Sep 27, 2026; Detroit, Michigan, USA; New York Jets quarterback Geno Smith (7) throws the ball against the Detroit Lions during the game at Ford Field. Mandatory Credit: David Reginek-Imagn Images
The Jets are 1-3 after four games, with 76 points scored and 84 allowed. They opened with a win in Tennessee, lost in overtime to Green Bay, then fell at Detroit and Chicago. The self-scout shows an offense with productive passing answers and a defense with contrasting game-to-game results. Sustaining those answers is the central problem.
This four-game look at New York pairs with our Browns scouting report, which examines Cleveland’s 3-1 start and the narrow margins behind it. The strongest offensive contrast: +0.140 EPA per drop back versus -0.170 per run-classified play. Twelve sacks allowed and a 0-for-7 third-down game in Chicago help explain why positive passing efficiency has not produced consistent scoring.
| Game | Result | NYJ third downs | Sacks allowed | Sacks made | ||
|---|---|---|---|---|---|---|
| W1 at Tennessee | W 23-10 | 18.2% (2/11) | 0 | 3 | ||
| W2 vs. Green Bay | L 17-20 (OT) | 42.1% (8/19) | 4 | 3 | ||
| W3 at Detroit | L 24-31 | 50.0% (6/12) | 5 | 2 | ||
| W4 at Chicago | L 12-23 | 0.0% (0/7) | 3 | 1 | ||
Geno Smith’s passing sample includes 85 completions on 117 credited attempts and no interceptions. The ordinary-play CPOE sample is +7.9 percentage points, indicating more completions than the model expected on those 110 attempts. That is a useful efficiency signal within this window, alongside the sacks and uneven third-down outcomes.
The defense held Green Bay to one third-down conversion in nine tries, but Chicago converted seven of fourteen and generated eleven explosive gains under the thresholds used here. Breece Hall and Mason Taylor were unavailable for the Chicago game, and the safety rotation also changed. Those personnel differences belong in the interpretation of Week 4.
On first down, Jets drop backs gained 485 net yards on 54 plays, 8.98 per play, compared with 171 yards on 44 run-classified plays, or 3.89 each. The passing average was higher in all four games. Explosive completions carry substantial weight in that result, so it is a prompt to examine first-down choices alongside score, field position and the opponent.
| Film alignment | All plays | Runs | Yards / run | ||
|---|---|---|---|---|---|
| Under center | 82 | 54 | 4.07 yd | ||
| Shotgun | 142 | 34 | 2.35 yd | ||
The complete ordinary-offense formation chart places 142 plays in shotgun and 82 under center. Under-center runs gained 4.07 yards each, compared with 2.35 from shotgun. This is an alignment-and-outcome association; runner, down, front and game state vary. Productive Tennessee hand offs from both alignments show that shotgun runs can work.
The Jets converted 16 of 49 third downs (32.7%). Their six conversions on 26 third downs with seven or more yards to gain represent 23.1%. Green Bay accounted for fourteen of those long-yardage opportunities and also supplied four conversions. Chicago’s 0-for-7 result warrants a focused review, while the fourteen combined conversions against Green Bay and Detroit show that answers were available.
Sacks became more costly after the opener: none on 26 Tennessee drop backs, four on 48 against Green Bay, five on 44 at Detroit, and three on 18 in Chicago. Three Detroit sacks occurred on five red-zone drop backs. The film identifies the negative plays; common protection, route and quarterback causes remain unresolved.
Garrett Wilson led with 31 named targets, followed by Kenyon Sadiq with 18, Adonai Mitchell with 15, Williams with 12 and Hall with 10. There are 110 credited throws with a named target; seven of 117 attempts lack one. Selected film examples show boundary throws, a running-back swing, tight-end vertical gains and a short goal-line completion.
Hall was inactive at Chicago, where Braelon Allen played 30 of 32 offensive snaps and handled all fourteen run-classified plays. Taylor was inactive in Detroit and Chicago; Sadiq’s snap share increased. Week 4 therefore represents a different available-player mix from the opening games.
