For decades, NFL performance was judged by traditional box-score stats: passing yards, touchdowns, sacks, and interceptions. Today, those surface-level numbers are only the beginning. Behind every snap lies a massive data operation powered by advanced analytics, artificial intelligence, and cloud computing—most notably through Amazon Web Services and NFL Next Gen Stats.
Modern football analytics blends probability theory, machine learning, and real-time tracking to answer deeper questions: How hard was that throw? Was the quarterback actually under pressure? Did the receiver create yards—or were they schemed open?
Core NFL Performance Metrics Explained
Let’s break down the core metrics and the technology that makes them possible.
DVOA (Defense-adjusted Value Over Average)
DVOA measures how efficient a team or player is compared to league average, adjusted for opponent quality and game situation.
- A 10-yard gain on 3rd-and-9 is worth far more than the same gain on 3rd-and-20.
- Unlike raw yardage, DVOA evaluates context and impact.
Why it matters: It reveals which teams consistently win high-leverage moments—not just who racks up stats.
As of the end of the 2025 NFL regular season and heading into the Super Bowl (January 2026), the Seattle Seahawks hold the #1 spot in overall DVOA (45.1%), driven by the top-ranked defense. They are also rated as the top defense in DVOA history, ranking 6th all-time when including playoffs.
- Total DVOA: Ranked #1 in the NFL.
- Defense DVOA: Ranked #1 in the NFL.
- Offense DVOA: Ranked 11th (notably dropping from 1st earlier in the season).
- Key Stat: The 2025 defense is ranked as the 12th best in DVOA history.
EPA (Expected Points Added)
EPA assigns a point value to every play based on how it changes a team’s expected score.
- A 5-yard gain on 3rd-and-4 produces positive EPA
- A 5-yard gain on 3rd-and-10 produces negative EPA
Why it matters: EPA directly connects play-by-play performance to scoreboard impact, making it one of the most trusted efficiency metrics.
The Seattle Seahawks’ defense ranked No. 2 in the NFL in EPA per play during the 2025 regular season, according to Next Gen Stats and PFF. They are noted for having the league’s top-ranked rushing defense by EPA.
Key 2025 season EPA statistics for the Seahawks:
Overall Team: Ranked 15th in total EPA.
Defense: Ranked 2nd in EPA per play for the season, with some metrics placing them as the best overall unit by the end of the year.
Rush Defense: Ranked 1st in the NFL.
Post-Bye Week: Since their Week 8 bye, the defense ranked 1st in EPA per play, while the offense ranked 8th.
CPOE (Completion Percentage Over Expected)
CPOE compares a quarterback’s actual completion rate to an expected completion rate calculated using machine learning.
The model considers:
- Defender proximity
- Throw depth
- Sideline distance
- Pressure and pocket movement
Why it matters: It separates elite accuracy from easy completions.
In the 2025 NFL season, Seattle Seahawks quarterback Sam Darnold ranked 4th in Completion Percentage Over Expected (CPOE) with a 5.2% rate, serving as a key factor in the team securing a top playoff seed. Throughout the season, Darnold was consistently among the league’s top quarterbacks in accuracy, at one point leading the NFL in CPOE.
- Elite Performance: Throughout 2025, Darnold was recognized for high-level efficiency, including placing 4th in CPOE.
- Season Trend: While starting strong and ranking 1st in CPOE earlier in the season, he ranked 5th in CPOE among 36 qualifying quarterbacks from Week 7 onward.
- Context: Darnold’s performance was bolstered by a top-ranked scoring defense and an improved running game, helping Seattle secure the NFC’s top seed.
aDOT (Average Depth of Target)
aDOT measures how far downfield a quarterback throws on average.
- Low aDOT quarterbacks may rely on quick-game concepts
- High aDOT quarterbacks take more vertical risks
Why it matters: It contextualizes efficiency metrics—short passes inflate completion rate, but not difficulty.
