Predicted Scores vs. Vegas Lines - Oct. 24-26

[djchad]

Here they are, with an average std dev of 4.82:

<pre>
Favors Line Avg Pred Diff Std Dev Play
Air Force -17.50 -25.95 8.45 6.15 **
Alabama 2.00 -2.88 4.88 5.39 *
Arizona St -3.50 -8.35 4.85 4.60 *
Baylor 30.50 27.67 2.83 11.11 -
BG -21.00 -23.96 2.96 5.18 -
Buffalo 6.50 2.95 3.55 3.65 *
Cal 5.50 0.09 5.41 5.48 *
Cent Mich 20.50 15.33 5.17 4.95 *
Colo St -12.00 -17.53 5.53 4.71 *
Idaho 4.00 1.16 2.84 5.10 -
Indiana 9.00 1.95 7.05 4.51 ***
Iowa St 13.00 7.39 5.61 5.46 *
LSU -3.50 -12.51 9.01 5.41 **
NC State -1.00 -3.74 2.74 3.56 -
Nevada 4.50 -2.81 7.31 5.00 **
New Mexico 6.00 1.06 4.94 4.96 *
Notre Dame 10.00 1.36 8.64 6.85 **
Ohio St -4.50 -10.54 6.04 5.51 **
Rutgers 15.50 12.44 3.06 6.99 *
SJ State 16.00 12.88 3.12 6.03 *
Texas Tech 6.00 3.29 2.71 2.32 -
UAB -8.00 -13.88 5.88 14.78 *
UConn 8.00 2.37 5.63 4.24 *
UTEP 2.50 -1.31 3.81 7.58 *
Va Tech -28.50 -33.28 4.78 7.14 *
Virginia 5.50 -1.65 7.15 5.36 **
Wash St -11.00 -16.31 5.31 5.35 *
Wisconsin -2.50 -5.45 2.95 3.10 -
</pre>

3 star play = The predicted score is at least 6.00 above the
posted line AND the std. deviation of the predicted scores is less
than the avg std dev.

2 star play = The predicted score is at least 6.00 above the posted line AND the std. deviation of the predicted scores is greater than the avg std dev.

1 star play = The predicted score is at least 3.00 above the posted line but less than 6.00 above the posted line.


A few "wrong team is favored" picks this week:
Alabama, Nevada, UTEP, and Virginia

And even though the Texas Tech pick does not merit a star rating, it should be given a look, especially with a std dev so low.

 

[TORONTO-VIGILANTE]

always look for your posts....!!!

thanks!!!

 

[Master Capper]

What do the stars mean at the end of the grid?

 

[Joe De]

thx djchad..as usual the stats do not show my dog of the week..no problem...it does have one which is Indiana
As i understand ...the lower the Std deviation and the higher difference ....makes a play ..example you said tex tech with a 2.32 std. deviation with a difference2.71 is not a play because it did not hit above 6.0,,,is that correct?

 

[moeclarrett]

out of curiosity, I was wondering if it is coinsidence or a trend that most of the two and three stars plays (6 out of 7) are the visiting team. Anyone else notice that and has anyone taken note if this is a regularity or not.

Also, gotta respect the Bucks qualifying for 2 stars. It definitly was a no play for me before, as I would never bet against my team, and rarely do I even bet for them - especially with the entire world calling for Penn St. to not only cover, but to win outright. But the way this system has been working thus far, I may throw a little money on them.

dj, that USC game was a great payoff - thanks much

Go Bucks

 

[djchad]

Yes, a low standard deviation and a high point difference makes a play. But 1-star plays hit because the diff is over 3 pts, and less than 6 pts.

With Texas Tech, the std dev is 2.32 and the average prediction is +3.29. By the rules of standard deviations, about 68% of the predictions are between +0.97 and +5.61. The line is at TT +6. So, these 68% of the predictions think TT will cover the 6 points.

Looking back over each of the actual predictions, 30 of 32 predictions think Texas Tech will cover the 6 points. It's very hard to argue with numbers like that.

 

[Joe De]

thx...i understand..i show utep as a No. 1 pick..blow them off the chart..but not my best opinion..Nevada comes up strong..and virginia...but i'm leary of this team covering will probably backoff to another play...fyi

 

[Jhpga]

Djchad.sorry if I dont completley understand your system.But I have LSU as my best game of the week.I have them winning 24-13.And I see you have them winning by 12.5.Why isnt this your strongest play.Just asking!!!Thanks

 

[heleanth]

When calculating the average std deviation, do you use all the games or just the ones that have a difference of more than 3?

I added all the std. deviations from the site that MadJack gave us earlier and I got an average of 5.14. This would make Nevada a 3* play, if I am understanding your system.

Would you please comment on this for us? Thanks for doing the work, djchad. Your efforts are appreciated.

Mike

 

[T-roe]

hey djchad how is your record so far this year thanks for the info

 

[Scarecrow]

To T-roe

To T-roe

Hey T-roe,

Just do a search for djchad and you can get his past posts and record for yourself. BTW, his record is excellent!

 

[djchad]

Jhpga - The LSU game may have a high prediction/line differential, but the standard deviation is a little too high to merit a 3-star rating.

