Predicted Scores vs. Vegas Lines

D

djchad

Registered User
Forum Member
Nov 8, 2001
233
0
0
Melbourne, FL
I caught one other mistake I posted. The UNLV play should have been a play on New Mexico.

Here is the final chart for Oct. 10-12.

<pre>
Favors Line Avg Pred Diff Std Dev Play Result Score
Air Force -4.50 -15.85 11.35 6.30 ** W 52-9
Arizona 13.50 7.53 5.97 3.79 * W 28-32
BG -15.50 -20.16 4.66 8.69 * L 45-35
Colo St -25.50 -29.15 3.65 5.56 * L 44-37
EMU 9.50 3.64 5.86 8.41 * L 27-55
Fresno St -23.00 -26.26 3.26 4.05 * T 30-7
Georgia -3.50 -10.91 7.41 7.18 ** W 18-13
Idaho -11.50 -15.92 4.42 9.26 * L 14-34
Indiana 13.00 10.66 2.34 3.82 - - -
Iowa -6.00 -13.71 7.71 5.00 *** W 44-16
Iowa St -6.50 -8.71 2.21 3.59 - - -
Kansas St -14.50 -18.86 4.36 5.27 * W 44-9
LSU 8.50 2.88 5.62 7.72 * W 36-7
Troy St 10.50 7.40 3.10 6.25 * W 8-11
NC St -6.50 -11.10 4.60 5.99 * W 34-17
ND -7.50 -13.55 6.05 5.87 *** W 14-6
Missouri 8.50 5.78 2.72 5.87 - - -
Oklahoma 3.00 -0.98 3.98 5.46 * W 35-24
Purdue -3.50 -7.65 4.15 4.03 * L 31-38
TCU -21.00 -29.13 8.13 8.61 ** L 46-27
New Mexico 7.00 3.73 3.27 4.16 * W 25-16
USF 3.00 -1.74 4.74 5.88 * W 16-13
Utah -10.50 -15.95 5.45 4.39 * L 17-36
Vandy -5.50 -10.81 5.31 8.28 * L 20-21
Virginia 5.00 0.72 4.28 3.30 * W 22-17
VT -8.50 -14.46 5.96 7.44 * L 28-23
Wash St -7.50 -12.96 5.46 7.33 * W 36-11
WMU 1.50 -2.55 4.05 4.23 * L 27-31
WV -13.00 -16.57 3.57 8.67 * W 40-0
</pre>

This week's results:

3-star: 2-0-0
2-star: 2-1-0
1-star: 11-9-1


Season results:

3-star: 6-0-0, 100%
2-star: 6-1-0, 85.7%
1-star: 17-15-1, 53.1%


Next week's should be up by Wednesday.

- djchad
 
K

kgar2121

Registered User
Forum Member
Jan 25, 2001
301
0
0
north carolina
just a little update

just a little update

hope you dont mind djchad, but ive been fiddling around with a spreadsheet based on your findings, and this weekend, and this admittedly is a small sample, shows a better ability to pick upsets or for the dog to cover numbers.....

this weekend, the money picks went a total of 15-11....of that, the favs were 8-9, for 47%. however, the dogs going for the cover were 7-2, for 78%. now i realize that one week does not a fortune make, but its certainly something worth looking at the next couple of weeks....thanks and keep up the good work!
 
D

djchad

Registered User
Forum Member
Nov 8, 2001
233
0
0
Melbourne, FL
These are the early, early numbers for Oct. 17-19. No std deviation numbers yet. This is with 22 of 35 prediction services reporting - so these numbers should still be considered preliminary. I'll be updating these numbers Wednesday evening.

So far, the best plays look to be South Florida, Houston, USC, Virginia, Georgia, Memphis, LA-Lafayette, NM State, and Virginia Tech. But the Va Tech play goes directly against the idea of +40 dogs.

