2014 NFL Betting Picks - Week 12

Pick 1: Denver -7.5 incorrect NFL point spread pick
Pick 2: Kansas City -7 incorrect NFL point spread pick
Pick 3: New England -7 correct NFL point spread pick

Every week is a learning experience. In retrospect, the New Orleans pick was a bad one given that the difference between prediction and vegas was small and when tinkering with the weights a bit, it would provide much different results. This means that the model was relying to heavily on recent success for the Saints. The Denver pick I think was the right one given the circumstance. I was a bit shocked with not the cover but the straight up win. I smelled blood in the Sunday night and Monday night game and again looking back I should have notified you through Twitter to go at if you had the chance. It wouldn't have changed the results here, but meant to tell you that I felt more comfortable with those picks than the confidence explained. Anyways, after this week we are at 56.8% ATS for the season and 8.1% ROI.

This week we have 3 solid premium picks. There are some free computer-based predictions that don't look too bad either. We have the Thurday night game and two day games on Sunday. Some notable mentions in the free picks include Arizona and Cleveland. This week we again use 18% of bankroll split evenly across our 3 premium games giving us a bet of about $650 per game. Below is our table with free NFL point spread predictions.

GameVegas LineEstimatePrediction-VegasConfidence
MIAMI @ DENVER -7.5 -14.7 -7.2 62.9%
KANSAS CITY @ OAKLAND 7 -15.6 -8.6 58.6%
DETROIT @ NEW ENGLAND -7 -12.3 -5.3 58.0%
ARIZONA @ SEATTLE -6 -2.2 3.8 57.4%
CLEVELAND @ ATLANTA -3.5 -1.4 2.1 55.6%
CINCINNATI @ HOUSTON -1.5 2.4 3.9 55.3%
JACKSONVILLE @ INDIANAPOLIS -14 -18.6 -4.6 54.3%
BALTIMORE @ NEW ORLEANS -3 1.7 4.7 53.8%
TAMPA BAY @ CHICAGO -6 -7.4 -1.4 52.5%
GREEN BAY @ MINNESOTA 10 8.3 -1.7 50.0%
WASHINGTON @ SAN FRANCISCO -9 -7.3 1.7 50.0%
ST LOUIS @ SAN DIEGO -5 -6.6 -1.6 50.0%
DALLAS @ NY GIANTS 3.5 4.5 1.0 49.2%
NY JETS @ BUFFALO -4.5 -5.1 -0.6 48.7%
TENNESSEE @ PHILADELPHIA -11 -15.9 -4.9 47.3%

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.

Comments

TotalBS said…
Games With Value Of 3 Or More:
BUF -4.5
CLE +3
MIN +9.5
SD -4.5
ARZ +6.5 (Best Bet)
WAS +9
DAL -3 (Best Bet)

johnbart said…
My line has the following on the top 3 picks:

@DEN (-3.7)
@OAK (+8.9)
@NE (-8.8)

I like ARI(+6.5) and CLE (+3) but I don't see +3.5 anywhere for CLE. I think it might eventually get there the way the money is coming in but it doesn't look to be widely available yet.

Best of luck all.
TotalBS said…
Who cares about what we think. It is ffgeeks we need to listen to!!
johnbart said…
Word! Let the dogs bark FF!
TotalBS said…
Classic example why using "systems" with STATS ONLY is dangerous.

Everyone knew the Chiefs statistically was the better team.

But like I said before games are decided by HUMANS not stats. And last night human factors TRUMPED the stats.

Chiefs off HUGE win...Check
Chiefs traveling to West Coast.. check
Chiefs less time for preparation...Check

Raiders desperate for win... check
Raiders on national TV...check
This game is like a Super Bowl for the Raiders.... Check

Jaime wants to make his model better?..Simple, you need to factor things like this in IN CONJUNCTION with stats. If you don't you are likely to lose a lot of this type games.
johnbart said…
Well you don't know if his numbers actually picked KC but it's a strong bet that they did.

I agree this was a huge trap game for KC. IMHO the way to use a stats model is to use it to set your line to find base value and then check basic emotion/situation criteria to make sure it's still a play. You still only bet on a team if the numbers are right so you need bet against the numbers.

A good book that sets out that approach and provides a number of good emotional situations is "Beat The Sports Books" by Dan Gordon. I have no idea if it's still in print but that's where I got most of my situational checks from. I don't use the power rankings he describes at all since I have my own models for that.

FFGeeks said…
This comment has been removed by the author.
FFGeeks said…
Only 3 games meet all my "best bet" criteria this week.

TB +6, PHI -11 and SD -4.5

I have to come up with a 4th for my pick 4 club, so I'm going with ARZ +7 as my 4th pick.

I agree with TotalBS. I pick about 80% by stats. The ARZ game makes me nervous. My stats pick ARZ but if SEA wants any chance at a home playoff game, it must win on Sunday.

TotalBS said…
All the top handicappers in the Hilton Contest are ALL on Seattle this week....

These guys are extremely talented handicappers and when they all agree they know something.

Best of luck to all...
FFGeeks said…
"All the top handicappers in the Hilton Contest are ALL on Seattle this week...."

