The Line: Ravens vs. Bills, Week 1

I don’t mean the O-line or D-line or the beer line. We’re talking betting line.

Among casual fans, something that is no longer taboo is sports betting. In some states, including nearby Delaware for you Baltimorians, you can place bets, risk free, just like you could when Vegas had the stronghold on the industry. Resident gambling expert Seth Bondroff has provided his plays for weekend but did not pick the Ravens/Bills matchup as one to bet on.

So let’s talk about the line for this matchup, kicking off the Ravens season, against the Bills on Sunday.

Discuss your thoughts on this topic, and more on our message board.

Vegas, offshore sites, other various places will use large amounts of data including numbers, statistical analysis and news to gives teams a rating and set an opening line for a game. From there the line can change based on current events and incoming money on each team in the matchups. The line will move in a way to generate ideally 50% of the money coming in on one team, and 50% of the money on another, so that they pay the winners with the losers money and profit from “the juice”. If 60% of the money goes to one side and 40% to the other, and the 60% wins, Vegas loses. In case you weren’t aware, the house never loses. In 2017 they had their best year ever, taking in $4.8 Billion in bets and profiting $248.7 Million. Vegas sportsbooks profited $76.8M just in football bets. Not to mention the other $90 billion or so bet illegally for those not in Vegas that states hope to tap into now that the federal ban is lifted.

There are many ways you can come up with a line on a football game but the one I’m going to look at here is analyzing how many points per drive (PPD) you expect a team to score. How many drives per game (DPG) an average NFL game will have. Information like this is found with a deep enough search of places like Pro Football Reference, Football Outsiders, and wherever else you can find numbers.

If you have a PhD in mathematics you can get really deep into statistics and talk about linear relationships, Kalman Filters, Correlated Gaussian Methods. I don’t have one of those. Algorithms are a foreign language to me, and if you could figure a way to predict the outcome of games like Biff Tannen in BTTF2, we’d all learn how real quick. But this is a pretty straight forward way to predict where a line should be around.

Since we don’t have a sample set for 2018 yet, we’ll use 2017 data to try to guess the line, then talk about what has changed between the teams over the offseason to see if there is something to exploit here.

Offensive drives per game (DPG)

In 2017 the NFL league average was 11.52 drives per team, per game. This is pretty standard over the last 19 years. 1998 was the first year of record on Pro Football Reference for this and there were 11.23 drives per team, per game. It went to 11.77 by 2002, 11.58 in 2007, 11.68 in 2012.

2017 Bills – 11.25 drives per game.  2017 Ravens – 12 drives per game

Defensive DPG

2017 Bills – 11.44 DPG.  2017 Ravens – 12.19 DPG

Offense – Points per drive (PPD)

In 2017, league average was 1.78 points per drive (PPD). Again, pretty standard. 1998 – 1.63 PPD, 2002 – 1.71, 2007 – 1.71, 2012 – 1.79. The difference in the 1998 number and the 2017 number equals 1.72 points per game based on 2017 drives per game.

2017 Bills – 1.65 PPD.  2017 Ravens – 1.88 PPD

Defense – PPD allowed

2017 Bills – 1.99 PPD allowed. 2017 Ravens – 1.56 PPD allowed

And now the math…

Predicting drives per game 

(Team offense DPG X Opposing team defense DPG) / League average DPG = Expected drives to occur in the game.

2017 Bills: 11.25 X 12.19 / 11.52 = 12 drives in this game

2017 Ravens: 12 X 11.44 / 11.52 = 11.92 drives in this game

Predicting points per drive

(Team offense PPD X Opposing Team defense PPD) / League average PPD = Expected PPD against that opponent.

2017 Bills: 1.65 X 1.56 / 1.78 = 1.45 PPD

2017 Ravens:  1.88 X 1.99 / 1.78 = 2.1 PPD

We’ll call this a 12 drive game, where the Bills score 1.45 points per drive and the Ravens score 2.1 points per drive. Let me stray off course for a second though.

I feel predicting drives in the game is difficult. For one, you want to eliminate kneel downs to end halves and games as “drives”, which Football Outsiders does to show you teams PPD. But also there is the variable of the coin toss, what a team choses to do, does a team get a drive to end the first half, score on that drive, then receive the ball to start the second, score again? It’s like if the offense threw a pick-six. Can’t predict those things. If a team throws three actual pick-sixes, they will get three more drives than the opposition by way of getting the ball back without giving the scoring team a “drive”. Turnovers are wildly unpredictable, often the turning point in a game, and the ultimate game changer for gamblers keeping an eye on the line.

Furthermore, the Arizona Cardinals led the league in 2017, averaging 12.63 drives per game. The Falcons were last, averaging 10.31 drives per game. On defense, the Falcons, because you often have to go offense, defense, offense, defense, offense….the Falcons had the least amount of defensive drives at 10.25 per game. The Cardinals had 12.5 defensive drives per game. So if these two extremes played each other….

Cardinals offensive drives predicted = (12.63 X 10.25) / 11.52  = 11.24 drives in the game.

Falcons offensive drives predicted = (10.31 X 12.5) / 11.52 = 11.19 drives in the game.

This game did not actually happen last year. So let us take the two offenses with the most drives, and defenses with the most drives. Arizona vs. Jacksonville

Cardinals offensive drives predicted = (12.63 X 12.81) / 11.52 = 14.04 drives in the game

Jaguars offensive drives predicted = (12.44 X 12.5) / 11.52 = 13.5 drives in the game.

In 2017 they did play this game and it was a 14 drive game. About what our equation predicted.

