We dropped another 3.9 units last week, pegging us back to +1.17 for the season.
If we were running hot at +10 units or so two weeks ago, I think we might be running cold now. We beat the moneyline price on the Giants by some 30 cents, while the Cowboys closed around 1.83 after we bet them at 2.14. That’s some pretty huge expected value.
On the other hand the worst pick of the bunch was the over 55 in Atlanta which closed around 52 and went under easily. Digging back through the numbers, the standout thing is that Atlanta had faced the toughest slate of opposing offenses in the league and perhaps isn’t as bad as its surface stats suggest.
In fact, this concept plays into the first couple of bets today; the idea that defensive metrics at this point are more noise than signal. I retweeted some excellent research this week which essentially found that defensive numbers accrued to this point have very little predictive value for the rest of the season.
To which extent did the team performance during the first X weeks explain the team performance during the last 17-X weeks? I split into the 6 categories: offensive/defensive EPA per play/dropback/run.
We have 9 seasons of data, so there are 288 data points. Results: pic.twitter.com/YYuJiD3Dkl
— Timo Riske (@PFF_Moo) October 23, 2018
Pass offense, by contrast, is very predictive. As the chap who did the study put it: “After two weeks of play we know more about offensive performance than we will ever know about defensive performance.”
It means defensive numbers are far more reactive to who you’ve played, and if you’ve had a very tough schedule to date, you can expect that defense to regress positively.
This brings us nicely to the Tampa Bay Buccaneers, who come in ranked 31st in net yards per attempt allowed and dead last in Football Outsiders defensive DVOA. However, those numbers have come against the 2nd toughest strength of schedule per Warren Sharp, behind only the aforementioned Atlanta Falcons. Even last week the Bucs looked improved, in spite of, or probably because of, the departure of defensive coordinator Mike Smith who had an unhealthy obsession with stopping the run. Last week the Bucs held the Browns to 4.9 yards per play (ypp) and just a 39% success rate (that’s not good).
Conversely the Bucs offense could be very good. Jameis Winston led the league in first downs per pass attempt last year, and in 2.5 games this year he leads the league in 1st downs per dropback at 44% compared to a league average of 34%. The issue of course is he can be a litte interception-happy, but I’d rather bet on play-to-play success.
The Bucs tandem of Jameis Winton and Cameron Brate will look to do damage against an average Bengals D on Sunday.
The Bucs face a Bengals team with a defense that is broadly similar based on metrics and strength of schedule, but whose offense is notably worse (by around 0.5 ypp). Its gotten even worse in recent weeks as injuries to ancillary weapons like John Ross and Tyler Eifert, and centre Billy Price appear to have taken their toll.
After scoring 30+ points a week for the first month of the season, the Bengals are down to 17 points per game in their last three.
I suspect they won’t be able to keep up with the Bucs here.
The second bet has a broadly similar underpinning but at the other end of the spectrum. The Baltimore Ravens have the clear best defense in the league this year in terms of ypp allowed and ppg allowed. However, they’ve done that against the sixth-easiest schedule per Warren Sharp, and given the randomness of defense as outlined above, I think we should project the Ravens to be a top-ten defense rather than the standout league leader. After all, what standout players do they really have when you think about it? Some very solid corners and an ageing Terrell Suggs?
Panthers QB Cam Newton will have to feature prominently on Sunday night for Carolina to thwart the Ravens.
Anyway, the current price of Ravens -2.5 suggest there are around 5.5 points better than the Panthers, with that difference entirely fuelled by the difference in defense.
In fact the offenses are almost identical by passing efficiency, with the Panthers actually ahead by rushing efficiency.
It means if we regress the Ravens somewhat from their defensive peak, I’d make this line right around pick-em, suggesting the 2.31 available on the Panthers is a steal.
Next up, I’m having a small nibble on the Jaguars for a couple of reasons.
First of all, I do buy into the idea that the Jags’ experience in coming to London is a real advantage.
NFL players are a little bit like Formula 1 cars; they’re highly strung and need things perfect for them to perform optimally, in terms of food, sleep, preparation and travel.
The Jags know how to prepare for this trip, and they’ve done well here in recent years, going 3-0 and averaging 27 points. Conversely, this is the Eagles first ever trip abroad. On the field, we may be getting an inflated line here by virtue of Bortles’ recent struggles.
However, I’d argue his play in recent weeks is nothing new and we’re just in a bad Bortles patch.
Jaguars QB Blake Bortles looked calm and composed in his last trip to Wembley vs the Ravens.
Historically he’s so volatile he is as likely to be good this week as bad. In fact, Five Thirty Eight wrote this week that Bortles is performing right in line with his career norms this season, and has even improved his yards per attempt. And normal Bortles could well be enough in this game, with the Jags a significantly better team than the Eagles by adjusted net ypp. For the season the Jags are third in the NFL at +0.7 ypp, while the Eagles under Wentz are -0.4.
Throw it all together and the Jags are worth a small moneyline investment.
Finally, I’m taking the Seahawks in Detroit.
It’s a pretty simple numbers play as the metrics suggest the Hawks are the better team, and of course, they get circa 1.5 points simply for coming off a bye. With broadly similar schedules, the Seattle defense is fourth in the league in net yards per attempt allowed, while Detroit is 16th. Seattle is also more efficient passing the ball.
Seattle Seahawks QB Russell Wilson has been extremley efficient with his passing game this season.
Massey Peabody actually has the Seahawks as a favourite in this game, while Football Outsider thinks they’re the fourth best team in the league, so whatever your choice of metric, they’re the superior team. Fire away.
Earlier this week Brad Allen as ever joined host Nat Coombs, Mike Carlson, Matchbook’s duo of Aidan ‘Sully’ O Sullivan and Jesse May to preview some of week 8’s key games while Nat and Sully almost come to blows also over the Amari Cooper trade. If you haven’t already subscribed, search for ‘Matchbook Betting Podcast’ on your preferred podcasting app.
