Man vs Machine: Pickswise expert vs the computer model NFL Week 16 best bets
Man vs Machine
Welcome to the 2021 edition of Man vs Machine. This NFL season, I’ll be making my five best NFL bets each week. Our Pickswise supercomputer will make its best five bets, and then we’ll compare. May the best man (or machine) win.
Our computer model gives probabilities and projections for the main NFL betting markets (spread, total & money line), best bets for spread, total, money line, and same game parlays. It also features prop bet probabilities for first and anytime touchdown scorer, plus individual player projections for passing, rushing, receiving yards, and touchdowns for betting and fantasy use.
For more information on how the supercomputer works and how it generates picks, scroll down to the bottom of this page. Now let’s look at the picks.
My 5 best bets
Jaguars +1.5 over Jets
Call me crazy, but I like the Jaguars a lot this week. Trevor Lawrence hasn’t had an ideal rookie season by any stretch of the imagination, but he hasn’t looked as lost as Zach Wilson. With the Jets dealing with a Covid-19 outbreak, as well, I think Jacksonville will pick up its first win of the post-Urban Meyer era.
Chargers -10.5 over Texans
I don’t often take double-digit favorites, but this one shouldn’t be close. The only thing making this Texans offense watchable recently has been Brandin Cooks, and now he’s out with Covid-19. About half of Houston’s defense is also out, so Justin Herbert should shred the Texans.
Vikings +3.5 over Rams
This number is just too high for the Rams to lay on the road. Everyone has jumped back on the Los Angeles bandwagon after just having buried this team a couple of weeks before. The Rams are now being overvalued.
Broncos pick ’em over Raiders
I don’t think Drew Lock is much of a downgrade from Teddy Bridgewater. And the Broncos’ ground game should have some success against a weak Raiders run defense. Denver’s defense is trending up, while Las Vegas’ offense is trending way down.
Washington +10 over Cowboys
Give me Washington on Sunday Night Football. Dallas’ offense isn’t playing well, so there’s no way I could lay double-digits with them here. They only won big last week because Mike Glennon pulled a Mike Glennon for the Giants, as Dak Prescott had yet another lackluster day.
Jaguars +1.5 over Jets, Bills +1 over Patriots, Vikings +3.5 over Rams, Broncos pick ’em over Raiders, Steelers +10 over Chiefs
The computer and I are thinking alike this week. We’re on three of the same sides: Jacksonville, Minnesota, and Denver. Interestingly, our system is going with the Bills over the Pats in the slate’s most highly anticipated matchup. It’s also willing to roll with Ben Roethlisberger at Arrowhead. Both myself and the computer went 3-2 last week; hopefully we can each have another winning record in Week 16.
2021 season results
Pickswise Expert (36-39)
Pickswise Computer (44-30-1)
How are the pre-game win probabilities calculated?
The Pickswise supercomputer creates pre-game probabilities using machine learning techniques as inputs to a Monte Carlo approach. This means we simulate a sporting event 10,000 times, catering for the deep range of outcomes and probabilities within a given event. We first predict the performance of each individual player based on hundreds of algorithmic variabilities, which we use to run our event simulations.
Why do the win probabilities and predictions change over time?
The probabilities are dynamic in nature and get updated as new information becomes available. This includes but is not limited to team news, weather conditions, and moves in the betting market.
What does a ‘value bet’ mean?
A value bet is where we have identified that the probability of the event occurring is more likely than the listed sportsbook odds. Taking these bets gives you an increased chance of being a successful gambler.
How is the change in win probability calculated during a live game?
Live win probability is a statistical metric that projects the team’s likelihood of winning at any given point by simulating the remainder of the game. Based on the performance of historical teams in the same situation and the current state of the game, the model updates in real-time following every major event during the game.