• The Buffalo Bills might have a new backup running back: Ty Johnson outplayed Ray Davis during the first two weeks of the season, leading to Johnson playing more snaps.
• Ollie Gordon II takes over short-yardage situations: Gordon scored the first touchdown of the game in a goal-line situation.
• Unlock your edge with PFF+: Access Premium Stats, dominate fantasy with in-season tools and projections and make smarter bets with the new PFF Player Prop Tool.

Estimated reading time: 5 minutes

PFF’s fantasy football recap focuses on player usage and stats, breaking down all the vital information you need to achieve fantasy success in 2025.

Ty Johnson potentially surpasses Ray Davis: Johnson played more early-down snaps than Davis.

Last season, James Cook was the Bills’ early-down back, while Davis was his backup, and Ty Johnson was the passing-down back. Cook played 62% of Buffalo’s offensive snaps on early downs, followed by Davis’ 26% and Johnson’s 14%. Over the first two weeks of the season, Davis and Johnson were much closer together on early downs, 20% by Davis and 18% by Johnson.

This week, Davis was the Bills’ first backup on the field, but the next several snaps by a backup were from Johnson, who ended the game with more early down snaps while still maintaining his third-down role. Davis has averaged 2.4 yards per carry in the first two weeks of the season, compared to Johnson’s 7.5, which could have contributed to the change.

It doesn’t matter how the backup snaps are distributed as long as Cook is healthy, but if Cook suffers an injury, there is no guarantee Davis would be the clear early-down back like he was in Week 6 last season when Cook was inactive. It makes sense to wait another week to see if Johnson remains ahead of Davis before dropping him. If Johnson remains ahead of Davis, then it would probably make sense to look for a different handcuff on your roster.

Dalton Kincaid shines despite an unideal role: Kincaid led the Bills’ receivers with five receptions for 66 yards and a touchdown.

Kincaid has been part of a three-man rotation to start the season. He played less than 50% of Buffalo’s offensive snaps over the first two weeks. He played 42.7% of the snaps on early downs out of 11 personnel, while he played 40.6% of the third downs out of 11 personnel, 54.5% of the two-tight-end sets, and all of the three-tight-end sets. Part of the problem has been Dawson Knox, but the other part is Jackson Hawes, who was our highest-graded tight end through two weeks, albeit on a smaller sample size than most tight ends at 51 snaps. This left Kincaid running a route on 56% of the Bills’ pass plays, which is far from ideal for a fantasy starting tight end. That is a lower rate than last season, despite Kincaid dealing with injuries all season.

The tight end rotation cost him in this game, as there was a two-tight-end set where he wasn’t on the field, and Hawes ended up scoring the touchdown.

Thursday night, his rates in each situation were very similar to the first two weeks. The difference this week is that his target rate per team target was much higher this week than the first two weeks combined. This came at all of the wide receivers’ expense. While there should be other big games by Kincaid, it’s much more likely that his target rate regresses from this game than his playing time increases notably. This makes Kincaid a sell-high candidate and someone who will be frustrating in fantasy lineups going forward.

Ollie Gordon II takes short-yardage snaps: The sixth-round rookie running back had an increased role this week, primarily in short-yardage situations.

Gordon has been De’Von Achane‘s primary backup, while Alexander Mattison is on injured reserve, and Jaylen Wright missed the first two games due to a knee injury. Wright was active this week but didn’t play any snaps on offense.

In the first two weeks, Achane played at least 78% of Miami’s offensive snaps in every situation. This included five of six snaps in short-yardage situations and the Dolphins’ only goal-line snap. This week, Gordon played in both short-yardage and goal-line situations. This included a goal-line snap on the first drive, leading to the Dolphins’ first touchdown.

There is a chance the Dolphins simply wanted to limit Achane’s playing time after playing a career-high 93% of snaps last week, 10% more than any other week in his career. This was also a short week, and it wouldn’t be the first time a team adjusted a player’s playing time for a Thursday night game. However, it would also make sense to give Gordon the short-yardage and goal-line snaps, given his size advantage over the starter. Achane would still be a must-start player, but it will be much harder for him to finish the season as a top-five running back if he’s not scoring the goal-line touchdowns.

Miscellaneous notes

• Dolphins wide receiver Jaylen Waddle was questionable heading into this game due to a shoulder injury, but that had no impact on his playing time.
• Tight end Darren Waller has missed his third consecutive game due to a hip injury. He also didn’t play in the preseason.
• The Dolphins continued the same tight end rotation as the first two weeks. Julian Hill played more on early downs and Tanner Conner played more often on passing downs. This typically would mean more playing time for Hill, but the Dolphins have typically played from behind this season, leading to relatively even playing time for both players.
• Miami elevated kicker Riley Patterson off the practice squad for the third straight game with Jason Sanders on injured reserve.
• The Dolphins made Tahj Washington a healthy inactive, as they’ve done throughout the season, giving them five active wide receivers.
• Similarly, the Bills made wide receiver Curtis Samuel a healthy inactive, as they’ve done all season.
• Joshua Palmer played notably less in snaps this week. His playing time in 11 personnel was similar to past weeks. He hasn’t played as much as other Bills receivers in one- or two-receiver sets this season, but he barely played in those personnel groupings this week.

Table Notes

• Snaps include plays called back due to penalties, including offensive holding or defensive pass interference. The other three stats have these plays removed.
• Targets may differ from official NFL sources. The most likely discrepancy would be from a clear thrown-away pass, where the NFL may give the target to the nearest receiver, while this data will not.
• Carries are only on designed plays. Quarterback scrambles won’t count for the total number of carries in the game.