There is a moment every September when a league table starts to exist but cannot yet be trusted, and the Atlantic Coast Conference is in it. Its standings already separate seventeen teams into eight groups. They also conceal the single fact that decides what any of it means: the teams in that table have not played the same number of games in it.
Two records, not one
Every team in this sport carries two records at once. The overall record counts every game it has played. The conference record counts only the games against members of its own league — and only that second number decides who plays for the league title. The college football guide sets out the wider structure; the practical consequence is what matters here.
The Atlantic Coast Conference’s own standings, as published by the conference, show how far apart those two numbers can sit. Virginia Tech is 3–0 overall and 0–0 in the conference: unbeaten, and absent from the race entirely. Georgia Tech is 1–2 overall and also 0–0 in the conference: a losing September that has cost it nothing in the standings. Boston College is 2–1 overall with no conference games played.
At the other end, Miami is 2–0 in conference play and Stanford and Syracuse are both 0–2. Between those extremes sit six teams on 1–0 and one, SMU, on 1–1.
Add the wins across that table and the conference has completed nine league games. The conference’s release of September 22 confirms the same figure and adds four more scheduled for the following weekend.
Why the numbers are uneven
This is not an accident or an error in the table. It is how the schedule was built: conferences stagger which teams open league play on which weekend, so a September table is always a partial one.
The contrast with another league makes the point without needing a league-by-league census. The Big Ten’s own release describes conference play there arriving on September 19, with two games. On the same date the ACC had already played most of its nine.
So a reader comparing the two tables in the third week of September is not comparing two races at the same stage. They are comparing a race that has started with one that is starting. The site’s analysis of the Big Ten’s opening covers the other side of that. The more general reason a schedule’s shape changes what a record means is set out in an earlier piece on the 2026 calendar.
What the early results have actually done
The most useful thing the nine games have produced is not a leader. It is a set of results that do not match what the conference expected of itself.
In the conference’s preseason poll, voted on by 188 credentialed media members in July, Miami was picked to win the championship with 165 of the first-place votes and 3,159 points. SMU was picked second, Louisville third, Clemson fourth and Georgia Tech fifth. That is the conference’s own record of expectation, not this publication’s ranking — and it is worth reading beside the preseason view of the season as a whole, which deliberately declined to forecast winners at all.
Set that against the standings now. Miami is the only team at 2–0, which is consistent with the poll. SMU, picked second, is 1–1: it already carries a conference defeat that no one else in the top half of the poll has. Georgia Tech, picked fifth, has not played a conference game and is 1–2 overall. Louisville and Clemson are each 1–0.
Meanwhile three teams that were not in the poll’s top five — Duke, Pitt and California — are each 1–0 in the league and 3–0, 3–0 and 2–1 overall.
The poll itself recorded how little its voters agreed. The conference described it as one of the closest in recent memory, with nineteen points separating sixth place from eighth. Stanford finished seventeenth and still received a first-place vote; Florida State finished eleventh and received three. Those two teams now sit 0–2 and 0–1 in the conference respectively — which says less about the voters than about how thin the evidence is in July.
This is interpretation rather than a conference finding: the early evidence has not overturned the preseason view, but it has stopped confirming it. One team is doing what was expected. Several others are doing something the poll did not anticipate, and a few of the teams expected to matter have not yet been asked a conference question at all.
Miami, and the limits of what its record proves
Miami is the team the table currently rewards, and the team whose record needs the most careful reading.
Its own athletics department records three wins: 45–6 at Stanford, 77–7 at home against Florida A&M, and 33–20 at Wake Forest. Two of those were conference games, which is why Miami sits alone at 2–0. Its quarterback Darian Mensah appears in the conference’s weekly release as the league’s passing-efficiency leader with eleven touchdowns and no interceptions.
What that establishes is a start without a blemish. What it does not establish is anything about the rest of the fall, and Miami’s own published schedule is the reason for caution: after a home game against Central Michigan on September 26, the Hurricanes visit Clemson, host Florida State and Pittsburgh, travel to North Carolina, and then go to Notre Dame on November 7 before finishing against Duke, Virginia Tech and Boston College.
That is a schedule whose difficult half has not begun. A reader can note that Miami has not lost, note that it has not yet played the teams most likely to beat it, and hold both thoughts at once. Nothing in the accessible official record supports a claim about where Miami will finish, and this article does not make one.
The individual performances worth knowing
Two names come out of the conference’s own weekly release rather than from anybody’s opinion.
Duke’s Nate Sheppard leads the entire top division of college football in rushing yards, with 528, and has eight rushing touchdowns. He scored three of them in Duke’s 35–7 win over Stanford, a conference game — which is how a team picked outside the poll’s top five arrived at 1–0 in the league.
North Carolina’s Xavier Lewis leads the conference with six sacks, on a team the standings show at 0–1 in conference play and 2–1 overall. That combination — an individually dominant player on a team with a league defeat already — is a fair summary of why individual statistics and conference standings answer different questions.
What the next weekend starts to resolve
Four more conference games follow on the weekend of September 26, according to the conference’s release: Clemson at California on the Friday night, Virginia Tech at Boston College, Wake Forest at Louisville, and Georgia Tech at Stanford.
Three of those involve teams that have played no conference football at all. By the end of that weekend, Virginia Tech, Boston College and Georgia Tech will each have a conference record that means something, and the table will be closer to comparable than it is today.
But only a little closer, and this is the part most likely to mislead. The same release lists the rest of the conference’s weekend, and most of it is not conference football: Central Michigan at Miami, Missouri State at SMU, Appalachian State at NC State, Bucknell at Pittsburgh, Central Arkansas at Florida State, William & Mary at Duke and Delaware at Virginia. Seven of the seventeen teams will finish the weekend with exactly the conference record they started it with.
So a reader returning to this table on Sunday will find it slightly fuller and still uneven — which is the normal condition of a conference table in September, rather than a flaw in it. The rankings attached to some of those names in the conference’s release, Miami at sixth, Louisville at sixteenth and SMU at twenty-second, are national poll positions and belong to a separate conversation from the standings.
That is the honest answer to the question this article set out to ask. The ACC race is genuinely underway, earlier than the calendar suggests and earlier than several other leagues. It has produced one clear early leader and several results the conference itself did not forecast. It has not yet produced a hierarchy, because a third of its members have not entered it.
For the games being played across the country on the same afternoon, the weekly preview carries the current national slate; the polls referenced above are tracked on the rankings desk. Neither is a substitute for the standings — and in late September, the standings are not yet a substitute for a season.