Geno’s 116 ordinary throws averaged 2.68 seconds from snap to release, with a 2.51-second median. 57 throws (49.1%) came out in 2.5 seconds or less; 34 (29.3%) took more than three seconds. Those bands describe the distribution and do not assign a decision grade.
| Sample | Throws | Mean TTT | Median TTT | Release mean (n) | ||
|---|---|---|---|---|---|---|
| Week 1 | 24 | 2.64 s | 2.49 s | 0.332 s (23) | ||
| Week 2 | 40 | 2.61 s | 2.39 s | 0.297 s (35) | ||
| Week 3 | 37 | 2.77 s | 2.64 s | 0.296 s (12) | ||
| Week 4 | 15 | 2.67 s | 2.75 s | 0.334 s (15) | ||
| Four-game pool | 116 | 2.68 s | 2.51 s | 0.313 s (85) | ||
All 117 credited attempts have TTT review. Including the separately retained spike lowers the overall mean to 2.66 seconds; the table excludes it so the ordinary throwing sample stays clear. Longer holds require the individual play context before assigning responsibility to Geno, protection or receiver availability.
Windup-to-release duration averaged 0.313 seconds on 85 measurable throws. The other 32 credited attempts have explicit unmeasurable outcomes, including the spike. Detroit has only twelve measurable releases out of 37 attempts, so its release mean represents a particularly limited observed sample. This experimental measure describes the throwing motion, while TTT includes the time before that motion begins.
| Metric | Jets offense | Opponents vs. NYJ | ||
|---|---|---|---|---|
| EPA / play | +0.018 (n=224) | 0.000 (n=247) | ||
| EPA / drop back | +0.140 (n=136) | +0.076 (n=143) | ||
| EPA / run | -0.170 (n=88) | -0.105 (n=104) | ||
| Positive-EPA success | 43.8% (n=224) | 49.4% (n=247) | ||
| CPOE | +7.9 pp (n=110) | +0.7 pp (n=119) | ||
| Average target depth | 7.21 yd (n=116) | 6.39 yd (n=127) | ||
EPA estimates the change in an offense’s expected scoring value. The Jets’ drop backs supplied positive value, while the run-classified sample reduced it; the overall offense was slightly positive. Opposing EPA was essentially even across the four games. Lower EPA in the defensive column favors the Jets. These raw values have no league ranking or opponent adjustment.
Success means positive EPA here. CPOE is actual completion rate above the model’s expected rate, in percentage points. Average target depth is the air-yard distance of an attempt. Sample counts show where missing model values narrow the denominator.
Opponents converted 17 of 45 third downs (37.8%), with sharply different game results: Tennessee 4/12, Green Bay 1/9, Detroit 5/10 and Chicago 7/14. The full window includes nine sacks on 143 drop backs and three takeaways. Chicago provided two takeaways despite its large offensive workload.
| Metric | Observed result | Sample | ||
|---|---|---|---|---|
| Observed pressure, classified reps | 12.3% (17/138) | 5 of 143 unclear | ||
| Sacks / drop backs | 6.3% (9/143) | 9 official sacks | ||
| Mean time to first pressure | 2.04 s | 12 of 17 pressure plays timed | ||
| Mean time to sack | 4.10 s | 9 of 9 sacks timed | ||
| Median time to sack | 3.39 s | Range: 2.32-7.15 s | ||
Pressure requires a visible effect on the quarterback, such as forced evasive movement or an altered throw. Seventeen of 138 classified drop backs show that effect, with five additional cases unclear. The 2.04-second pressure average comes from twelve timed pressure plays; five classified pressure onsets were unmeasurable. All nine sacks have bounded endpoint estimates, averaging 4.10 seconds.
The pressure and sack means describe different conditional samples. They cannot establish coverage quality: protection losses, missed receivers, quarterback movement and finishing all matter. The retained review leaves coverage contribution unclear on the nine sacks.
