Allen
RB · NYJ vs CHI
Proj pts
16.6pts
market
Williams
RB · LA vs PHI
Proj pts
14.4pts
market
Braelon Allen or Kyren Williams: who should I start in Week 4?
Start Braelon Allen. Bench Kyren Williams. Braelon Allen leads at 16.6. Kyren Williams is the first one out, 2.2 short.
Crowd: started Allen 49% · Williams 97%
Highlights
Pros
- Reliable99% consistency
Cons
- Committee back3% of touches
- Negative game scriptTrailing
Pros
- Goal-line back1.2 GL carries/g
- Positive game scriptFavored
- High floor12.7 floor
Cons
- Below your start line2.2 pts short
Some recent form, opportunity and matchup figures still lean on the 2025 season (not enough games played this season yet).
Week 4 · Updated Tue 6:00 AM ET
Why Braelon Allen
Start Braelon Allen — projects 16.6, an even matchup vs CHI, with a rising role.
Braelon Allen projects 16.6 points on a 4.3–18.2 range. The matchup grades as neutral against CHI. Usage: carries / game at 4.5. Net: Braelon Allen is start this week.
Working for Allen: Reliable (99% consistency).
Working against Allen: Committee back (3% of touches), Negative game script (Trailing).
Where Allen leads the head-to-head: Projection (16.6 vs 14.4); Consistency (99% vs 62%); Bust rate (0% vs 6%).
Why Kyren Williams
Sit Kyren Williams — projects 14.4, an even matchup vs PHI, with a rising role.
Kyren Williams projects 14.4 points on a 12.7–20.4 range. The matchup grades as neutral against PHI. Usage: carries / game at 15.2. Net: Kyren Williams is a bench this week.
Working for Williams: Goal-line back (1.2 GL carries/g), Positive game script (Favored), High floor (12.7 floor).
Working against Williams: Below your start line (2.2 pts short).
Where Williams leads the head-to-head: Floor (12.7 vs 4.3); Ceiling (20.4 vs 18.2); Touch share (35% vs 3%); Opportunity score (13.9 vs 4.0); Matchup grade (recency weighted) (B- vs C); Implied team total (23.3 vs 20.0); Last 3 avg (17.7 vs 5.7); Last 5 avg (16.0 vs 4.9); vs season (+1.0 vs +0.7); Boom rate (6% vs 0%); Carries / game (15.2 vs 4.5); Touches / game (17.4 vs 5.0); Target share (8% vs 3%); Opportunity score (exp pts) (13.9 vs 4.0); Yards per carry (4.8 vs 4.2); Snap share (68% vs 23%); Goal line carries (20 vs 1); Red zone carries (61 vs 7); Grade vs RB (recency weighted) (B- vs C); Implied total (23.3 vs 20.0).
Recent form
Braelon Allen's last 5 games: PPR points 6.7, 0.6, 4, 8.5, 4.7 (oldest to newest).
Kyren Williams's last 5 games: PPR points 16, 11.1, 15.5, 15.7, 21.8 (oldest to newest).
Williams's worst of these 5 (11.1) beats Allen's best (8.5) — the run argues the other way.
Allen has already beaten Williams's 14.4 projection in 0 of his 5; Williams has beaten Allen's 16.6 in 1 of his 5.
At least 1 of these 5 were played before the 2026 season — the card does not date them.
Matchup
Braelon Allen (NYJ) faces CHI, kickoff Sun 1:00 PM ET. The matchup is rated neutral. Betting line: implied team total 20.0, spread +3.5.
Kyren Williams (LA) faces PHI, kickoff Sun 1:00 PM ET. The matchup is rated neutral. Betting line: implied team total 23.3, spread -3.0.
What flips it
Allen's edge is 2.2 points (16.6 to 14.4). Williams's ceiling (20.4) clears Allen's projection, so a big game from Williams flips it. Allen's floor (4.3) sits under Williams's projection, so a quiet game from Allen flips it too. Confidence in the call: high.
How to read these numbers
- Confidence in the call
- How clear-cut this start/sit call is — how decisively the player beats the alternative he competes with (for a starter, the best player on your bench; for a bench player, the last one in your lineup), on both projected points and the overlap of their likely ranges. A clear play reads high; a genuine toss-up reads lower. It rates the decision, not the player’s own volatility, and is not a measured accuracy rate.
- Floor
- The 20th percentile of the same simulated distribution.
- Ceiling
- The 90th percentile of the same simulated distribution.
- Opportunity score
- Expected points from projected volume alone, before any efficiency of this player’s own: each projected touch and target valued at the positional average for the scoring format.
- Matchup grade
- The opponent’s points allowed to this position per game, weighted toward recent games, then graded on a curve across the 32 defenses.
- Implied team total
- The points the betting market expects this team to score, derived from the game total and the spread. Market-derived, not modelled by Nuvela.
- Trend
- The direction of his scoring across the trailing window, classified as heating, steady or cooling.
- Touch share trend
- The same classification, measured on his share of team touches rather than on his points.
- Consistency
- The share of his games that landed within a set band of his own per-game average.
- Boom rate
- The share of his games that finished inside the top 12 scorers at his position for that week.
- Bust rate
- The share of his games that came in under half of what he was projected for that week.
- Vacated touches
- Carries and targets belonging to team-mates who are ruled out, on IR, or listed as non-participants for this week, taken at their recent per-game rate.
- Defense funnel
- Whether the opponent’s rush defense and pass defense are far enough apart to push offenses one way. A gap inside the band reads Neutral.
- Game script projection
- The expected state of the game for his team — favoured, level or trailing — taken from the closing spread.
More start/sit comparisons from this week's PPR projections. Open a pair for the call and the reasoning.