NFL Week 1: Tyler Shough, Jahmyr Gibbs and the Limits of a Single Game

In Detroit, Tyler Shough walked off the field with 410 passing yards and 3 touchdowns. The NFL's individual stat sheet placed his name at the top of the league in passing yards. The scoreboard said the opposite: the New Orleans Saints lost to the Detroit Lions. Elsewhere, Jahmyr Gibbs of the Lions closed out his afternoon with 156 rushing yards and 2 touchdowns in a win. In Tennessee, Anthony Hill Jr. recorded 16 tackles, the highest total in the league after Week 1.

Three players. Four numbers. One game each. That is everything the Week 1 individual leaderboard left behind for the rest of the season.

I have read leaderboards of this kind for years, across many leagues, and what makes them interesting is not what they display but what they keep silent. A weekly leaderboard is an editorial document before it is an analytical one. It selects three categories, selects three names, and by selecting, it shapes how millions of readers understand a season that has only just begun.

What is worth noting is that Week 1 had only just ended.

Context: one eighteenth

The NFL regular season runs 18 games per team. Week 1 covers 5.6 percent of the road. Set beside a European club season of 38 rounds, the NFL's opening week is roughly the first two matches, a stage at which anyone who has worked with sports data knows there is nothing yet worth concluding.

The American league structure creates its own pressure. Each team plays exactly one game per week. The whole league crowds into a single weekend window. That concentration turns the weekly stat sheet into a standalone media product, reproduced every seven days, with a headline that is almost fixed: who leads.

To read that product correctly, two types of metric must be separated.

Volume metrics count actions: passing yards, rushing yards, tackles. One action, one unit. Efficiency metrics measure quality per action: yards per attempt, success rate per play, EPA, CPOE. The two groups answer two different questions. The first answers how much. The second answers how well.

The Week 1 leaderboard supplies the first group, and only the first group. The three names at the top occupy three different positions, belong to three different teams, and come with no efficiency metric attached.

The first principle I learned from a data editor still holds: never conclude from a sample, conclude from a series. Here, the series is exactly one element long.

Tyler Shough's 410 yards

Passing yards are the most easily misread of the four numbers named, and the reason lies in the structure of the sport rather than in the player.

In American football, passing volume depends directly on game state. A team that trails throws more. A team that throws more accumulates more yards. That is a causal relationship at the tactical level. It operates independently of the thrower's ability.

The Saints lost to the Lions. That is the only fact tied to the outcome of Shough's game. From that fact, a reasonable inference can be built: New Orleans most likely fell into chase mode, and chase mode inflates pass attempts. Passing yards then rise with it. The three touchdowns may belong to the same current, either as scoring plays that narrowed the gap or as plays that arrived once the result was already settled.

I do not have enough data to assert that. But I have enough data to say that the opposite conclusion, that Shough is the league's best quarterback, has no matching foundation either. Both directions exceed the evidence.

This is where absent metrics matter more than present ones. To read 410 yards, one needs the number of attempts. Four hundred and ten yards on 30 attempts is a different tier from 410 yards on 55 attempts. One needs yards per attempt, the metric that flattens the volume factor. One needs success rate per play. One needs the distribution of yards over time: how many came while the game was still balanced, how many came after the margin had opened. One needs snap count and time of possession.

None of these appear in the weekly leaderboard. That does not mean they do not exist. It means the leaderboard is built to skip them.

I have rerun models many times just to test an assumption like this, and every time, the omitted variable sat on the game-script side rather than the player side.

Jahmyr Gibbs' 156 yards

If I had to pick the cleanest signal among the three from Week 1, I would pick Gibbs.

One hundred and fifty-six rushing yards and 2 touchdowns, in a win. That combination means something. Effective rushing in a winning game usually reflects genuine control of the contest: the team leads, runs to protect the ball, and that sequence feeds itself. It is the inverse structure of Shough's case. In Detroit, the volume came from initiative. In New Orleans, the volume came from reaction.

A cleaner signal is still not a strong signal.