Chicago ran fifty times for 231 yards, with 33 runs tagged through left or right guard. That concentration differs from the first three opponents, whose official charts contained twelve guard runs combined. Dense film examples show Monangai and Johnson reaching the second level, alongside later stops at or behind the line. The evidence does not establish one repeated gap error.
Explosive concessions increased from two at Tennessee and three against Green Bay to five at Detroit and eleven at Chicago. The late increase covered both run and pass plays. The defensive presnap chart usually shows two separated deeper defenders: 230 of 232 classified open-field appearances, with fourteen goal-line looks and one unresolved case treated separately. That spacing alone does not identify the post-snap coverage call.
The Jets made seven of eight field goals and five of six extra points. One Green Bay punt return traveled 63 yards and accounts for more than half of the 114 signed punt-return yards allowed. Three opponent kickoff returns reached at least 37 recorded yards among fourteen returns. Those outliers make placement, coverage and field position useful review topics alongside the ordinary execution.
| Metric | Jets offense | Opponents vs. NYJ | ||
|---|---|---|---|---|
| Drop back share | 60.7% (136/224) | 57.9% (143/247) | ||
| Third-down conversion | 32.7% (16/49) | 37.8% (17/45) | ||
| Third and 7+ conversion | 23.1% (6/26) | 7.1% (1/14) | ||
| Third and 1-3 conversion | 50.0% (5/10) | 52.6% (10/19) | ||
| Sacks / drop backs | 8.8% (12/136) | 6.3% (9/143) | ||
| Explosive runs: 10+ yards | 8.0% (7/88) | 11.5% (12/104) | ||
| Explosive drop backs: 20+ yards | 8.8% (12/136) | 6.3% (9/143) | ||
First, connect the efficient passing examples to the plays that put the offense behind the sticks. The first-down yardage edge, the under-center rushing split and the twelve sacks give concrete situations to examine. Third-down and red-zone losses deserve individual explanations before changing a protection or route plan.
Second, treat Chicago’s run concentration as a specific matchup test. Compare productive interior tracks with the later stops, and account for the changed safety participation. The positive Green Bay third-down sample provides a useful contrast; neither game alone defines the defense.
Third, retain the available-player context and special-teams outliers in the weekly plan. Hall’s absence, Taylor’s absence, Allen’s condensed workload and the changed secondary rotation make a four-game average less useful unless the next lineup is comparable.
The window is Tennessee, Green Bay, Detroit and Chicago in 2026 Weeks 1-4. This edition uses the retained Jets self-scout, NFLverse play-by-play and official game book references. EPA, success and CPOE are nflfastR-derived measures. The copy describes this four-game sample without a league ranking or opponent adjustment.
Efficiency covers 224 Jets offensive plays and 247 opposing offensive plays. No-play penalties, kneels, spikes, conversions and special teams are excluded. Drop backs include sacks and non throwing scrambles. Run-classified plays include the retained aborted-snap outcome. Explosives mean gains of 10+ yards on runs and 20+ on drop backs. Red-zone play counts are distinct from drive scoring rates.
All 117 credited Geno throws have TTT review: fifty measured and 67 estimated, with source uncertainty retained. The ordinary-throw timing table excludes one spike and pools 116 throws. Eighty-five release duration are measurable; 32 are unmeasurable, including 25 Detroit cases. Release averages exclude those exceptions. Both All-22 angles count as one physical event.
All 628 retained physical events have paired sampled-overview review. Offensive formation is classified on all 224 ordinary plays; exact personnel is identified on 36. Defensive shell is classified on 246 of 247 ordinary plays. Broad run tracks, target paths, motion, actual rush counts and named coverage calls remain partial; selected examples do not establish their full-window frequencies.
Pressure review covers all 143 opposing drop backs: seventeen pressure-positive, 121 with no observed pressure and five unclear. TTP has twelve bounded estimates and five unmeasurable pressure onsets. TS has nine bounded estimates across all nine official sacks. The 12.3% observed-pressure rate uses 138 classified cases, and the uncertain cases remain visible. Pressure timing does not establish coverage contribution.