Last season, IBM introduced IBM Insights, a tool that uses Watson AI to analyze thousands of fantasy articles, blogs, websites and podcasts and provide support data to assist with fantasy football decisions.
Watson outputs an estimated scoring range for each player and projects the chances that a player will exceed the upside estimate (e.g., "boom") or fall short of the low-end estimate (e.g., "bust") on any given week.
For those who are looking for a tool to concentrate various web opinions into one location to help make roster moves, IBM offers Watson's Insights as an unsupervised, quantitative supplement to the more in-depth, personalized analysis that you would normally expect from an ESPN analyst.
The following article points out some of the recommendations that Watson has for Week 3 of the NFL season, as of Friday afternoon.
Wide receiver decisions
Projections higher than rank
While wide receivers can provide some of the best fantasy value in a given week, it can be difficult to decide beforehand which receivers to start. Because most of the 32 teams in the NFL regularly run three-WR sets, there are a plethora of options available each week who have the potential to score. IBM Insights provides information that can help make these decisions easier.
With the general assumption that the top 10 wide receivers are weekly starts, the following analysis was performed on the 30 wide receivers ranked by IBM Insights as being below the top 10 who are expected to play (and did not already play on Thursday).
Marquise Goodwin, San Francisco 49ers vs. Kansas City Chiefs
Weekly rank: 97, 34th among eligible WRs
High projection: 18.9 fantasy points
Low projection: 7.9 fantasy points
Boom chance: 0.23
Bust chance: 0.1
Jordy Nelson, Oakland Raiders vs. Miami Dolphins:
Weekly rank: 131, 40th among eligible WRs
High projection: 18.3 fantasy points
Low projection: 5.2 fantasy points
Boom chance: 0.07
Bust chance: 0.24
For this section, the high projections and low projections were averaged for each of the 30 receivers. Goodwin (third) and Nelson (seventh) were the only receivers ranked outside of the top 20 to have their averaged projection finish in the top eight of the analysis.
Goodwin faces a Chiefs defense with an injured secondary that already has been involved in two high-scoring aerial games this season and is allowing an NFL high (by far) of 430 yards per game through the air thus far. Insights suggests that Goodwin not only has a great expected value for this week but also that he is a relatively safe bet with a boom potential more than twice as high as his bust potential.
Nelson faces a Dolphins defense that has allowed 270 yards per game through the air, 11th in the NFL. Nelson has had a slower-than-expected start to his season with the Raiders, but Insight suggests that his expected value is still high in Week 3. However, Insight also suggests that Nelson is more than three times more likely to bust than he is to boom, so proceed with caution.
Watson's insight: Goodwin is expected to have a good game, with a solid chance at a great game. Nelson is expected to outproduce his rank, but he is a risky play with a good chance to bust.
Most likely to 'boom'
Robert Woods, Los Angeles Rams vs. Los Angeles Chargers:
Weekly rank: 84, 28th among eligible WRs
High projection: 17.5 fantasy points
Low projection: 4.7 fantasy points
Boom chance: 0.28
Bust chance: 0.09
Kenny Golladay, Detroit Lions vs. New England Patriots:
Weekly rank: 108, 36th among eligible WRs
High projection: 17.5 fantasy points
Low projection: 4.7 fantasy points
Boom chance: 0.27
Bust chance: 0.11
For this section, the boom and bust likelihoods were averaged for each of the 30 receivers. Woods (first) and Nelson (sixth) were the two lowest-ranked receivers to have their averaged boom/bust projection finish in the top seven of the analysis.
Woods was estimated to have more than three times the likelihood to boom as to bust, and on Sunday, he faces a Chargers defense that has allowed five receiving touchdowns through two games, tied for fourth most in the NFL.
Golladay faces a Patriots defense that was just torched for 377 yards and four touchdowns through the air by Blake Bortles and the Jaguars. The Lions seem likely to be playing from behind against the Patriots, which would suggest that Golladay could get plenty of looks.
Watson's insight: Both Woods and Golladay have strong matchups against pass defenses that have been vulnerable this season, making them strong upside plays.