<pre>

Favors Line Avg Pred Diff
ND 3.00 0.68 2.32
Ark St 8.50 5.75 2.75
BG -15.50 -20.35 4.85
Cal off -1.68 #VALUE!
Toledo 4.00 1.71 2.29
Wake 9.00 6.15 2.85
UConn 0.00 -3.29 3.29
USF 1.00 -6.22 7.22
Georgia -26.00 -31.47 5.47
Houston -14.00 -22.52 8.52
Kansas St -1.50 -4.69 3.19
LSU off -15.65 #VALUE!
Memphis 0.00 -5.24 5.24
Minnesota 3.50 0.42 3.08
LA-Lafay 10.00 4.25 5.75
NC State -24.50 -27.98 3.48
SJ State 6.00 1.53 4.47
New Mex St -17.00 -22.40 5.40
Michigan -4.00 -6.64 2.64
Sou. Miss off -10.72 #VALUE!
USC -7.50 -15.14 7.64
UTEP off 10.13 #VALUE!
New Mex -3.00 -5.91 2.91
Virginia -4.50 -10.89 6.39
VT -42.00 -50.00 8.00
West Va -7.50 -11.62 4.12
Ohio St -7.00 -9.85 2.85

</pre>

Chad
 
J

Joe De

Registered User
Forum Member
Sep 10, 2002
1,440
1
0
thx...i'll match this up with mine and let you know the picks...thx again
 
G

Gabuk

Registered User
Forum Member
May 8, 2002
1
0
0
Chad...are my calculations right?

Chad...are my calculations right?

Hi Chad,

Wanted to say thanks for the work that you have put in.

Based off of your previous posts, I am trying to determine if what I did is considered correct by your calculations.

This is based off of the formula calculations at 9:51 p.m. PST.

I calculated the average of the standard deviation at 4.775.

Out of the nine picks you list, I am trying to calculate for five of them. Three of the picks, GA, Louisiana-Lafayette and Memphis did not cross the six point "hump" so I temporarily left them out. Also, I don't have a line for the New Mexico St. game.

Out of the five remaining teams, I came up with these picks based off of your two or three star selection process.

2 Star
USF +1
Houston -14
Virgina -4.5
VT -42

3 Star
USC -4.5

I appreciate you taking the time to read my message. Hopefully, I didn't mangle your calculations =).

Forgot to add, I know these are still temporary calculations (not all formulas are in, which will change the STD deviation along with the average of the STD deviation) but I am trying to learn the process so I don't end up asking you a million dumb questions later in the week.

Good luck to everyone!
SD
 
Last edited:
H

howdy

Registered User
Forum Member
Sep 20, 2002
146
0
0
65
djchad,
THis is very interesting indeed!
I tried to email you for the url but the system says you are not accepting any email.
If you would send the url to [edit out email] I would be most appreciative.

Also,
I hate to ask stupid questions but I seem to be the only person who does not understand exactly how you come up with the std. dev. would you mind explaining.


thanks
 
D

djchad

Registered User
Forum Member
Nov 8, 2001
233
0
0
Melbourne, FL
Thank you for posting the link Jack. I didn't know if you had any restrictions on posting URL's here to such things, so I played it safe and kept it off.

Explaining standard deviation:

First, for those that may not know. Definition of std dev.
"The standard deviation is a statistic that tells you how tightly all the various examples are clustered around the mean in a set of data. When the examples are pretty tightly bunched together and the bell-shaped curve is steep, the standard deviation is small. When the examples are spread apart and the bell curve is relatively flat, that tells you you have a relatively large standard deviation."

If I list the std dev for a game as 6.5 and the average prediction is 14 points, this means that 68% of the predicted scores give a margin of + or - 6.5 away from the 14 points. In other words, 68% of the predicted scores fall between 7.5 and 20.5 points. Also, 95% of the scores will fall between + or - two standard deviations away from the average prediction.

What this means to us is, the lower the standard deviation, the closer the predictions are to the average. This a good thing, because it means that the predictions come close to agreeing on a margin of victory.

