NFL Wild Card Playoffs | Coin vs. Machine
JAN 6 2018
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
JAN 6 2018
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
Happy New Year and welcome to Wild Card Weekend!
Looking Back at Week 17’s Picks
The models struggled in Week 17, as anticipated. The three-model majority ("machine") picks went 6-10 against the spread, equaling Week 4 for their worst week of the season. The five-model majority and five-model points picks both went 4-12 against the spread. The coin, meanwhile, went 10-6 against the spread, tying Week 3 for its best week of the year. Three-model unanimous selections went an abysmal 1-6 against the spread, five-model unanimous picks (Tier I "best bets") went 1-4, and Tier II picks went 2-5. Long story short, Week 17 is a throwaway. The total standings I'm showing below are the results through Week 16.
The models struggled in Week 17, as anticipated. The three-model majority ("machine") picks went 6-10 against the spread, equaling Week 4 for their worst week of the season. The five-model majority and five-model points picks both went 4-12 against the spread. The coin, meanwhile, went 10-6 against the spread, tying Week 3 for its best week of the year. Three-model unanimous selections went an abysmal 1-6 against the spread, five-model unanimous picks (Tier I "best bets") went 1-4, and Tier II picks went 2-5. Long story short, Week 17 is a throwaway. The total standings I'm showing below are the results through Week 16.
|
|
Looking Ahead to Wild Card Weekend
Given there are only four games, I changed the table (below) to show which team of the five models "likes" against the spread. This provides a clearer picture across the five-model spectrum. Recall that Tier I "best bets" are those favored by all five models against the spread. They boasted a 59.3% win rate against the spread from Weeks 1-16, although the last time they were above .500 in a single week was Week 10. There are no such picks during Wild Card Weekend.
Tier II picks are those favored by four of the five models. There are none of those this week either. At current spreads, every game is Tier III or worse, meaning three models like them and two models don't. In the case of Buffalo and Jacksonville, the three-model majority pick and five-model majority pick flip flop, making it a Tier V selection. (Others could obviously swap, too, depending on which models are included in the three and the five. But these are the groupings we've used.)
The three-model and five-model points systems account for how different each model's implied lines are relative to the game's actual lines. (In other words, if one model really really liked a certain team, it could outweigh the fact that a majority of models liked the other.) This week, the point-weighted picks all match their majority counterparts. For what it's worth, Atlanta +6.0 is "most favored" by the models. On average, the five think the Rams should only be 2.35-point favorites.
As always, watch out for your lines relative to what's listed below. Sagarin Recent and Sagarin Offense-Defense both like Kansas City -8.5 but Tennessee +9.0, meaning that Tennessee +9.0 becomes a Tier II pick. Sagarin Offense-Defense likes New Orleans at -6.5 but Carolina at +7.5, which would swing Carolina up to Tier II as well. And FiveThirtyEight likes Tennessee +8.5 but implies a push at +7.5. I've included the Week 1-16 against-the-spread results for each method below.
Given there are only four games, I changed the table (below) to show which team of the five models "likes" against the spread. This provides a clearer picture across the five-model spectrum. Recall that Tier I "best bets" are those favored by all five models against the spread. They boasted a 59.3% win rate against the spread from Weeks 1-16, although the last time they were above .500 in a single week was Week 10. There are no such picks during Wild Card Weekend.
Tier II picks are those favored by four of the five models. There are none of those this week either. At current spreads, every game is Tier III or worse, meaning three models like them and two models don't. In the case of Buffalo and Jacksonville, the three-model majority pick and five-model majority pick flip flop, making it a Tier V selection. (Others could obviously swap, too, depending on which models are included in the three and the five. But these are the groupings we've used.)
The three-model and five-model points systems account for how different each model's implied lines are relative to the game's actual lines. (In other words, if one model really really liked a certain team, it could outweigh the fact that a majority of models liked the other.) This week, the point-weighted picks all match their majority counterparts. For what it's worth, Atlanta +6.0 is "most favored" by the models. On average, the five think the Rams should only be 2.35-point favorites.
As always, watch out for your lines relative to what's listed below. Sagarin Recent and Sagarin Offense-Defense both like Kansas City -8.5 but Tennessee +9.0, meaning that Tennessee +9.0 becomes a Tier II pick. Sagarin Offense-Defense likes New Orleans at -6.5 but Carolina at +7.5, which would swing Carolina up to Tier II as well. And FiveThirtyEight likes Tennessee +8.5 but implies a push at +7.5. I've included the Week 1-16 against-the-spread results for each method below.
Tiered picks are based on the five models listed below (methodology)
Enjoy the playoffs!
|
|
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 Jan 6 at 10am 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