One game. One opponent. One script. Three uncontrolled variables. To turn 156 yards into a conclusion about a running back's value, three more things are needed: carries, the share of offensive snaps the player was on the field for, and the quality of the opposing defense that day.

A weak defense allows high rushing yards. A strong defense worn down in the fourth quarter does the same. A defense depleted by injury does the same. All three scenarios produce an identical number, and none of them says anything about the runner.

There is a simple check I still use: compare the Week 1 ranking with the cumulative ranking after Week 6. If a player is still in the leading group after six games, then there is something to discuss. Before that, the weekly leaderboard is describing noise, not signal.

Anthony Hill Jr.'s 16 tackles

Tackles are the metric most dependent on position and defensive scheme among the three named.

Linebacker is a position designed to be near the ball. A linebacker in a deep-drop system has more tackle opportunities than one in an aggressive, compressed system. A linebacker on a weak defense stays on the field longer, faces more plays, and therefore gets more chances. The tackle count measures access to the ball before it measures the quality of the man making the stop.

This does not diminish Hill Jr. He still had to be in the right place, at the right moment, 16 times in one game. But it places the number in its proper frame: 16 tackles say he had a game that matched his role. They do not say he is the league's best linebacker.

I once spent four weeks re-counting every movement of a defensive unit just because a similar leaderboard made me suspicious. The result that time: roughly one third of the league-leading tackles came from situations in which the defense had already lost its shape. The number was correct. The reading of it was the problem.

There is one point here that needs separate verification. Hill Jr. is recorded as a Tennessee Titans linebacker. His team affiliation should be checked against official NFL rosters before it is used for any other purpose. This is precisely the kind of detail that automated aggregation sheets get wrong.

The contrarian angle: the leaderboard does not measure what it claims to measure

Three metrics. Three positions that cannot be compared with one another. A quarterback, a running back, a linebacker. Placing them side by side in the same bulletin is an editorial decision, not an analytical conclusion.

The deeper issue is that the weekly leaderboard is designed to regenerate. It appears every seven days. It will have new names next week. That is not a technical flaw. It is the point of the format. A media product needs continuous flow, and that flow only sustains itself if the rankings move.

The consequence: readers absorb a sequence of short-term leaders and accumulate them into a long-term impression of quality. The feeling that a player is breaking out is constructed from the shortest-lived fragments in the entire sports statistics system.

Every stat sheet is a confession: it reveals what the person who built it chose not to look at.

There is one overlap worth noting, which I noticed after years of following both association football and American football. In football, efficiency metrics such as expected goals became mainstream within about a decade, and ordinary fans now read them the way they read a scoreline. In American football, an equivalent advanced-metrics system has existed for a long time, yet the mass-market leaderboard still runs on counting logic. Two sports, two different adoption speeds, one shared trap.

That is why I always read a weekly leaderboard as a draft: it tells you what just happened, not what will follow.

What to track from Week 4 to Week 6

Regression to the mean is the most reliable statistical principle in sport. Extreme early results tend to drift back toward average as the sample grows. This does not mean Shough will decline or Gibbs will stop producing. It means a Week 1 lead predicts a season-ending lead very poorly.

Four things are worth tracking:

Cumulative rather than weekly ranking for all three players, measured at the Week 4 and Week 6 marks.

Shough's attempt count and yards per attempt, to test whether the 410 yards came from volume or from efficiency.

Gibbs' carries and snap share, to test whether the workload is being sustained.

Hill Jr.'s tackle count placed beside his defensive snap count, to separate opportunity from quality.

If Hill Jr. is still leading after six weeks at a similar average snap count, the metric will have shifted from measuring opportunity to measuring ability. If Shough is still near the top after six weeks with a high yards-per-attempt figure, then there will be something to write about.

For now, after one game, what we have is a counting sheet. It is correct. It also has not said anything yet.

I believe in structure, but structure exists to collapse, and a good analyst is the one who predicts the exact point of collapse. For the Week 1 leaderboard, the collapse point sits at Week 4.

Data gives us a map, but only chaos points to the real road. And Week 1, in the end, is pure chaos.