Based on 2025 season data, the Seattle Seahawks, with Sam Darnold at quarterback, featured an offense that frequently utilized deep passes to wide receiver Jaxon Smith-Njigba, with Darnold ranking 4th in EPA/play and 4th in CPOE (Completion Percentage Over Expected). While specific league-wide rankings for Average Depth of Target (aDOT) are not explicitly ranked in the provided search results, the team’s top-tier passing efficiency indicates a high-functioning deep passing attack, supporting a top-10, and possibly top-5, ranking in passing success.
- Sam Darnold’s 2025 Performance: Finished with 25 touchdowns, 14 interceptions, 4,048 yards, and a 99.1 passer rating,,, SI.com.
- Key Metrics: Ranked 4th in Expected Points Added (EPA) and 4th in CPOE (5.2%), , SI.com.
- Offensive Context: The offense supported a 14-3 record, securing the #1 seed in the NFC, , Pro-Football-Reference.com.
Darnold’s high EPA/play and CPOE, combined with the team’s overall offensive success (3rd in points scored), suggest a top-tier aDOT, likely placing in the top 5-10 range, though a precise, exact 1-32 ranking for aDOT is not provided in the text.
The Engine Behind the Numbers: AWS & NFL Next Gen Stats
Since their partnership began around 2020, the NFL and Seattle-based AWS have transformed raw tracking data into real-time intelligence.
RFID Tracking at Scale
Every player wears RFID chips in their shoulder pads, and the football itself contains a sensor. These devices track:
- Speed
- Acceleration
- Direction
- Position (10+ times per second)
Across a season, this produces over 300 million data points, with each game generating roughly 3.5 terabytes of data.
Real-Time Analytics During Games
AWS processes player, ball, and referee movement instantly—allowing broadcasters to display metrics during live plays, not after the fact.
This include
- Live speed readouts
- Route visualizations
- Defensive pressure indicators
Key Next Gen Stats Metrics Powered by AI
Expected Yards After Catch (xYAC)
xYAC estimates how many yards a receiver should gain after the catch, based on:
- Defender angles
- Separation
- Speed and field position
Use case: Differentiates receiver skill from scheme design.
Pressure Probability
A machine learning model that quantifies how likely a pass rusher is to disrupt a play—even if no sack occurs.
It analyzes:
- Time to pressure
- Pocket collapse rate
- Pass rusher leverage
Why it matters: It finally gives edge rushers credit beyond sacks.
Rush Yards Over Expected (RYOE)
RYOE measures how many rushing yards a ball carrier gains beyond what the blocking and defensive alignment would predict.
Why it matters: It isolates individual rushing talent from offensive line performance.
Completion Probability & Expected Return Yards
Used primarily for passing and special teams:
- Completion Probability estimates how likely a pass was to be completed
- Expected Return Yards evaluates kickoff and punt returns based on coverage positioning
Digital Athlete: Football Meets Simulation
One of the most ambitious initiatives is the Digital Athlete, a joint NFL–AWS project.
Using real player data, AWS builds virtual replicas of athletes to:
- Simulate collisions
- Study injury risk
- Test rule changes before implementation
This technology is now central to player safety research and equipment design.
Super Bowl LX Matchup: the Seattle Seahawks vs New England Patriots
While not all of the data from the 2025 NFL Season, including the 2025 Playoffs, where the Seattle Seahawks and quarterback Sam Darnold seemed to shift their game into a higher gear, there are some baseline metrics that can be factored in.
