2021 NFL Week Five: Projections, Sides and Analysis (Part Two)

43
2021 NFL Week Five- Projections, Sides & Analysis- Part Two

Now that we are through week four of the NFL season, the model is starting to have enough current data to utilize its power rankings. From this point forward the model will rely on simulations as well as its power rankings to find weak spots in the market. 

Since the power rankings are referenced often below, I include the latest power rankings here:

You can use the power rankings to create a raw expected spread (RES) for each game by simply combining the score for each team. Remember that is the RES is on a neutral field. Home field advantage is not included in the power rankings.  

You can also use this to predict next week’s lines. For example, the RES indicated that the Buccaneers should be -8.99 on a neutral field against Philadelphia. The game is in Philadelphia. If you give the home team 3 points, you have a line of Bucs – 5.99, or -6. The look-ahead line is currently Bucs -6.5.

And now for something completely different

The model is placing a two-team 6 point Round Robin Teaser at -110 on the following sides. Normally, I would deem this approach misguided. However, the model has all four teams over, at, or near 77% to cover with the 6 points. That results in a 59.29% chance to win each bet, a 6.91% edge on each ticket. Too much to pass. If you can only find -120, then the model says it still has value, 4.74% edge.

Dallas -7.5

Seattle +2.5

Minnesota -7.5 

Cleveland +2

Bills +2.5

That is a total of 10 bets at ½ unit each. As long as you are teasing down from 7.5 to 8.5, or up from 1.5 to 2.5 the model finds value. Otherwise, just bet the team straight at the number. I debated long and hard about leaving Seattle off this list. The model finds them more likely to lose significantly than any other team. But it still has the game most likely decided by three.

What follows will only be single sides. 

New England Patriots v Houston Texans (9.5, 39.5)

Patriots 25.42 – Texans 11.30

The mode’s RES has the Patriots as -7.8 on a neutral field. The line is -9.5. The model simulations expect a 14.42 point Patriot victory. 

This game is very tempting especially considering two factors. One, Davis Mills’ performance in his two games. Two, Belichick’s record against rookie quarterbacks. Still, it is a massive number for a road team. 

Since I want a piece of this game that rests on Houston being bad, I am going to take a different angle. I am going to look at odds for the Patriots pitching a shutout. I will be stunned if the Texans blitz Jones more than a handful of times. They will drop 7. Jones is too good against the blitz.

This game will be updated on Friday if the model decides to take a position. 

For now, the model is not taking a position in this game. 


Chicago Bears v Las Vegas Raiders (-4.5, 45.5)

Bears 17.44 – Raiders 24.57

The model’s RES makes the Raiders 2.28 favorite on a neutral field. The spread is listed at -4.5. The model’s simulations have the Raiders winning by 7.13 points. Some of the Shine may have been taken off the Raiders with their loss to the Chargers, but I disagree. If Carlson makes the 52-yard field goal with 10:42 remaining in the fourth, it is a 21-17 game. The Raiders are legitimate, and the model loves them at +146 to make the playoffs.  

Nagy has finally given in to reason and announced Fields as the starter the rest of the way. On top of that, it appears Bill Lazor will be calling the plays instead of Nagy. Those are both great news for the Bears. 

The model is not taking a position in this game.


Cleveland Browns v Los Angeles Chargers (-1.5, 47)

Browns 23.77 – Chargers 19.07

The total has moved dramatically down since it opened at 49.5. The model jumped it, but the value on that side is gone now. The model’s RES makes the Browns a 1.23 point favorite on a neutral field. So the line at Chargers -1.5 is only in line with expectations if you give the Chargers a 3 point boost for their home field.   

As we saw last week, the Chargers do not have much of a home yield advantage, but that was the Raiders and this is the Browns. The simulations see 4.7 point favorite. 

I love Herbert. Twice I have stated that If all the players were put into a common draft and I had the first pick, I would take him and his rookie contract over Mahomes. I backed him in week one, but faded him, unsuccessfully (Newman!), last week. I am going to fade him again.  The under also has value. 

The model is taking the Browns +110 MLB


New York Giants v Dallas Cowboys (-7, 52)

Cowboys 33.89 – Giants 18.19

The model’s RES (raw expected spread) makes Dallas -4.64 on a neutral field. This number was -8.5 at the look-ahead and worked its way down to -7. The simulations have Dallas as a 14+ point winner. The model is admittedly slow to trust Mr. Jones and has such a negative view of Jason Garrett … it is almost as if the model dislikes the way Garrett calls a football game so much, it has become personal.  My model is self-aware, but only to the extent that it knows it distrusts Garrett. 

I have the Cowboys in a teaser at -7.5. I probably should have held off and just waited for them to drop to 7 and taken them ATS. 

The model is not taking a position in this game. 


San Francisco 49ers v Arizona Cardinals (-5.5, 50.5)

49ers 26.49 – Cardinals 30.82

This game was a pick’em before the season began. The look-ahead line was Arizona -2.5. Now the line Arizona -5.5. I point that out to suggest that this line might be a bit inflated after the Cardinals dismantled the Rams. The model’s RES suggests the line at a neutral field should be Cardinals -1.3. The line appears to give Arizona a whopping 4.2 points in home-field advantage. The simulations suggest a line of Cardinals -4.33.

If it is Trey Lance, it will be different than what we have seen. It will be the entire game plan designed for his skillset. Every Shanahan loyalist (I am not a card-carrying member, but I am thinking about signing up full time) will be tuned in to see the offense. This should be something unique, for which Arizona will have zero preparation. This should be elite-level play and series sequencing.

If you watch this game I encourage you to try to guess each offensive play by the 49ers. You can start basic run/pass, left/right/middle. But notice how they will (I believe) run so many different plays from the exact same look).   

Murray is the MVP leader after four games. When he gets the ball out quickly, he appears to have very few limitations. 

I will take the over on Trey Lance rushing yards  


Buffalo Bills v Kansas City Chiefs (-2.5, 56.5)

Bills 27.76 – Chiefs 30.64

This line is fishy. The model’s power ranking RES makes the Chiefs a 1.18 favorite on a neutral site. The Chiefs have a legitimate home-field advantage. The model gives them 2 points at home. That works out to a line of Chiefs -3.18, or -3.5. It is really hard for me to understand why this line started at -2.5. Under the key number of 3.  It has moved to -3 at many places, but it will not move to 3.5. The Chiefs are only on 41% of the spread bets, yet represent 56% of the handle on the spread. The Chiefs are only 25% of the moneyline tickets, but 61% of the cash. The books courted Chiefs action with these numbers. 

The simulations have the Chiefs -2.88.

I have the Bills in a teaser from +2.5 to +8.5. I am not interested in anything else except letting the teams play this one out.  Everything about this line is fishy to me. 

I am not taking a position in this game. 


Indianapolis Colts v Baltimore Ravens (-7, 46)

Colts 19.58 – Ravens 26.34

The model sees this game nearly identically to the book. It would have made the exact same line and total. The RES says Ravens -4.95 on a neutral field. The simulations say Ravens -6.76. 

There is no value in this game. 

The model is not taking a position in this game.

About the author:

+ posts

I write about data and sports. I created my first model in 1997 using nothing more than Excel. Currently, I have data-driven models for the NFL, NBA, and World Cup Soccer.

Mathematics is the music of reason.
— James Joseph Sylvester, English mathematician

Leave a Reply