2016 NFL Point Spread Picks - Week 16

Pick 1: Seattle -8 incorrect NFL point spread pick
Pick 2: Atlanta -2.5 correct NFL point spread pick
Pick 3: Dallas -7 correct NFL point spread pick
Pick 4: San Diego -6.5 incorrect NFL point spread pick
Pick 5: Kansas City -4 correct NFL point spread pick

We continue to either win big or go home by continuing the 30% of bankroll. This puts us at $400 a game for the 5 games we have going on this week.

GameVegas LineEstimatePrediction-VegasConfidence
ARIZONA @ SEATTLE -8 -15.6 -7.6 72.4%
ATLANTA @ CAROLINA 2.5 13.9 11.4 62.5%
DETROIT @ DALLAS -7 -9.1 -2.1 58.7%
SAN DIEGO @ CLEVELAND 6.5 11.6 5.1 57.3%
DENVER @ KANSAS CITY -4 -10.8 -6.8 57.0%
CINCINNATI @ HOUSTON -2 -8.3 -6.3 55.1%
INDIANAPOLIS @ OAKLAND -4 -10.1 -6.1 53.5%
NY JETS @ NEW ENGLAND -16.5 -12.9 3.6 52.6%
NY GIANTS @ PHILADELPHIA 2.5 1.5 -1.0 52.0%
MIAMI @ BUFFALO -3.5 -7.6 -4.1 51.4%
SAN FRANCISCO @ LOS ANGELES -3.5 -3.0 0.5 50.4%
WASHINGTON @ CHICAGO 3 -3.0 -6.0 50.0%
TENNESSEE @ JACKSONVILLE 5 11.0 6.0 50.0%
TAMPA BAY @ NEW ORLEANS -3 -17.3 -14.3 50.0%
MINNESOTA @ GREEN BAY -7 -15.9 -8.9 49.4%
BALTIMORE @ PITTSBURGH -4.5 -3.0 1.5 48.3%

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.

3 comments:

fiftytwopointfour said...

Hi Jaime, the 72.4% confidence that your model has on the Seattle game is the highest I've ever seen. Is it a record-breaker? That leads me to wonder, have you ever looked back and seen if your, say, 65% confidence picks won about 60% of the time over the years, your 55% confidence picks won about 55% of the time, etc.

Fiftytwopointfour likes the following this week:
Bears (+3)
Cowboys (-7)
Bills/Dolphins (Over 41 1/2)
Cowboys/Lions (Over 44 1/2)

P.S. As of a few weeks ago, your comments system became more complicated/strict. That's why I haven't been bothering to deal with it up to this week. It would be great if you could figure out how to loosen it back up so commenting is quick and easy again!

Jonathon Grant said...

Going back to 2008, his combined record is 252-228, which is 52.5%. Considering he only ever plays games that have a confidence of 58% or higher, it's clear that confidence does not equate to probability of winning.

The way his model works is that each game goes into a bucket based on the Vegas line and his line. The historic performance of that bucket determines the confidence of that pick.

fiftytwopointfour said...

Ah! Thank you, Jonathon.