Jaylen Warren or J.K. Dobbins: Who Should You Start?

The call

• Role is the problem: a 37% snap share means he's not the every-down back, and the 11.4 projection lands him at RB29 — right on the 12-team flex line.
• New England's defense grades 10 of 32 vs the position (32 = softest), so this is a tougher-than-average box to run into, though it's a 1-game sample with 0.00 confidence attached.
• Game script cuts both ways: a 19.5 implied total and 4.5-point spread limit the ceiling, but trailing football feeds his 16% target share, which is where his PPR floor lives.
• Tipping point: if he's the pass-game back on third downs and the Steelers fall behind, he's a usable flex; if this stays a low-scoring grind with a split backfield, bench him for a back with a locked-in workload.

The call

• The Week 1 usage is the tell: eight carries, team-high 36 yards, and no targets. A back with 0% target share in PPR needs volume or a score, and 43% snaps says he isn't getting the volume.
• The matchup is a negative, not a boost — Jacksonville grades 9 of 32 in the DvP composite (1 = stiffest), so the soft-spot argument isn't there even on a one-game sample.
• Game script is the one thing working for him: a 23.0 implied total as a 2.5-point favorite keeps the early-down rushing script alive, which is exactly where he gets his work.
• Tipping point: if Jacksonville plays it close and this becomes a grind, RB37 at 11.6 projected is a usable flex. If the model's expectation is a game that gets away like Kansas City did, he's off the field late — bench him for any back who catches passes.