Nine Places Apart

There are two Atlanta defences on last season's record, and anybody previewing this one gets to pick which team to tell you about.

One of them ranked ninth in the league in defensive rating. The other ranked eighteenth in points allowed. Same roster, same eighty-two games, nine places apart, and both numbers are correct.

The gap is pace. Atlanta scored 118.5 a game and gave up 116.0, and a team playing that fast gets punished in the raw column for possessions it chose to create. Adjust for them and Atlanta's 113.77 has eight teams below it and nobody tied. Leave it alone and this is a mid-table defence.

This is not a complaint about outsiders. It is a problem about instruments. The team a national writer watches on a scoreboard and the team an analyst reads in a model are genuinely different teams, and a city has to live inside both at once.

theScore runs the win total both ways. Its over case: "Adding Lu Dort should strengthen a competent core that boasted the league's ninth-best defensive rating last season." SI's July projection calls that same unit "an admittedly weak perimeter defense" and cautions that "Dort alone isn't going to make them a top-ten defense." theScore names its metric. SI names none, and its verdict only holds in the raw column. One is reading the rate, the other the total, and the same defence answers to both.

The rest of the market works the same way. BetMGM set the number at 43.5 and then published the under as its own pick. Its October 9 post prices the over at -120 and the under at +100 and takes the under anyway; two days earlier the same writer had the under at -110 on its Hawks page, a pick dated August 24. Pricing a number and believing in it are different jobs, and a book does both in public. The 43.5 has held for seven weeks; the cents did not, sliding from -110 in that August pick to +100 by the ninth, so take every price here as the artifact of its publication date, not as this morning's market.

DocSports took the over at +100 on September 16, on the grounds that Atlanta "surpassed this total with a 46-36 record last year after starting the season 26-30." BetMGM's stated reason for the under is an unwillingness "to put too much stock into the Hawks' late-season push."

Those are the same twenty-six games. A 26-30 record at the break became 20-6 after it, including 13-2 in March, and then the season ended in a six-game first round that finished 140-89 at home. The over case and the under case are one stretch of basketball, valued differently. Nothing else is really in dispute.

Yahoo Sports projects a 43-39 finish, half a game off a line it never mentions, and adds that "fifth seed feels about right in a conference that got considerably heavier at the top." SI has Atlanta seventh and in the play-in, clearing forty wins but not forty-five.

Underneath sits one weighing. Luguentz Dort and Dyson Daniels against Zaccharie Risacher and Jonathan Kuminga, in an East that SI's piece counts Miami, Toronto, Philadelphia, Charlotte and Washington as having improved. The defensive half is real: Dort was All-Defensive first team in 2024-25 and fourth in Defensive Player of the Year voting, on the same five-man list as Daniels, for a different franchise, and Daniels has gone back to back, first team then second. The cross-examination is just as real. Dort made neither All-Defensive team last season, his defensive win shares slipped from 3.2 to 3.0, his minutes fell to a post-rookie-year low of 26.8, and he shot .385. Risacher, the No. 1 pick in 2024, averaged 9.6 points and 3.8 rebounds. Kuminga lasted sixteen regular-season games here.

In August, writing about two roster cuts, I said non-guaranteed contracts hitting the waiver wire "does not move a win total." Six weeks on, the win total is the whole conversation, which is a fair description of what a preseason produces.

My lean is over, and it is not a cheerful one. BetMGM's own split has Atlanta at 23-6 against teams below .500 and 23-30 against winning teams, and the four numbers sum exactly to last season's finish. A defence already inside the top ten before Dort arrived will keep beating the bottom of the league, and the over gets bought right there — in games nobody will remember, with Trae Young now playing in this division.

Atlanta has been almost there for a decade. What is new is that almost now depends on which column you read it in. On October 21 at Orlando, somebody starts filling in the only one that counts.