NFL Point Spread Picks Week 12 - 2013

Pick 1: San Francisco -6 correct NFL point spread pick
Pick 2: Chicago +1 incorrect NFL point spread pick
Pick 3: New Orleans -10 incorrect NFL point spread pick

All NFL picks this week are visiting team. I'm not too surprised with this week's NFL predictions. Atlanta has been terrible so I expect NO to blow them out. No big injuries that I know of except for Jay Cutler for which I think the Bears are better off anyways so I'm sticking with the prediction model's result.

We continue to use 18% of bankroll. For week 12 of the NFL, we have a total of 3 predictions for a total of $600 per game. Here are the NFL picks against the spread for week 12 of 2013.

GameVegas LineEstimatePrediction-VegasConfidence
SAN FRANCISCO @ WASHINGTON 6 13.3 7.3 61.1%
CHICAGO @ ST LOUIS -1 3.5 4.5 59.8%
NEW ORLEANS @ ATLANTA 10 19.0 9.0 59.0%
INDIANAPOLIS @ ARIZONA -1.5 7.4 8.9 57.7%
DENVER @ NEW ENGLAND 3 8.3 5.3 55.1%
NY JETS @ BALTIMORE -4 -7.4 -3.4 51.3%
DALLAS @ NY GIANTS -2.5 -2.3 0.2 51.2%
TENNESSEE @ OAKLAND -1 -10.4 -9.4 50.0%
CAROLINA @ MIAMI 4.5 0.5 -4.0 50.0%
PITTSBURGH @ CLEVELAND -1.5 -9.6 -8.1 50.0%
MINNESOTA @ GREEN BAY -5 -10.6 -5.6 50.0%
SAN DIEGO @ KANSAS CITY -5.5 -8.7 -2.2 50.0%
TAMPA BAY @ DETROIT -8.5 -10.9 -2.4 49.4%
JACKSONVILLE @ HOUSTON -10 -3.7 -3.7 48.8%

How to read the table:
  • Vegas Line: A NEGATIVE number implies the point spread favors the HOME team
  • Estimate: NFL Pickles' point spread prediction
  • Pred-Vegas: Subtraction of the previous two. POSITIVE implies VISITING team will cover point spread.
  • Confidence: The probability that the point spread pick is on the correct side.

Comments

darkside said…
Hi,
I have been a reader of your blog for the past couple years. Keep up the good work, and thank you for posting your weekly picks.

I recognize that you use statistics are the basis for your picks. Have you considered also using other factors such as Home/Road team etc to weigh whether one selection is better than another?

For example, this week one of your selections was NOR -10.
Now, there is no doubt that the Saints are clearly the better team than the Falcons, by any metric: statistics or even the eye test.
However, the Saints are not the same team on the road, statistically or win loss record or their ATS record.
This was a divisional game (both teams know each other quite well) played on a short rest, with the Saints coming off a bruising game against the 49ers.
In other words, NOR -10 being statistically a solid selection but perhaps, not as strong situationally?

I hope my comment is not taken as a criticism of your system. Rather, I am posing a question: can a statistical approach be augmented by taking into factors such as situations, trends etc?
Jaime said…
Dark side you have a valid point on integrating trends through metrics. Statistical models should capture trends per se, meanin that if home dogs are much more likely to win then it should be reflected in results. We do account for home advantage but not for each team. I tried that once but accuracy fell since now we were estimating 31 coefficients of home field advantage, one per team -1. I'm preparing a high levelverview of stats and trends and hope to publish it some time this year. I'll keep you posted.
darkside said…
Jaime,
Thank you for you response and I look forward to reading this review.

In the interim, I wish you the best of luck in your picks.
Unknown said…
I have enjoyed reading your blog. While I don't have a 10,000 bank role, I've been doing parlay's based on your recommendations. I want to start building up my bank role now for next season. Are you planning on keeping the bank role at 10,000 for next season too?
Unknown said…
Kourosh - You could factor up or down depending on what your own bankroll is. If you're starting with $1000, just bet 10% of whatever Jaime recommends.

Jaime - I had fired you a message a little while ago hoping to discuss smoothing techniques, don't know if you got it.

Another thing I've been meaning to ask about. I can't find the article, but I remember when you switched to the formalized "amount to wager" strategy. I can't remember exactly what the formula was, but I'm pretty sure it was designed to prevent catastrophe (i.e. losing everything). You've often been using 15% and 18% of bankroll each week. Was the formula supposed to calculate the amount to wager each week, or was it suggesting that you bet 15-18% per bet?
Jaime said…
I may have gotten it and forgot to reply. I'll find it and discuss. For the formula it's called Kelly's formula and suggests 15-18%, here is the link http://www.nflpickles.com/2008/11/betting-size-and-kellys-formula.html?m=1
Unknown said…
This comment has been removed by the author.
Unknown said…
Regarding Kelly's Formula, you say that it is for sequential wagers. However, because of Sundays (having games at the same time, or overlapping) bets are made simultaneously.

I get your point, but I think you're missing out on opportunity here. If you are confident in your 60% probability, then you would get more out of this formula if you were not to place the wagers simultaneously.

I'm sure you'll agree with me on that, and I suspect that one of the biggest limitations as to why you don't want to do that is because it is simpler to do it at once (only requires updating the website once a week, and wouldn't confuse others).

That being said, if you are playing n games this week, why don't you calculate what your wager on the nth game would be if you were to lose the first n-1 games, and apply that wager to all the games. It would be on the conservative side, and should (will?) be larger than what you are currently wagering. This logic would still be within the framework of Kelly's formula with respect to lowering the probability of ruin, and would offer a much bigger upside if your probability is correct.

Let me know your thoughts.
calvinewers said…
Thank you for the excellent content I did experienced studying it, I will be sure to save your website and definitely will come returning from again. I want to motivate that you proceed your excellent job, have a excellent day. nfl picks
Unknown said…
How can I use your statistics to just pick the winner of each game? I'm in a confidence pool where I chose my confidence that a team will win from 1 - 16 points.