What about the two offenses with the least drives and defenses with the least, Dallas vs. Atlanta

Cowboys offense drives predicted = (10.69 X 10.25) / 11.52 = 9.51 drives in the game

Falcons offensive drive predicted = (10.31 X 10.69) / 11.52 = 9.57 drives in the game

In 2017 they did play this game and it was a 10 drive game. Rounding up from 9.5, about what our equation predicted.

Maybe there is something to trying to figure out the drives when it is two extremes playing one another. Likely safe to say that no team will have less than nine drives, and no team will have more than 14 drives in a game and if they do it is a crazy anomaly.

The 11.52 league average for drives per team per game is just the smallest fraction from being dead on between nine and 14.

Back to the Ravens, they had two games in 2017 where they only had nine drives, and a two games where they got to 14 drives. 15 in the OT loss to the Bears. But they hit close to the average of 11.52 with 11 or 12 drives, nine times in 16 games. Both the Ravens and the Bills are closer to the average for drives than the extremes.

For the sake of this exercise, we’ll predict the score based on the 11.52 number. Then again for round numbers of 10, 11, 12, and 13 drives.

If the Ravens and Bills play 11.52 drives (which we know they won’t) and multiply their expected points per drive in this matchup…

2017 Bills: 1.45 X 11.52 = 16.7.  2017 Ravens: 2.1 X 11.52 = 24.2

Predicted spread = Ravens (-7.5).  Predicted total for over/under = 40.9

Actual spread and over/under total….As of early Friday morning

Well looky there. Now I’ll apply the math to the round number drives in a game.

10 drives: Ravens 21 – Bills 14.5  (-6.5) 35.5

11 drives: Ravens 23.1 – Bills 16  (-7.1) 39.1`

12 drives: Ravens 25.2 – Bills 17.4 (-7.8) 42.6

13 drives: Ravens 27.3 – Bills 18.9 (-8.4) 46.2

Mind you, more drives don’t necessarily mean more points, less drives don’t necessarily mean less points. But it’s clear why the team who is the underdog, like whomever it is playing the Patriots, would be wise to do whatever it takes to keep the Patriots and Tom Brady off the field. Limit their drives. The Patriots led the NFL in points per drive last year at 2.69 PPD. You give them an extra drive, you’re lucky if they don’t put a field goal on the board. Maybe you stop them, but you give them a second extra drive, maybe they score a TD on you. In a 10 drive game with the Patriots you can expect them to score 27. In a 13 drive game you can expect them to score 35.

So there is your line…based on 2017 stats. This is 2018 though. What had changed between the two teams set to do battle on Sunday? The news of Nate Peterman starting, a lot of money coming in on the Ravens so they need to get some money on the Bills side? BSL’s Gabe Ferguson gives a preview of the matchup.

Looking at the offseason, the biggest change has been the quarterback position. Joe Flacco is still the signal caller. However all signs point to a healthy Joe Flacco who looked tremendous all summer long taking the field with a brand new set of receivers. Flacco dealt with back issues in the start of 2017, but as he got healthier so did his game. The first half of the season he threw six TDs vs. eight INTs. In the second half the threw 12 TDs vs. five INTs.

Opposite Flacco won’t be Tyrod Taylor, who led the Bills back to a long overdue playoff appearance. It will be 2017 fifth round pick Nathan Peterman. Who plays football probably as well as J. Peterman. He has two TDs and six INTs in about one games worth full of action spread out over a few games last year. Peterman getting the start is welcome news after seeing Josh Allen get drafted by the Bills in round one this year. Allen will have to wait his turn.

The rest of the Bills lowly offense seems to have gotten worse this offseason with an offensive line in shambles, no real threats in the receiving game, and heavy reliance on a 30-year-old LeSean McCoy to carry the load once again.

On defense they swapped out basically Marcell Darius for Star Lotulelei, trading Darius last year and signing Star in this past free agent frenzy. Trading up to draft Tremaine Edmunds could be a big deal as they are already calling him a cornerstone in Sean McDermott’s defense. One that ranked 18^th in points allowed and 26^th in yards allowed last year. 15^th in DVOA and 26^th in weighted DVOA with emphasis on the latter part of the season. Something tells me it’s going to take a little longer to turn that unit around.

The Ravens defense on the other hand is down Jimmy Smith from suspension, but looks deep at all levels and likely won’t miss Smith against the likes of Kelvin Benjamin and Zay Jones. Their philosophy under new DC Don Martindale looks more like the organized chaos looks Rex Ryan might show, rather than the Dean Pees defense which lacked in gam adjustments and was notoriuos for giving up points and leads at the end of halves and games. They struggle to just put teams away. The deep Ravens pass rush against a poor O-line should give a bad quarterback fits.

So if data and numbers and Vegas say the line is Ravens by 7.5 based on what looks like 2017 data estimations, and if the Ravens looked improved while the Bills look in transition at best, where are you putting your money? If the line says 24-17, I’ll take Peterman to throw a costly pick six instead of a touchdown and make it Ravens 31 – Bills 10. I wouldn’t be surprised if more money keeps coming in on the Ravens and the line eventually ends up at -8, -8.5 to try to get some money on the Buffalo side. 7.5 point spreads are tricky because you need to win by eight, which in football scores (often seven or three at a time) is often like saying you need to win by 10. If the money comes in on Buffalo and the line moves to Ravens -6.5, jump all over that and take the Ravens by the TD margin.

Keep an eye on the weather though too. 80% chance of rain. If this one gets out of hand early, expect the Ravens to sit on the ball in sloppy conditions and the Bills failing to do much through the air to play catch up. Offenses might struggle out of the gate. Last nights Falcons/Eagles matchup between some offenses with some weapons was a low scoring affair. The under 40 might be in play in something like a 20-3 game.