TotalBS That actually makes me feel a little MORE confident. Looks like SEA is the #1 pick overall. We call it the CURSE of pick 4 club. The team everyone is on usually loses (PIT in week 10) !! Ha Ha
TotalBS said…
No, it is a curse if it is the number one pick for all the contestants I am talking about when the top 10 all do that means something usually.
johnbart said…
Larry is a game time decision for the Cards which is probably part of it. The other big thing is that SEA still has to travel to ARI after this game so they need to win this one. If ARI loses they still have the home game to for taking control of the division.

This is a must win game for Seattle in even sense of the word. With that said the line is still too high. The Hawks don't need to cover to win.
Unknown said…
I'm not convinced that when all the top handicappers on the Hilton contest agree that it means much. Further, I'm not even convinced that these guys are top handicappers. While they do seem to have some good weeks, when I look at the results of the Top 5/10 any given week, they are typically around 50/50.

I don't have standings to look back any further, but last week, here is their performance:
Leader: 3/5 (60%)
Top 2: 6/10 (60%)
Top 6: 16/30 (53%)
Top 12: 25/60 (42%)
TotalBS said…
Johnathon I have followed that contest for years, and believe me when they (the top10 not ALL) are all on the same side there is a reason.

You can believe what you want, but I know what I speak.
Unknown said…
TotalBS - You may be right. I just started following it this year, after someone on here (probably you) mentioned it. I need to get comfortable with it.

While the top player is 42-13, the bottom player is 17-38 (of those who have picked all games). There's no way that the top guy actually picks with a 76% success rate. There's a luck component to his success rate (maybe he's 60 or 65% in the long run? He could also be closer to 50% with a high degree of luck).

I'll get comfortable with it eventually - I was stunned a few weeks back when they all agreed on NYJ over PIT - but I just need to figure out how to best use it.
Unknown said…
For anyone interested, here are the top selections of the Top 10 in the Hilton contest:

NYJ+4.5 6-0
SEA-6.5 7-1
BAL+3.5 5-0
MIN+9.5 4-0
DAL-3.5 3-0
TEN+11 2-0
ATL-3 3-1
JAX+14 3-1
CIN+1.5 2-1
MIA+7 3-2
NE-7 2-2
SF-9.5 1-1
johnbart said…
Obviously the pros know what they're doing :)

Larry not playing for ARI took away a big offensive weapon. The D played strong and gave them a shot but they couldn't get anything going on offense.

What are the chances there was an inside scoop on the injury above and beyond guessing?

Either way a good week for the dogs except for TB which had the better game minus two picks. That's the way the ball bounces though :)
Unknown said…
I'm sure that there are some people in the contest who know what they are doing. But who's to say that those guys are the leaders. I haven't counted, but there's around 1000 people in that contest. If I took 1000 quarters, and flipped each 55 times, it wouldn't be terribly surprising if one of them hit heads around 40 times. Doesn't mean there's a 75% chance of it happening next time.

The top 10 I posted above were picking at 71% going into this week. Through SNF, they have picked a combined 54% this week (21-18). Does that seem more consistent with a pro, or a coin toss?

Sure, they mostly agreed on the Seattle victory, but without being able to say that the Top 10 are indeed better than everyone else, I struggle to see how this is any more meaningful than attributing it to Seattle being the most popular pick among all participants this week.

Does anyone know if there is an archive of weekly picks (for this year and/or prior years) anywhere?
FFGeeks said…
Back to reality.... (1-2 on "best bet" picks)

I have been looking at my "system" for a way to take into account 3 turnovers in the TB/CHI game that gave CHI two very short field drives, to turn around a 10 point TB victory into a spread covering 8 point loss. The answer is, there ISNT a way!!

Just like their isnt any way to take into account C.J. Andersons decision to give himself up on the way to a spread covering touchdown in the DEN/MIA game (I had MIA in that one).

Vegas makes adjustments to game lines in order to push bettors one way or another. Their goal is to have an equal number of bettors on both side. That is how they make money. And as I said before, they are battling bettors tendancies to go with the favorite. So , how much does the spread actually have to do with past performance?

I will continue to try and "predict" game outcomes, but the bottom line is, "its all about that bass", I mean "luck".
johnbart said…
Football is a sport played by people. You're never going to be able to predict anything 100% due to human errors or random luck. As a handicapper you are just looking for an edge that allows you to over come the luck in the long run.

If TB and CHI played each other head to head 10 times in a row on Sunday how many would TB win against the spread? If your answer is more than five then you made the correct bet even though you lost. Yes it sucks in the short term but you'll get some lucky wins as well. If you're a winning handicapper then you'll have more bad luck games than lucky games due to the fact that you're betting on the correct side more often.

As for the DEN/MIA game Anderson giving himself up on that run was the best play he could have made as it locked up the victory. It allowed them to run out the clock with nothing riskier than a kneel down.

If he goes in for the TD they are up by 10 but MIA has the ball back with a minute left on the clock where they can get a TD/Onside Kick/TD to win. Yes the odds of that happening are tiny, maybe 0.5% but 0.5% is still greater than 0%.
Jaime said…
All of you have made this year the best for this blog. There is still time to make it the best money wise, although I'm starting to run out of time :)

I appreciate all the well thought comments coming out every week. It's a tough business and yes models can't account for everything, even if you have the variable it might be insignificant.