NFL Week 17 | Coin vs. Machine
DEC 30 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
DEC 30 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
Welcome you to Week 17 of Coin vs. Machine. There's only 24 hours left in 2017, so let's get right down to it.
Looking Back at Week 16’s Picks
Sevens were wild in Week 16, as the three-model majority (“machine”), five-model majority, and five-model points systems all went 7-7-2 against the spread (ATS). Three of the five models (FiveThirtyEight, Sagarin Rating, Nutshell Sports) also went 7-7-2. It was the machine's second straight 7-7-2 week.
The most noteworthy Week 16 result belonged to the coin, which bad-lucked its way to 3-11-2 against the spread. That's the worst weekly mark not only for the coin, but among any of the five individual models or four derived model-based methods (three-model majority and points, five-model majority and points).
The coin's best week was a 9-4-1 result against the spread in Week 10, so good luck's upside wasn't quite as good as bad luck's downside was bad in Week 16. (The best week for a derived method was five-model points in Week 8, when it went 11-1-1 against the spread. Nutshell Sports went 10-2-1 that week.)
Three-model unanimous picks rebounded from a 1-3-1 mark in Week 15 to go 5-3-1 in Week 16. They're a modest 47-43-6 against the spread on the season. Five-model unanimous picks (Tier I "best bets") went 2-2 and still haven't topped .500 in a week since Week 10. But they're still an attractive 32-22-2 heading into Week 17.
(Jacksonville was asterisked as a Tier I "best bet" last week thanks to the Garoppolo effect. They gave up 44 points to Jimmy G. and failed to cover. I had asterisked Seattle as a Tier I "best bet" too, but they did manage to upset Dallas. My last name is Guglielmo and yet I still have to Google how to spell Garoppolo every week.)
Sevens were wild in Week 16, as the three-model majority (“machine”), five-model majority, and five-model points systems all went 7-7-2 against the spread (ATS). Three of the five models (FiveThirtyEight, Sagarin Rating, Nutshell Sports) also went 7-7-2. It was the machine's second straight 7-7-2 week.
The most noteworthy Week 16 result belonged to the coin, which bad-lucked its way to 3-11-2 against the spread. That's the worst weekly mark not only for the coin, but among any of the five individual models or four derived model-based methods (three-model majority and points, five-model majority and points).
The coin's best week was a 9-4-1 result against the spread in Week 10, so good luck's upside wasn't quite as good as bad luck's downside was bad in Week 16. (The best week for a derived method was five-model points in Week 8, when it went 11-1-1 against the spread. Nutshell Sports went 10-2-1 that week.)
Three-model unanimous picks rebounded from a 1-3-1 mark in Week 15 to go 5-3-1 in Week 16. They're a modest 47-43-6 against the spread on the season. Five-model unanimous picks (Tier I "best bets") went 2-2 and still haven't topped .500 in a week since Week 10. But they're still an attractive 32-22-2 heading into Week 17.
(Jacksonville was asterisked as a Tier I "best bet" last week thanks to the Garoppolo effect. They gave up 44 points to Jimmy G. and failed to cover. I had asterisked Seattle as a Tier I "best bet" too, but they did manage to upset Dallas. My last name is Guglielmo and yet I still have to Google how to spell Garoppolo every week.)
|
Looking Ahead to Week 17’s Picks
All of Week 17's games will be played at 1 o'clock and 4 o'clock on New Year's Eve. Seven of those 16 games are three-model unanimous and five are Tier I "best bets." (Philadelphia is a Tier I "best bet" without being three-model unanimous because four models like the Eagles at +3.0 and the fifth's projection matches the game's actual line, which is a Tier I corollary that comes up frequently. Washington is a Tier II pick based on the fact that three models like them and two models have the line spot-on.)
Week 17 is usually pretty weird given some teams have everything to play for, other teams have nothing to play for, some teams sit their starters, and others bench them once the game's in hand. Oddsmaking is no picnic in Week 17, but Vegas does its best to factor these considerations into their lines. The models, however, have no idea who's already clinched, who's going all out, and who's benching their starters. The Rams and Eagles, for example, are Tier I "best bets," meaning the models love them. But both teams' playoffs seeds are set. Nothing changes if they win, nothing changes if they lose. But the models see both as underdogs as think they'll cover.
You can read about who has what to play for here. Along with the Eagles and Rams in Tier I, the Chiefs and Jaguars have nothing to play for in Tier II. The Steelers can clinch the one-seed in the AFC if they win the Patriots lose, but with the Pats at home to play the Jets, it seems Pittsburgh is going to sit some starters. The models love Miami, but they face the Bills, who are a playoff long-shot but need to win to have any chance. And they love Arizona, but they play Seattle, which has a decent shot at the six-seed in the NFC. And so on and so on. I asterisked a lot of teams below (asterisk = proceed with caution), and you could damn near asterisk the whole board.
And of course, you'll want to keep an eye on your lines relative to what’s listed below. The three-model majority (“machine”) pick is Washington at -3.0 but the N.Y. Giants at +3.5; Minnesota at -12.0 but Chicago at +12.5; New England at -15.5 but the N.Y. Jets at +16.0; and Tampa Bay at +7.0 but New Orleans at -6.5.
All of Week 17's games will be played at 1 o'clock and 4 o'clock on New Year's Eve. Seven of those 16 games are three-model unanimous and five are Tier I "best bets." (Philadelphia is a Tier I "best bet" without being three-model unanimous because four models like the Eagles at +3.0 and the fifth's projection matches the game's actual line, which is a Tier I corollary that comes up frequently. Washington is a Tier II pick based on the fact that three models like them and two models have the line spot-on.)
Week 17 is usually pretty weird given some teams have everything to play for, other teams have nothing to play for, some teams sit their starters, and others bench them once the game's in hand. Oddsmaking is no picnic in Week 17, but Vegas does its best to factor these considerations into their lines. The models, however, have no idea who's already clinched, who's going all out, and who's benching their starters. The Rams and Eagles, for example, are Tier I "best bets," meaning the models love them. But both teams' playoffs seeds are set. Nothing changes if they win, nothing changes if they lose. But the models see both as underdogs as think they'll cover.
You can read about who has what to play for here. Along with the Eagles and Rams in Tier I, the Chiefs and Jaguars have nothing to play for in Tier II. The Steelers can clinch the one-seed in the AFC if they win the Patriots lose, but with the Pats at home to play the Jets, it seems Pittsburgh is going to sit some starters. The models love Miami, but they face the Bills, who are a playoff long-shot but need to win to have any chance. And they love Arizona, but they play Seattle, which has a decent shot at the six-seed in the NFC. And so on and so on. I asterisked a lot of teams below (asterisk = proceed with caution), and you could damn near asterisk the whole board.
And of course, you'll want to keep an eye on your lines relative to what’s listed below. The three-model majority (“machine”) pick is Washington at -3.0 but the N.Y. Giants at +3.5; Minnesota at -12.0 but Chicago at +12.5; New England at -15.5 but the N.Y. Jets at +16.0; and Tampa Bay at +7.0 but New Orleans at -6.5.
Tiered picks are based on the five models listed below (methodology)
Happy New Year!
|
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 Dec 30 at 10pm ET. 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