Week 15 2019 picks
Welcome to Week 15 in the NFL. With 3 weeks remaining in the regular season, now is the time to make moves up the leaderboard. There seem to be more than a few decent plays this week but first let’s review the results of Week 14 mixed bag:
Raw algorithm picks: 1-4 (20%)
Unofficial plays: 2-4 (33%)
Official plays: 4-1 (80%)
For the record, the official plays last week were: CHI +3, BUF +5.5, CIN +8.5, WSH +12.5, NYG +9.5. The games I write about, for the most part, are the games I consider playing. The CIN-CLE game last week experienced some very late line movement up to kickoff that prompted us to take off NO and put on CIN. I don’t normally suggest doing that but it does go to show how fluid the market action is, particularly on Sunday morning. So considering I’m writing this on Sunday morning, you’re getting the latest information I have.
With that, let’s take a performance health check on the other algorithms I track versus my algo through week 14 of the 2019 NFL season:
FPI - 46.4%
DVOA - 48.1%
Sagarin - 49.3%
My Algorithm - 51.4%
Yikes, none of the models are performing well at all. There’s debate to be had as to why this season has ended up being more unpredictable than ever. Tony Romo thinks it has to do with the strong QB play—specifically the fact that there’s a large age dispersion at the most important position—and the older QBs are still performing at a high level. Whatever the reason, the models are having a hard time keeping up.
With that, let’s take a peek at the Week 15 games:
Raw algorithm picks:
NO -8
GB -4
MIN -2.5
WSH +4.5
ARI +2.5
Now let’s decipher the sharp and contrarian sides for today:
Sharp/Contrarian sides:
DEN (@KC) - Moderate - I’ve been hearing the radio personalities like DEN in this matchup. The public tickets are pretty even and the line’s moved towards DEN from +11.5 at the open to +10. At double-digits, the game looks appetizing to take DEN, but the SuperContest only has it as DEN +9…sometimes I feel like they shade the “Sharp” side but with the algo liking KC (84% win probability) this might be a pass on DEN +9 if it wasn’t for the snowy weather forecasted. DEN +9 could be a play.
HOU (@TEN) - Weak - Line movement here went from TEN laying 1.5 to now laying a full FG. Even ticket count. Algo leans TEN in this matchup (suggests the line should be TEN -4). My gut tells me this could be a game HOU shows up for and Tannehill leads the league right now in the football metric of the moment, Completion Percentage Over Expected (CPOE). I’m not sure how predictive that metric is to game outcome but TEN certainly has become a sharp team to watch. Lean HOU but depending on the other games, I’ll probably lay off.
CHI (@GB) - Moderate - The Packers opened as a TD favorite and have been getting the public backing at a 60/40 split, but the line’s moved sharply down to -4.5. I like GB with an 81% win probability but the eye test seems to think GB isn’t as good as their record. In fact, the Pythag W/L indicates the record differential of these two teams is closer than actual. Getting 4.5 points doesn’t seem to be enough and that’s actually why I like CHI here—both a contrarian pick and a Sharp side.
CAR (SEA) - Weak - Word on the street is, the firing of Riverboat Ron Rivera didn’t go over well in the Carolina locker room. And they’d be right if the on-field product last week was any indication. The line in the SuperContest seems fair when compared to my algo-expected line but the market hasn’t quite caught up. The Seahawks opened as 4.5-point favorites and have been bet to 6-point favorites off 5:1 (83%) tickets. Carolina hasn’t looked good at all and Seattle has beaten teams worse than them consistently this season. Still, laying 6.5 on the road isn’t a good spot so I will be passing here.
ARI (CLE) - Weak - I must write about this game since the algorithm suggested the play. There’s nothing I want here though. The Browns getting 60% of the public tickets predictably the line has moved from 2.5 to 3 in favor of the Browns. Yes, the algo thinks this is really a pick ‘em but Arizona’s offense looks like it’s been solved and the defense can’t stop anything. Pass.
JAX (@OAK) - Weak - Who can back the Jaaaaagggssss? Actually, who will be watching this game at all? The public is all over OAK at a 3:1 clip. The SuperContest and the algo agree on 6.5 being the right number. If the market gets to a full TD, following the public might be the right side here. As much as I like the mustachioed Minshew, I think you’re getting the better QB-Coach duo in OAK. This will not be a play this week.
LAC (MIN) - Strong - Chargers getting 2.5 in the SuperContest while the market is at 1.5. There’s value here even as the algo thinks the Vikings should be 4-point favorites. The Sharps have clearly weighed in here as well. The opening number was a full FG but has moved in favor of LAC even though they’re only getting 30% of the public tickets. I concede that MIN is the better team, but like the CHI-GB game, the Pythag here is raising a yellow flag. LAC has been historically unlucky (or however you want to look at it), while the Vikings’ record is about 1 win better than the numbers would suggest. The Chargers are a play this week.
DAL (LAR) - Moderate - Conflicting signals here. The algo likes LAR to win (56%), Uncle Colin picked them in his Blazing 6 this week. On the flip side, the model aggregate slightly favors DAL and the public is all over LAR at an insane 4:1 clip—suggested contrarian value on DAL. The line actually moved to DAL (pick em’) as I was writing this post. The spot is questionable here too: the Cowboys are playing for a division and while a must-win doesn’t mean they’ll win, I do think the motivation angle favors the ‘Boys. Additionally, the Rams are off a big SNF win last week over their divisional rival, SEA. Are the Rams finding their stride or will this be a let-down spot? Lean LAR.
PIT (BUF) - Weak - Public weighing in heavily here. Nearly 3:1 tickets on BUF and the line hasn’t moved. This was another trendy pick this week on the radio. SuperContest line is at PIT -2.5, the algo thinks PIT -1.7, and the market is currently at PIT -1. I lean PIT but it’s very weak.
IND (@NO) - Weak - As I wrote above, the algo likes NO as a top-5 play. This line should be 11.5 but the SuperContest has the Saints -8 and the market is at -9. I think this is another game where since the market hasn’t caught up on the SuperContest, it might mean the Colts are the side showing value. In addition, the Pythag is waving the yellow flag again here as the differential between these two teams is 2.6 wins—meaning the actual games back, 4, should really be 1.5. Colts are a play.
WSH (PHI) - Moderate - The algorithm likes WSH ATS this week (+2 suggested) but the market has moved slightly towards PHI. There’s been pretty even public participation on both sides here, though slightly favoring the Eagles. The SuperContest only gives 4.5 to the Redskins, so versus the market there’s no value to be had. Slight lean towards WSH but will not play this game.
Let’s summarize.
Unofficial plays: