Welcome to Week 2 of the fantasy football season!
Last season, IBM introduced IBM Insights, a tool that uses its Watson technology to scan the web and parse it for fantasy football advice on a weekly basis.
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 but still quantitative supplement to the more in-depth, personalized analysis you normally see from our ESPN analysts.
The following article points out some of the recommendations Watson has for Week 2 of the NFL season, as of Friday afternoon.
Replacement running back decisions
One of the more difficult decisions for any fantasy manager is what to do in cases where a starting running back is injured. Should you play his real-life backup, or look elsewhere for your RB needs? IBM Insights presents several data points for estimating a player's production in a given week, which combine to create its predictions on each player's upside, downside, likelihood to boom or bust, and overall expected value range.
With this information, a fantasy manager can decide whether to lean toward maximizing upside, minimizing downside or perhaps to go in a completely different direction. Let's look at four situations in which the starting running back has either already been ruled out or is listed as questionable due to injury:
High projection: 22.5 fantasy points, low projection: 5.8 fantasy points
Boom chance: 0.08, bust chance: 0.16
Upside gain: 4.0, downside loss: 7.7
Le'Veon Bell has already been ruled out for Week 2 and has been removed from the Steelers' depth chart as he continues his holdout. There were questions about Conner going into Week 1, but he answered them emphatically with a monster game against the Cleveland Browns. These projections reflect a very high upside for Conner in Week 2, as only seven players have higher "high projection" values this week (Todd Gurley, Antonio Brown, Ezekiel Elliott, Alvin Kamara, Saquon Barkley, Michael Thomas, Aaron Rodgers).
On the other hand, Conner has a very low "low projection," along with a bust chance twice as great as his boom chance and a downside loss almost twice as large as his upside gain, according to Watson. Thus, IBM Insights would peg Conner as an extremely high-upside play Sunday against the Kansas City Chiefs, but one that also comes with significantly more risk than the other premium options.
Watson says: Play Conner in much the same way you would an active Bell, even if he's a bit more risky.
High projection: 15.8 fantasy points, low projection: 3.7 fantasy points
Boom chance: 0.07, bust chance: 0.07
Upside gain: 5.1, downside loss: 7.0
Devonta Freeman has been officially ruled out for Sunday's game against the Carolina Panthers, leaving Coleman as Atlanta's feature back. Because the algorithm includes data from all weeks, the "low projection" might be residual from when it was uncertain whether Coleman would start. However, it should be noted that Coleman is considered just as likely to boom as he is to bust, and that his upside gain and downside loss are of more similar caliber when compared to Conner.
Watson says: Play Coleman as a very low-risk RB1/RB2.
High projection: 15.8 fantasy points, low projection: 3.6 fantasy points
Boom chance: 0.26, bust chance: 0.08
Upside gain: 5.0, downside loss: 7.2
Rex Burkhead is questionable to play due to his participation in the league's concussion protocol, while Sony Michel missed the opener and remains questionable for Week 2 with a knee issue. Thus, White could remain in his usual role as New England's pass-catching running back, but it's also possible that he could end up as the primary feature back. That certainly accounts for Watson's projection that his chances to boom are more than three times his chances to bust. IBM Insights suggest that White should have baseline value either way, but if Burkhead and/or Michel sit, then White has a pretty good chance to improve upon that value.
Watson says: Play White as a RB2 with boom potential.
High projection: 11.6 fantasy points, low projection: 1.8 fantasy points
Boom chance: 0.08, bust chance: 0.25
Upside gain: 5.1, downside loss: 4.7
Leonard Fournette is listed as questionable with his hamstring injury, which means it is possible he'll play Sunday, though it is expected to be a game-time decision. This likely influenced Yeldon's very low "low projection" and high bust chances per Watson. Unlike Coleman or White, Yeldon would have had little to no involvement in the offense if his lead back were completely healthy. He's more pure backup than a change-of-pace or pass-catching back. Thus, for Week 2, IBM Insights projects that Yeldon has the upside of a modest fantasy starter (if Fournette sits), but also the lowest floor of any of the backs featured here.
Watson says: Only play Yeldon if Fournette is officially declared inactive.