NFL Week 7 | Coin vs. Machine
Not a single team is favored by 6.5 or more
(First time in at least 15 years if it holds)
OCT 19 2017
Week 1 - Week 2 - Week 3 - Week 4 - Week 5 - Week 6
Week 7 - Week 8 - Week 9 - Week 10 - Week 11 - Week 12
Week 13 - Week 14 - Week 15 - Week 16 - Week 17
Wild Card - Divisional - Conference - Super Bowl
Not a single team is favored by 6.5 or more
(First time in at least 15 years if it holds)
OCT 19 2017
Week 1 - Week 2 - Week 3 - Week 4 - Week 5 - Week 6
Week 7 - Week 8 - Week 9 - Week 10 - Week 11 - Week 12
Week 13 - Week 14 - Week 15 - Week 16 - Week 17
Wild Card - Divisional - Conference - Super Bowl
Last week gave fans another pretty interesting batch of NFL football. Underdogs went 11-3 against the spread (or 10-3-1 if you had Vikings-Bucs as a pick'em) and won nine of 14 games outright. They’re now a whopping 54-36-1 against the spread on the season (plus or minus a game or two depending on your final lines). Three teams were favored by double digits in Week 6 – two of them were upset (Atlanta and Denver) and one barely won (Washington).
Headed into Sunday Night Football, Week 6’s average margin of victory (7.1 points) was the lowest for a single week since Week 13 of 2012. Sunday night and Monday night’s games were decided by 13 and 14 points, bumping Week 6’s average margin to 8.0 points – right behind Week 5’s 7.9. That makes Weeks 5 and 6 the first back-to-back weeks with average margins of 8.0 or less since Weeks 1 and 2 of 2013.
After a slow start to the season, 35 of 60 games (58%) over the past four weeks have been decided by seven points or less. That’s the fifth-highest four-week clip in the past five years. Not too shabby.
Headed into Sunday Night Football, Week 6’s average margin of victory (7.1 points) was the lowest for a single week since Week 13 of 2012. Sunday night and Monday night’s games were decided by 13 and 14 points, bumping Week 6’s average margin to 8.0 points – right behind Week 5’s 7.9. That makes Weeks 5 and 6 the first back-to-back weeks with average margins of 8.0 or less since Weeks 1 and 2 of 2013.
After a slow start to the season, 35 of 60 games (58%) over the past four weeks have been decided by seven points or less. That’s the fifth-highest four-week clip in the past five years. Not too shabby.
|
Week 7’s Betting Lines Are Unique
This season’s betting lines have been interesting to track, too. As discussed last week, Week 5 finished with the second-lowest average line (3.1 points) since at least 2002. Week 6’s average came in at 6.4 points – highest since Week 17 of 2014. Week 7’s current average (4.1) is on the low side of normal.
But this week’s NFL betting lines are unique in their own right. As of Thursday night, Week 7's biggest favorites (Tennessee and Dallas) are laying only 6.0 points, which would the lowest maximum for a week since at least 2002. If one of those ticks up half a point, then Week 7 of 2017 will join Week 8 of 2016 and Week 1 of 2010 as the only three weeks since at least 2002 with a maximum line of 6.5. Pretty wild.
This season’s betting lines have been interesting to track, too. As discussed last week, Week 5 finished with the second-lowest average line (3.1 points) since at least 2002. Week 6’s average came in at 6.4 points – highest since Week 17 of 2014. Week 7’s current average (4.1) is on the low side of normal.
But this week’s NFL betting lines are unique in their own right. As of Thursday night, Week 7's biggest favorites (Tennessee and Dallas) are laying only 6.0 points, which would the lowest maximum for a week since at least 2002. If one of those ticks up half a point, then Week 7 of 2017 will join Week 8 of 2016 and Week 1 of 2010 as the only three weeks since at least 2002 with a maximum line of 6.5. Pretty wild.
On top of that, only one Week 7 team (L.A. Chargers) is favored by less than a field goal. That means that 14 of this week’s 15 games have point spreads between three and six points. Tough week for survivor pools.
|
Looking Back at Week 6’s Picks
The three-model machine won six and lost eight against the spread in Week 6. Lady luck fared a little better with all those underdogs covering, as the coin went 8-6. FiveThirtyEight’s “Elo” was the only model of the five covered here to crack 50%. R-P Excel batted below .500 for the first time since Week 1.
A few quick updates to note. First, with Marcus Mariota starting for the Titans, the Monday Night Football line settled at Tennessee (-6.5 or thereabouts), up from the no-Mariota placeholder of -2.5. At 6.5, the three main models liked the Colts (the other two still liked Tennessee). So Indianapolis was ultimately booked as a three-model unanimous selection (albeit only a Tier IV selection overall given the other models).
I had the Giants penciled in for exclusion given the lines account for injuries but the models do not. They would have been a Tier I “best bet,” and they beat up the heavily favored Broncos on Sunday night. I’ll continue to note potential exclusions in the bottom right of the image with their would-be tier in parentheses (see Green Bay in Week 7). But I’m going to bake the exclusions back into the standings.
There were two exclusions in Week 1 (DAL covered, IND didn’t) and two in Week 6 (NYG covered, IND didn’t). So rest assured I’m not fluffing the stats by putting them back in. Deciding which games to exclude is fairly arbitrary anyway. I’ll mention major injuries and potential exclusions going forward, but for the sake of completion (i.e., having a Tiered pick in every game), everything is included in the standings.
