NFL Betting Picks - Week 2

Pick 1: Washington -5.5 correct NFL point spread pick
Pick 2: Arizona -2.5 correct NFL point spread pick
Pick 3: Kansas City +13 correct NFL point spread pick
Pick 4: San Francisco -7 incorrect NFL point spread pick

My wife was listening to NFL Radio on SiriusXM on her commute to work. As a psychiatrist, she started analyzing every caller and convinced me that my picks should be very relevant from the beginning of the season. Her argument is that people are very emotional at the beginning of the season because they have high hopes for their teams given the clean slate. She argues that could be a big reason why the first week picks were successful.

My response to that is that given I only analyze team level data, my models have a big assumption(that I have to continuously check and hope you help): team composition/competitiveness is nearly constant in recent games. That means that if for example St Louis has a new QB and not enough data has been gathered with Shaun Hill, then the Rams' computer spread forecasts should be avoided because their skill level is not the same as previous recent games.

We will start tabulating results this week. All are official picks from now on. To be more conservative, we'll start with a smaller bankroll percentage in the first few weeks 12% until week 5 where will step up to 15% and go up to possibly 25% towards the end of the season depending on success rates. Here are week 2 NFL point spread predictions:

GameVegas LineEstimatePrediction-VegasConfidence
JACKSONVILLE @ WASHINGTON -5.5 -9.3 -3.8 63.0%
ARIZONA @ NY GIANTS 2.5 5.6 3.1 59.8%
KANSAS CITY @ DENVER -13 -9.1 3.9 57.1%
CHICAGO @ SAN FRANCISCO -7 -19.0 -12.0 57.0%
PHILADELPHIA @ INDIANAPOLIS -3 -4.5 -1.5 51.4%
NEW ENGLAND @ MINNESOTA 3.5 8.7 5.2 51.1%
DETROIT @ CAROLINA -2.5 -5.2 -2.7 50.0%
ST LOUIS @ TAMPA BAY -6 -5.0 1.0 50.0%
PITTSBURGH @ BALTIMORE -2.5 -0.6 1.9 50.0%
DALLAS @ TENNESSEE -3.5 -2.3 1.2 50.0%
ATLANTA @ CINCINNATI -5 -10.5 -5.5 49.9%
SEATTLE @ SAN DIEGO 6 5.3 -0.7 49.0%
HOUSTON @ OAKLAND 3 -5.4 -8.4 49.0%
NEW ORLEANS @ CLEVELAND 6.5 13.5 7.0 48.3%
MIAMI @ BUFFALO 1 -0.2 -1.2 48.3%
NY JETS @ GREEN BAY -8.5 -3.8 4.7 47.4%

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: POSITIVE implies VISITING team will cover point spread.
  • Confidence: The probability that the point spread pick is on the correct side.

Comments

Walleye said…
I don't quite understand the confidence percentage. For example, in this week‘s NYJ @ Green Bay game the pred-Vegas is 4.7 (quite high) suggesting that the Jets will cover the spread, yet the confidence % is only. 47.4%. Does this mean you actually prefer Green Bay in this case?
Jaime said…
Not really although you could argue it that way. I never do these reverse bets, they haven't worked out for me. The confidence is based on historical predictions and the spread dating back to 2000.
Unknown said…
Jaime long time reader just wanted to thank you for taking time out of your personal life and posting interesting and good information.
Jaime said…
Thank you Adrian. Providing quality information is my top priority. Any benefit I get from this starts there.
Walleye said…
Thanks Jaime,I was just trying to get my head around the correlation (if any) between the two numbers.I`ve just recently come across your stuff and really like what your doing. Keep up the good work.
Unknown said…
Hi Jaime,

I stumbled upon your website, and the "experiment" you are doing with the bankroll and everything is the kind of mathematical/statistical thing I like to do when trying out something... I am however confused as to how to you bet on your picks... do you simply place an equal unit on each individual game? do you parlay any of them? do you bet more units on pick #1? etc...

Thanks!
Jaime said…
Kevin, great question. Currently I simply split the total bankroll equally. I don't parlay, but have thought about incorporating the confidence percentage into the bet.