Seattle Seahawks — 2025 Snapshot
Data fingerprint
- ↑ Explosive-play rate
- ↑ QB scramble EPA
- ↑ Defensive speed, ↓ defensive discipline
- Volatile efficiency, high ceiling
Seattle’s roster construction trends toward:
- Higher aDOT
- Positive Rush Yards Over Expected
- Above-average pressure without blitzing
- Slightly negative late-down EPA variance (boom/bust)
New England Patriots — 2025 Snapshot
Data fingerprint
- ↑ Offensive efficiency, ↓ explosiveness
- ↑ CPOE and Completion Probability
- ↑ Situational EPA (3rd down, red zone)
- Lower volatility, higher floor
New England’s post-Brady rebuild has leaned into:
Strong defensive EPA allowed inside the 20
Quick-game passing
Scheme-generated xYAC
Pressure Probability through disguise
Projected Advanced Metrics Comparison (2025-Based Model)
| Metric | Seahawks (Proj.) | Patriots (Proj.) | Edge |
|---|---|---|---|
| Offensive EPA/play | +0.11 | +0.15 | Patriots |
| Defensive EPA/play allowed | -0.07 | -0.04 | Seahawks |
| Net EPA | +0.04 | +0.11 | Patriots |
| CPOE (QB) | +1.5% | +4–5% | Patriots |
| aDOT | Higher (aggressive) | Lower (efficient) | Seahawks |
| RYOE (team) | Positive | Neutral | Seahawks |
| Pressure Probability (def.) | High without blitz | High via scheme | Even |
| Red Zone EPA | Average | Top-5 | Patriots |
| Late-Game EPA (4Q) | Volatile | Elite | Patriots |
Next Gen Stats–Style Matchup Breakdown
When Seattle Has the Ball
Advantages
- Positive RYOE → real run-game stress
- Explosive pass upside (aDOT + scramble EPA)
- Receivers outperform xYAC when plays break down
Risks
- Lower Completion Probability vs disguised coverage
- Negative EPA when forced into long-yardage, timing throws
Projection:
Seattle wins chunks, loses consistency.
When New England Has the Ball
Advantages
- High Completion Probability
- Strong CPOE on intermediate routes
- Elite red-zone play sequencing
- Low sack rate despite pressure
Risks
- Limited explosive-play ceiling
- If forced into shootout mode, EPA flattens
Projection:
New England wins downs, not highlights.
Digital Athlete & Injury Risk Angle
Using AWS-style Digital Athlete modeling assumptions:
- Seattle’s speed-heavy defensive profile → higher soft-tissue fatigue risk
- New England’s rotational depth → lower in-game injury variance
This subtly favors New England late in games.
Analytics-Based Game Script Projection
Most likely flow
- Seattle jumps ahead via explosive plays
- Patriots stabilize with high-EPA drives
- Game tightens in the 4th quarter
- Outcome hinges on one high-leverage red-zone or 3rd-down sequence
Why the Seahawks Rate ~5 Points Better on the Spread (Analytics View)
Explosiveness Drives Point Spreads
Point spreads are highly sensitive to explosive-play potential, and this is Seattle’s biggest edge.
Analytics drivers:
- Higher aDOT
- More explosive EPA swings (20+ yard plays)
- Positive Rush Yards Over Expected
- QB scramble EPA adds hidden variance
Why it matters:
Explosive teams cover spreads more often because they create nonlinear scoring events (quick TDs, short fields).
Seattle’s offense produces wider score distributions → higher projected margins → spread inflation.
Defensive Ceiling > Defensive Consistency
Seattle’s defense:
- Generates pressure without blitzing
- Maintains tighter coverage shells
- Forces negative EPA in bunches
New England’s defense:
- Excellent at situational downs
- Less dominant snap-to-snap
- Relies more on disguise than raw disruption
Betting models reward defensive ceilings, not just averages.
A defense that can erase drives entirely is worth more against the spread than one that merely bends.
Variance Is Your Friend in Spread Betting
Seattle is a high-variance team:
- Big wins
- Occasional ugly losses
New England is low-variance:
- Close games
- Narrow margins
Spread models prefer:
- Teams that can win by 10–14
- Even if they sometimes lose outright
New England wins close.
Seattle wins big when they win.
That alone can be worth 3–4 points in spread math.
Put It All Together
Why Seattle –5 Makes Sense
| Factor | Edge |
|---|---|
| Explosive-play EPA | Seahawks |
| Defensive disruption | Seahawks |
| RYOE | Seahawks |
| Variance (spread value) | Seahawks |
| Late-game efficiency | Patriots |
| Close-game win rate | Patriots |
The Bottom Line (This Is the Answer)
Seattle is favored by ~5 because they are more likely to win by multiple scores—
even if New England is more likely to win a close game.
That’s not a contradiction.
That’s modern betting analytics. [24×7]