With that in mind, the three-model system went 1-3 against the spread on unanimous picks and is now 20-13 YTD. The five-model system went 2-1 on Tier I “best bets” (11-5 YTD), 0-3 on Tier II picks (15-19 YTD), 0-2 on Tier III picks (6-5 YTD), 1-3 on Tier IV picks (8-5 YTD), and 3-0 on Tier V “toss-ups” (9-8 YTD).
The three-model machine won six and lost eight against the spread in Week 6. Lady luck fared a little better with all those underdogs covering, as the coin went 8-6. FiveThirtyEight’s “Elo” was the only model of the five covered here to crack 50%. R-P Excel batted below .500 for the first time since Week 1.
A few quick updates to note. First, with Marcus Mariota starting for the Titans, the Monday Night Football line settled at Tennessee (-6.5 or thereabouts), up from the no-Mariota placeholder of -2.5. At 6.5, the three main models liked the Colts (the other two still liked Tennessee). So Indianapolis was ultimately booked as a three-model unanimous selection (albeit only a Tier IV selection overall given the other models).
I had the Giants penciled in for exclusion given the lines account for injuries but the models do not. They would have been a Tier I “best bet,” and they beat up the heavily favored Broncos on Sunday night. I’ll continue to note potential exclusions in the bottom right of the image with their would-be tier in parentheses (see Green Bay in Week 7). But I’m going to bake the exclusions back into the standings.
There were two exclusions in Week 1 (DAL covered, IND didn’t) and two in Week 6 (NYG covered, IND didn’t). So rest assured I’m not fluffing the stats by putting them back in. Deciding which games to exclude is fairly arbitrary anyway. I’ll mention major injuries and potential exclusions going forward, but for the sake of completion (i.e., having a Tiered pick in every game), everything is included in the standings.
With that in mind, the three-model system went 1-3 against the spread on unanimous picks and is now 20-13 YTD. The five-model system went 2-1 on Tier I “best bets” (11-5 YTD), 0-3 on Tier II picks (15-19 YTD), 0-2 on Tier III picks (6-5 YTD), 1-3 on Tier IV picks (8-5 YTD), and 3-0 on Tier V “toss-ups” (9-8 YTD).
|
Year-to-Date Standings
Three-model unanimous picks (“unanimous”) and five-model unanimous picks (“Tier I best bets”) remain strong on the year (despite a poor Week 6 for the former). Tier II picks – for which four of the five models agree – continue to struggle and have won only 44% of games against the spread this season. That underperformance muddies the system's water a little bit. We’ll continue to keep an eye on those.
Three-model unanimous picks (“unanimous”) and five-model unanimous picks (“Tier I best bets”) remain strong on the year (despite a poor Week 6 for the former). Tier II picks – for which four of the five models agree – continue to struggle and have won only 44% of games against the spread this season. That underperformance muddies the system's water a little bit. We’ll continue to keep an eye on those.
Looking Ahead to Week 7’s Picks
Week 7 features a ton of agreement among the models and promises to be a really interesting test case as a result. The three main models agree on eight of the week’s 15 games. All five models agree on five picks (Tier I “best bets”): N.Y. Giants (+5.5), Kansas City (-3.0), Dallas (-6.0), Philadelphia (-4.5), and Denver (+1.0). Four of the five models agree on another seven picks, representing Week 7’s “Tier II.”
Green Bay (+4.5) is up for potential exclusion given the injury to Aaron Rodgers. The Packers would otherwise be a Tier III pick. And there are a handful of lines to watch out for. The three-model majority (“machine”) likes Minnesota at -5.5 but Baltimore at 6.0; Buffalo at -3.5 but Tampa at 4.5; Cleveland at +6.0 but Tennessee at 5.0; Pittsburgh at -5.5 but Cincinnati at 6.5; and Philadelphia at -4.5 but Washington at 5.5.
Week 7 features a ton of agreement among the models and promises to be a really interesting test case as a result. The three main models agree on eight of the week’s 15 games. All five models agree on five picks (Tier I “best bets”): N.Y. Giants (+5.5), Kansas City (-3.0), Dallas (-6.0), Philadelphia (-4.5), and Denver (+1.0). Four of the five models agree on another seven picks, representing Week 7’s “Tier II.”
Green Bay (+4.5) is up for potential exclusion given the injury to Aaron Rodgers. The Packers would otherwise be a Tier III pick. And there are a handful of lines to watch out for. The three-model majority (“machine”) likes Minnesota at -5.5 but Baltimore at 6.0; Buffalo at -3.5 but Tampa at 4.5; Cleveland at +6.0 but Tennessee at 5.0; Pittsburgh at -5.5 but Cincinnati at 6.5; and Philadelphia at -4.5 but Washington at 5.5.
Tiered picks are based on the five models listed below (methodology)
Good luck and enjoy the week!
|
The five models referenced are created by and sourced from FiveThirtyEight ("Elo"), Jeff Sagarin via USA Today ("Rating" and "Offense-Defense Method Projections"), RP-Excel.com ("R-P Excel I"), and NutshellSports.com ("Nutshell Sports NFL Ratings"). Lines are sourced from FootballLocks.com as of Oct 19 at 4pm PT. Data was compiled and analyzed by ELDORADO. All charts and graphics herein were created by ELDORADO.
ELDORADO | Berkeley, CA | New York, NY
eldo.co | @eldo_co
ELDORADO | Berkeley, CA | New York, NY
eldo.co | @eldo_co