Week 3 usage report: snap-share risers and fallers through Week 2

Players compared

390

Risers (+10 pts)

72

Fallers (−10 pts)

53

RB committees

14 teams

How did we measure it?

Snap share is the percentage of a team's offensive plays a player was on the field for. For every QB, RB, WR and TE with a Week 1 snap, we compared his Week 2 share with his Week 1 share. A riser gained at least 10 points and played at least 40% in Week 2. A faller had been at 50% or more and lost at least 10. A player whose team played but who didn't gets 0%, marked DNP. Byes don't count.

Every number comes from the GridIron API: /players, /snaps and /stats, with injury statuses as of Saturday, September 26 at 11:00 am ET. One week of history is a thin baseline. We say so where it matters.

Who gained the most snap share in Week 2?

What it means for your lineup

• Aaron Jones Sr. is a start. With Jordan Mason on IR, 81% is a workhorse share, and 10.5 points on that much volume is closer to his floor than his ceiling.
• Rome Odunze is the buy. His points went 7.2, then 7.3, so his league-mates haven't noticed that his role nearly doubled. Role first, points later is the pattern snap share caught in Amon-Ra St. Brown's 2021 breakout. The catch: Caleb Williams is listed Out, so the quarterback play caps his Week 3.
• Jaylen Warren is Pittsburgh's back, since Rico Dowdle dropped to 26% and is now Out. Warren is Questionable himself, so this one rides on the 12:55 pm ET injury run.
• Germie Bernard and Konata Mumpfield are fill-ins, not breakouts. Both rose because the starter ahead of them sat (Pittman Jr. and Nacua). Stash Bernard in deep leagues; if Pittman plays Sunday, Bernard goes back to a rotational role.
• Seattle and Philadelphia are committees. Emanuel Wilson (41%) and Jadarian Price (34%) split Seattle's work. In Philadelphia, Will Shipley (51%) and Tank Bigsby (34%) soaked up Saquon's missing snaps. Neither backfield has a player you can trust in Week 3.

Who lost the most snap share?

What it means for your lineup

• Saquon Barkley is the one number here that doesn't add up. 16% of snaps and 3.0 points is what an in-game exit looks like, yet he's listed Healthy. Don't lock him in off the morning report: check his status after the Sunday injury run before his kickoff. If he's still Healthy, treat Week 2 as a one-game blip and start him.
• Rhamondre Stevenson is a sell. TreVeyon Henderson played 60% in his first week with snaps, and Stevenson dropped from 85% to 36%. That's a depth-chart change, not noise.
• DNP isn't a demotion. Pittman Jr., Nacua and Collins lost their snaps by not playing. Their roles are intact, so their Sunday status is the whole question.

Which Week 2 point spikes should you trust?

A big fantasy week backed by more snaps is a role change. A big week on fewer snaps is usually a touchdown or a long play, and those don't repeat on schedule. We split Week 2's biggest point risers by whether their snap share rose with them (by 5 points or more).

What it means: the Rams' spike is real. Davante Adams and Terrance Ferguson both played more with Nacua out, which also explains Matthew Stafford's 28.0. Tre Tucker and Jake Ferguson scored big on smaller roles, so don't pay waiver money for that. Ja'Marr Chase is Ja'Marr Chase; a 3.2-point Week 1 was the outlier, not Week 2.

How accurate were Week 2 projections?

We graded every player projected for at least 5 PPR points against what he actually scored. 57% of 250 landed within 5 points, and the median player scored 1.6 points under his projection: projections ran a little hot.

What it means: running back projections were the most trustworthy number in Week 2: 73% landed within 5 points. Quarterbacks were the least: only 31% within 5, with an average miss of 7.3. For a close QB decision, weight the matchup and the offense's snap trend over a one-point projection gap.

Which backfields are committees?

In Week 2, 14 teams had no running back above 60% of snaps (fullbacks excluded). Their backs are a riskier start than their names suggest, because the carries are split before the game begins.

What it means: Christian McCaffrey at 52% and Derrick Henry at 53% are still starts, because they score on the snaps they get. Down the list, "committee" is a real label. Bijan Robinson (51%) had Brian Robinson Jr. at 48%, and Houston's David Montgomery and Woody Marks split 52% and 48%.

How can I pull this data myself?

This report comes from about 20 API calls. /players returns each player's weekly snap share, so one paginated sweep gets the whole league:

import requests

API = "https://api.gridirondata.com/api/v1"
H = {"x-api-key": "YOUR_KEY"}

players, cursor = [], None
while True:
    params = {"limit": 100, **({"cursor": cursor} if cursor else {})}
    body = requests.get(f"{API}/players", headers=H, params=params).json()
    players += body["players"]
    cursor = body.get("next_cursor")
    if not cursor:
        break

def snap(p, week):
    row = p.get("seasons", {}).get("2026", {}).get("snap_counts", {}).get("weekly", {}).get(str(week))
    return row["offense_pct"] if row else None

movers = [(snap(p, 2) - snap(p, 1), p["player_id"]) for p in players
          if p["position"] in ("RB", "WR", "TE") and snap(p, 1) and snap(p, 2)]
for delta, pid in sorted(movers, reverse=True)[:10]:
    print(f"{delta:+5.0f}  {pid}")

That's about 18 requests, well inside the free plan's 100 a day.

Frequently asked questions

What is snap share?

The percentage of his team's offensive plays a player was on the field for. It moves before fantasy points do, because a coach decides the role before the box score shows it.

Why compare Week 2 with only one earlier week?

Because that's all the season has. One week is a small sample, so we flag a riser only when the jump is at least 10 points and the new share is at least 40%. The picture sharpens every week.

When will Week 3 snap data be in the API?

Sunday's stats and snaps land Monday at about 11:00 am ET, and Monday night's on Tuesday. The Week 4 edition of this report follows.

Are these projections?

No. Every number here is an actual: snap share, PPR points, or how far actual points landed from the weekly projection.