATS X $90.91 = +$ 454.55
3 times the line pushes ATS    = + 0
==============================

                                                   = $3992

Moneyline DOG player: (ie, bet $100 to win avg +$122)
29 outright winners X +$122 avg  = + $3538.00
5 cover ATS, but lose ML! X -$100  =  - $500.00
3 pushes ATS are all ML losers!      =  - $300.00
==================================
                                                          = $2738

So ML dog bettor $2738 vs. ATS dog player $3992

I'll argue with you. You said, "numbers don't lie"?  Have you paid any attention to the presidential elections, or better yet, to any elections??? Probably around 85% of any elections, mayor, City council, judge, presidential etc? The numbers lie all the time.

DISREGARD. SORRY

HD and line < 3.5 and season > 2013

shows (185-217-3) 45.% SU winners, supporting fubah2's conclusions!!