The standard deviation of all the predicted scores is a formula, and the gentleman that runs the site listing all the predictions happens to calculate this for us.

Then, to differentiate between the 2-star and 3-star plays, I need to determine in which games are the std dev's fairly close. So, I add up ALL the standard deviations of ALL the games predicted. 2-star games will be above that average (having a higher std dev) and 3-star games will be below that average (having a lower std dev).

I hope that helps clear up some of the confusion. I'll be updating the numbers shortly.
 
D

djchad

Registered User
Forum Member
Nov 8, 2001
233
0
0
Melbourne, FL
Here are the numbers for Oct. 17-19.

A couple of notes:
- The Memphis line is at -2.5 in most places now. Well, I got it at pick, so that's what I am listing here ;)
- This week's avg std dev is 5.16. I took out the two highest std dev's out of the average because they were too far out.


<pre>
Favors Line Avg Pred Diff Std Dev Play
Ark St 8.50 3.91 4.59 9.63 *
BC -28.50 -31.21 2.71 6.52 -
BG -15.50 -21.23 5.73 6.18 *
Cal 3.50 -1.99 5.49 3.63 *
Colo St 0.00 -2.97 2.97 7.24 -
East Mich 15.00 12.00 3.00 6.12 *
Georgia -25.50 -34.07 8.57 8.98 **
Houston -14.00 -21.88 7.88 9.69 **
Kansas St -1.50 -4.55 3.05 3.89 *
LA-Lafay 9.50 3.52 5.98 7.32 *
Marshall -19.00 -23.16 4.16 5.63 *
Memphis 0.00 -3.89 3.89 5.27 *
Michigan -3.50 -7.97 4.47 4.38 *
Minnesota 3.50 0.01 3.49 3.14 *
NC State -24.50 -29.34 4.84 5.45 *
ND 3.00 -0.30 3.30 4.30 *
New Mex -3.00 -6.61 3.61 3.93 *
New Mex St -17.00 -21.94 4.94 4.68 *
North. Ill -9.50 -12.60 3.10 3.20 *
Ohio 11.50 8.47 3.03 4.06 *
Ohio St -7.00 -9.96 2.96 4.51 -
SJ State 5.00 0.93 4.07 4.12 *
Texas Tech -6.50 -10.40 3.90 3.81 *
Toledo 4.50 0.62 3.88 4.62 *
Tulane -4.50 -8.20 3.70 14.19 *
UConn 1.00 -2.87 3.87 5.73 *
USC -7.50 -14.91 7.41 4.18 ***
USF 1.00 -7.57 8.57 5.71 **
UTEP 13.00 9.18 3.82 7.11 *
Virginia -4.50 -12.04 7.54 5.91 **
VT -41.50 -49.83 8.33 11.75 **
Wake 8.50 5.20 3.30 2.93 *
West Va -7.00 -13.02 6.02 7.68 **

</pre>
 
Last edited:
S

stackdaddy

Registered User
Forum Member
Aug 1, 2002
23
0
0
51
Reno
Thank you DJ, I never thought of taking out the sd that would throw off the data. Have you ever looked into other sports (basketball, baseball) and how this formula would apply? I know many of those prediction services do report them. Thanks again for your work.
 
T

Tduji

Registered User
Forum Member
May 20, 2001
168
0
0
Denver, Colorado USA
Just an (ignorant?) question for you if you dont mind....

Just an (ignorant?) question for you if you dont mind....

I was just wondering if the line is greater than the predicted average and the standard deviation is much less than the average, would this constitute a play on the opposite team ?

Just trying to refine this system that is a great way of narrowing down my list of games I have capped....

Keep up the great work and good luck :D



Just as an example....

Clemson is favored by 9 at home vs Wake Forest.

The prediction average is 5.2 (3.8 less than the line) and the standard deviation is 2.93 (less than the average std deviation)

So does this indicate a possible play on Wake Forest with the 9 ?

Thanks
 
Last edited:
You must log in or register to reply here.
Bet on MyBookie
Top