re: Official Week 9 College Football Bet Thread

quote:

---

ASU is receiving the overwhelming majority of bets, and the line has increased accordingly

---

Obviously that isn't what I'm seeing. The information could be wrong but I'm not seeing that at all.

quote:

---

What do you consider low for your standard deviation?

---

Anything under 4.5, The lower, the better.

I would pay more attention to the "ATS Standard Deviation" for sides. This pays more attention to the value each side has according to my model, and is what I was researching when I found the link between low standard deviation and the model's predictive success.

The O/U Standard Deviation is also something to look at.

Basically it is measuring variance.

If Method 1 says Team A wins by 20, Method 2 says Team A loses by 20, and Method 3 says Team A wins by 50, there is a lot of variance there because they are all wildly different. Therefore there couldn't be much confidence in my model's predictions.

But if Method 1 says Team A wins by 7, if Method 2 says Team A wins by 6, and Method 3 says Team A wins by 8, there is barely any variance, so more confidence can be placed in what my model is projecting there.

Standard Deviation is why I added the Navy -7.5 bet. Standard Deviation was 3.3. South Florida projected to score 16, 20.6, 18.37 points. Navy projected to score 38, 34.84, 34.51 points. Low variance on both sides.

So we need to take the sum of the std deviation numbers on both sides and look for 4.5 or lower is the best?

Revised early leans... Cal and UVA. If the numbers are right, probably my only two plays.

quote:

---

So we need to take the sum of the std deviation numbers on both sides and look for 4.5 or lower is the best?

---

No

The number for sides you want to look for is the "ATS Std. Deviation Value". It will be the same on both sides. If this number is under 4.5 that is a good sign.

So for example, if my model picks a side across all 3 methods, and the ATS Std. Deviation Value is under 4.5, that's a good sign.

Wisconsin is an example. My model picks Wisconsin ATS on all 3 methods. ATS Standard Deviation Value is 1.79. Therefore Wisconsin would be a play my model is going with that has low variance.

East Carolina/UCONN has a really low total standard deviation. All 3 of my model methods pick the Under in that game, and the O/U Std. Deviation is 0.29. Low variance, universal agreement on the under. Also for totals its OK to also look at the "Standard Deviation" line as well. There can be times where the Standard Deviation for both team is low, but the O/U Std. Deviation is high. That's OK, as long as one of the two are low.

If I were going big on a play, I'd look for these criteria:

-Low standard deviation
-Universal agreement on all 3 model picks
-High cover(or over/under for totals) percentage

It's rare that all 3 of these boxes will be checked.

I went with UMass/Ball State OVER because it was given a 90.2% over chance, all 3 methods picked the over, and the individual standard deviations were low (3.21 for UMass, 5.45 for Ball State).

I went with Navy -7.5 because Navy is given a 71.4% cover chance, with a 3.3 standard deviation and universal agreement.

Louisville/Wake Forest Under. 86.9% under chance, universal agreement, 4.96 O/U standard deviation(with a 1.42 Louisville Standard Deviation)

Hope you get the approach.

what are you, the bet thread Wizard?

someone wanna explain how ND isnt the lock of the week vs Temple? im sure this is one of those games where vegas "knows something you dont" lines but Im honestly not seeing it unless someone can convince me otherwise?

most of the talk has been about "temples defense" but the offenses they have face are poverty.

(according to FEI), they faced the following.

#45 Cincy
#63 East Carolina
#78 UMass
#80 Penn St.
#89 Tulane
#126 UCF
#127 Charlotte (There are 128 FBS teams)

(For reference, ND has faced four in the top 23; #3 Navy, #10 USC, #21 Ga. Tech and #23 Clemson)

cincy had over 500 yards vs temple, but turned the ball over 5x and lost

their defense is extremely undersized

they havent faced an offense anywhere close to the caliber of ND (ND is 4th in the country in yards per play)

they trailed 1-6 Umass in the 4th qtr
they trailed 0-8 ucf in the 4th qtr
they trailed 4-3 ecu in the 4th qtr

so they could easily be 5-2 or even 4-3

unless ND derps this game away it should be an easy cover.

quote:

---

poverty.

---

i love your usage of this word

really wish youd get in on the groupme

and yes, I'm with you. almost locked em in this morning

ASU -1.5
Kentucky +8.5
ND -10

I'm 0-2 betting against Temple this year (PSU, ECU) so hopefully third times a charm. ND simply outclasses them. I expect multiple interceptions thrown by PJ.

Also leaning Maryland +17 and Syracuse +20.5 but road dogs have never been kind to me. Still debating.