2009 NFL Point Spread Picks Week 10

Pick 1: New Orleans -13.5 incorrect NFL point spread pick -$90
Pick 2: Baltimore -10.5 correct NFL point spread pick + $81.82
Pick 3: Denver -3.5 incorrect NFL point spread pick -$90

It is time to bounce back from last week's debacle. Lets increase the bankroll by banking on the Tomato Cans of the NFL Theory. The tomato can phrase came from ESPN's Chris Berman, but the theory that the line is unwilling to give very high spreads to struggling teams playing against strong teams is a NFLpickles model discovery. These teams including the Rams, Redskins, Lions, Chiefs, and Browns are probably thinking what Marv Cook said: “We've got to find a way to win. I'm willing to start cheating.”

This week I was all smiles when the model not only predicted the spreads to be well above the line, but the confidence measure (probability that the point spread estimate is in the correct side) was highest for 3 games that will test this theory. Lets look at these 3 NFL games more carefully:

NO @ STL $90 - The best team in the league against possibly the worst team in the league and the spread is 13.5 or 14? The estimate is 36, might be a statistical aberration, but this price is not right. The Saints #1 in points per game with 37.9 and the Rams last with 9.6. I think NO's last two "weak" games has hurt people's confidence in the team, i.e. recency bias.

BAL @ CLE $90 - Baltimore has not been able to finish games against tough opponents. This 1-4 record in the last 5 games has brought this spread down much lower than it should be. Are we going to see a 34-3 blowout as their last encounter? Probably not. But whether Cleveland decides to change their QB or not, Baltimore will win by at least two TDs.

DEN @ WAS $90 - Am a reading the spread correctly or did someone forget to add a 1 in front? The Kyle that played in Chicago came out yesterday and threw 3 INT. This was bound to happen, but not against the Redskins. The Denver defense, 3rd in the NFL in points allowed, plays a big role on Sunday and shuts down the Redskins.

Without further adieu, here are the NFLpickles free point spread estimates for Week 10 of 2009.

GameVegas LineEstimatePrediction-VegasConfidence
NEW ORLEANS @ ST LOUIS13.536.022.566%
DENVER @ WASHINGTON3.511.27.759%
TAMPA BAY @ MIAMI-10-8.21.853%
DETROIT @ MINNESOTA-16.5-19.1-2.650%
BUFFALO @ TENNESSEE-6.5-0.9-7.450%
DALLAS @ GREEN BAY310.07.048%
SEATTLE @ ARIZONA-9-23.0-14.044%
ATLANTA @ CAROLINA1.57.56.043%
KANSAS CITY @ OAKLAND-1.5-3-1.543%

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


Unknown said…
I'm going to make a ton of bets this week, more games anyway.

Bears +3
Jaguars +7
Steelers -7
Bills +7
Vikings -17
Rams +14
Bucs +10
Seahawks +9
Packers +3

I'm still going with my model for what it is worth (not very much haha)

You will see I'm picking Rams +14. Why? I really think the saints would blow the rams out. I though the packers would blow out the buccaneers last week as well. So I'll stick with the model for right now. Honestly, I'm going to wait till Sunday to bet on this game, because I have a feeling they will increase this.

I agree with your Denver vs Washington pick, as does my model, but I'm only predicting a 4 point victory margin. Yes they have beaten some good teams this year. Are they good? Well they have Orton... and his lucky no interception streak might be over.

For whatever reason, my model has the most confidence for the Vikings -17. They have the fire power, and the Lions seem to be getting worse and worse after their win this season.

Well good luck to everyone this week!
Anonymous said…
What do you guys think about parlays? Think its worth it? All opinions welcome.
Jaime said…
Great question Brian and something I've been wondering myself for a while. I don't have an answer. Every time I try it, I have failed and when I don't use it it works. Should I consistently use it? Are the expectations better? Can anyone provide a more analytical approach than mine?
Unknown said…
If your model is consistent enough for you to make good picks, then by all means go for it. I don't know of a model that can do that though. So I'm against parlays, I always lose!
Unknown said…
90% of my bets are 3-team and 4-team parlays. Let’s take a look at a 3-team parlay. 3-team parlays pay out 6 to 1 odds. If you can pick 3 teams to cover 1 out of every 7 times you will break even. If you can do better, then you’re profitable.

What is the probability that all 3 of your picks will win this week? If you take your confidence (probability of covering) and multiply them you get .25 or a 25 percent chance that all three will cover (.66*.64*.59). If you wagered a dollar and got paid 3 to 1 odds then your expected value would be zero (.25*$3+.75*-$1). 25% chance of winning $3 and 75% chance of losing $1. You would be indifferent about the bet. If you are getting paid more than 3 to 1 then you should make the bet. Bookies pay 6 to 1 so the bet should be made. Do you agree?
Unknown said…
I'm changing my picks this week because I spotted a SAS coding flaw in the week 10 data (I misspelled a name and it when WACK, EAT ME!). Go figure right?

My new bets this week, really all I did was take a few out.

Steelers -7
Vikings -17
Buccaneers +10
Chiefs +2
Seahawks +9
Packers +3

Anthony, the problem with using the probability of covering, is that is that truly the probability of covering? Yes, if you have more then a 60% probability of covering you will win in the long haul. Dealing with NFL modeling, variability is a ^$#%^!&. While we use it to make our decisions since it is our best guess, I never have extreme confidence since the MSE is always so high.
Jaime said…
Hey Joel, I'm a SAS user/fan. Use it to crunch the numbers that go into this blog.

Anthony, I will think about this on the train on my way home and answer later.
Anonymous said…
Wow. I'm never going to make the mistake of wishing you extra-good luck again.

I like the picks this week.. I am still settling into my new home, but maybe I can actually contribute something besides noise in a week or two.
HappyBreathNet said…

I agree with all of your picks, though I only bet with you on one. I have best bet (2 units) NO -13.5. I do also have Den and Bal, but both with very low confidence. I have Jac +7 as my other best bet and CHI +3 and MIN -16.5 as my other line bets. I also have two over/under best bets this week.
HappyBreathNet said…

I'd like to elaborate on what Anthony said. Parlay betting increases action for a given wager amount. If you have a positive expectation (you believe in your model) it increases your expected win. It does, however, also increase your standard deviation. It decreases the probability of a winning week even while increasing your overall expectation, because part of your expectation is wrapped up in low
probabilities for big wins. As an example, compare our Week Four Probability Distribution with our Week Four Parlay Analysis.

If you're betting with money you can afford to lose and you have a positive expectation (such as Jaime's system or mine) then Parlay betting can be a nice way to increase your expectation value
without putting out a lot of capital up front. If you just want to feel good and have a few extra dollars at the end of the week, straight bets are a better option.
Anonymous said…
If you're interested in betting parlays, but also maximizing your expected bankroll growth, SBR has a Kelly calculator, which includes support for multiple simultaneous events.

I'm not a fan of the interface, but at least it's something and might give you a basis for considering value.
HappyBreathNet said…
Rough week for Tomato Can theory. I went 5/11 for my 2nd losing week in a row. ) :

We'll get it turned around. Good luck to Baltimore (especially Todd Heap) tomorrow.