Here is my first post regarding these plays . You can find my record there

 

[heleanth]

Djchad, would you please look at my question a few posts up in this thread? I am trying to understand how you calculate the average std deviation. Do you use all the games that were predicted? Or only the ones that have a predicted score average over 3 points different than the line?

I was trying to duplicate your 4.82, but I come up with 5.14, which would make Nevada a 3*.

Thanks for taking the time to post your system. I think it has a lot of merit and I am going to bet some of the plays this weekend.

Mike

 

[Catdawg]

djchad-

I'm sorry I didn't have a chance to see this thread earlier in the year. Great work!

So I know what I'm looking at, I assume a line move in the PSU/OSU game to OSU-5 would knock OSU down to a one * play?

Thanks.

 

[finsfan]

Where do you get the avg predicted scores?

A guy here at work is doing something similiar as well

 

[MasterTX]

Heleanth...avg of std dev's comes from ALL the plays...I asked the same questions last week, hope that helps!

 

[heleanth]

Thank you, MasterTX. If that is the case, maybe the numbers changed since Wednesday, when Djchad posted. I ran the average a couple of times and came up with 5.14, which, if I understand the system, would make Nevada a 3*.

Good luck to all and thanks to everyone that takes the time to post information that is helpful in handicapping.

 

[djchad]

heleanth,

Sorry I missed replying to you earlier. Yes, I do calculate the average standard deviation from all the plays. However, I remove the standard deviations that are too far away from the average. This week, those deviations that I removed are from the UAB/Army game and the K State/Baylor game. I don't have a hard-set rule on this, but the general guideline is - if a standard deviation is more than twice the average, I do not use that std dev in my average. That is why you and I have a descrepancy. And as you can see, it makes a decent difference (0.32 points in this case).

I do this for one main reason. When a game has a very high standard deviation, say 14 points, then the prediction services have a wide range of predictions. Its like saying 68% of the predictions fall in a 28 point range. 28 points! That's just simply too big of a range.

What I have seen is that when you have a very high standard deviation, the game usually involves a piss poor team. In this case, Army and Baylor. One of the predictions says Army will win by 54.5 points, while another says UAB will win by 41! I can't weed out every single bad prediction, but by not counting the very high standard deviations, I can weed out some of the poorly predicted games.

finsfan,
The page I get the scores from is here

Catdawg,

Yes, if there is a line move, then the star rating must be adjusted. I used the current lines from my book as of Wednesday evening.

 

[heleanth]

Thanks for taking the time to reply to my question, djchad. I understand and I think you're methodology makes sense.

Good luck this weekend and thanks again for going through the trouble of posting this interesting system.

 

[djchad]

Overall a very average (and disappointing) day for the system. The 3-star pick was demolished. Overall 50% of the picks win. 2 of the 4 "wrong team is favored" wins.

Hopefully much better results this coming week. Prelim numbers will be out Monday eve.


<pre>
Favors Line Avg Pred Diff Std Dev Play Result Score
Air Force -17.50 -25.95 8.45 6.15 ** L 26-34
Alabama 2.00 -2.88 4.88 5.39 * W 34-14
Arizona St -3.50 -8.35 4.85 4.60 * W 27-16
Baylor 30.50 27.67 2.83 11.11 - - 10-44
BG -21.00 -23.96 2.96 5.18 - - 38-20
Buffalo 6.50 2.95 3.55 3.65 * W 12-16
Cal 5.50 0.09 5.41 5.48 * L 13-24
Cent Mich 20.50 15.33 5.17 4.95 * W 18-23
Colo St -12.00 -17.53 5.53 4.71 * W 37-10
Idaho 4.00 1.16 2.84 5.10 - - 21-18
Indiana 9.00 1.95 7.05 4.51 *** L 14-45
Iowa St 13.00 7.39 5.61 5.46 * W 10-21
LSU -3.50 -12.51 9.01 5.41 ** L 7-31
NC State -1.00 -3.74 2.74 3.56 - - 38-6
Nevada 4.50 -2.81 7.31 5.00 ** W 47-50
New Mexico 6.00 1.06 4.94 4.96 * W 42-35
Notre Dame 10.00 1.36 8.64 6.85 ** W 34-24
Ohio St -4.50 -10.54 6.04 5.51 ** W 13-7
Rutgers 15.50 12.44 3.06 6.99 * L 14-45
SJ State 16.00 12.88 3.12 6.03 * L 8-45
Texas Tech 6.00 3.29 2.71 2.32 - - 13-37
UAB -8.00 -13.88 5.88 14.78 * L 29-26
UConn 8.00 2.37 5.63 4.24 * W 24-28
UTEP 2.50 -1.31 3.81 7.58 * L 0-20
Va Tech -28.50 -33.28 4.78 7.14 * L 20-10
Virginia 5.50 -1.65 7.15 5.36 ** L 15-23
Wash St -11.00 -16.31 5.31 5.35 * L 21-13
Wisconsin -2.50 -5.45 2.95 3.10 - - 42-24



</pre>


<pre>
Week W L T %
*** 0 1 0 0.000
** 3 3 0 0.500
* 8 7 0 0.533


Season W L T %
*** 7 1 0 0.875
** 13 5 1 0.722
* 36 34 1 0.514
</pre>