Sports betting is all about finding value. While prop bets come to the forefront leading up to the Super Bowl each year, they are also offered on a weekly basis during the NFL season, both in Vegas and offshore.
Each week, ESPN Insider John Parolin will take a look at some of the prop bets offered before the weekend's games and recommend a play based on analytics.
Overall record: 9-3-1
Outstanding bets: SEA to allow 30-plus points in another game (minus-140)
Any team to score more than 56 points in a game (plus-600)
Antonio Brown not to catch at least five balls in the next two games (plus-150)
Note: All prop bets taken from Bovada's website.
Over/under: Andre Williams 75.5 rushing yards
(O/U -115)
Williams is no stranger to an increased workload, having rushed for 2,177 yards in his senior season at Boston College. He will be the primary back for the New York Giants on Sunday against the Philadelphia Eagles after Rashad Jennings hurt his knee in the third quarter last week. Headlines from New York this week have all referenced Williams' increased workload, but I still believe 75.5 yards is too high. After all, Jennings, the starter, was under that total in four of five games this season (albeit one because he had to leave early due to injury).
Williams is a strong north-south runner, but does not excel moving laterally, and that's where the Eagles are most vulnerable. Philly's defense has been much stronger inside the tackles (allowing 3.4 yards per rush, sixth-best) than outside (7.3, 30th) this year. Opponents have rushed between the tackles an average of 22.6 times against the Eagles this season. If you give Williams that total and apply his 3.19 yards per rush, he still comes up short of the number by about 3 yards.
You also have to consider Williams' third-down limitations. The Giants may use Peyton Hillis in some of those situations, which will hurt Williams' total. This hinders his upside, though you should keep an eye on the health status of Philly's inside linebackers. DeMeco Ryans (groin) and Mychal Kendricks (calf) have both been limited in practice this week, with Ryans returning to practice Thursday. Ryans is likely a difference-maker. He hasn't missed a game since joining the Eagles in 2012, so this is a bet on him playing.
The play: Under
O/U: Branden Oliver 70.5 rushing yards
(O/U -115)
Welcome to the scene, Mr. Oliver. The Chargers rookie had just 12 career carries before exploding for 114 yards on 19 rushes last week against the Jets. Ryan Mathews isn't coming back this week, and Donald Brown missed practice on Wednesday after suffering a concussion Sunday. The Chargers signed Ronnie Brown this week, but Oliver's performance Sunday was too much to ignore.
What does playing the Raiders (0-4) do for Oliver's chances? Oakland has allowed 158.3 rush yards per game this season, including at least 150 yards in three of four. The only team that didn't rush for at least 150 yards was the New England Patriots in Week 3, and their offensive-line problems were well-documented at that time. Teams have averaged 36.8 rushes per game against the Raiders, most of any team in the league. This team is as susceptible against the run as they come, and certainly more than the Jets. If Brown can't play, Oliver is easily the most dynamic of San Diego's options, which also include Shaun Draughn. The player who gets the snaps will get yards, and there's no reason for San Diego to force Brown back when Oliver was so productive.
The play: Over
O/U: Ronnie Hillman 55.5 rushing yards
(O/U -115)
Hillman is the starter with Montee Ball out, but what does that mean? ESPN Broncos reporter Jeff Legwold hit this very topic on Thursday, offering up some helpful splits on how Denver distributed its carries. Ball received 55.6 percent of the Broncos' rushes before leaving Sunday's game with a groin injury. This was fairly consistent with lead back percentages in 2013 (52.3 percent for Knowshon Moreno) and 2012 under offensive coordinator Mike McCoy (62.1 percent for Willis McGahee in Weeks 1-11 before he missed the last six games). Using 60 percent as a benchmark for Hillman seems fair enough.