Reading an NBA table: games behind, the play-in and ties
A basketball table looks simpler than a football one: wins, losses, a percentage. Three of its columns still trip people up. Here they are, with the final 2025-26 tables as the example.
curl -H "X-RapidAPI-Key: YOUR_KEY" \
-H "X-RapidAPI-Host: nba206.p.rapidapi.com" \
"https://nba206.p.rapidapi.com/standings?season=2025-26"
The top of the East, trimmed to the columns that matter here:
| # | Team | W | L | winPct | gamesBehind | zone |
|---|---|---|---|---|---|---|
| 1 | Detroit Pistons | 60 | 22 | 0.732 | 0 | playoffs |
| 2 | Boston Celtics | 56 | 26 | 0.683 | 4 | playoffs |
| 3 | New York Knicks | 53 | 29 | 0.646 | 7 | playoffs |
| 7 | Philadelphia 76ers | 45 | 37 | 0.549 | 15 | play-in |
| 8 | Orlando Magic | 45 | 37 | 0.549 | 15 | play-in |
| 11 | Milwaukee Bucks | 32 | 50 | 0.39 | 28 | null |
The table is ordered by percentage, not by wins
Teams play different numbers of games at any point in the season, so wins alone would rank a team
that has played more above a better one that has played less. winPct is wins divided by
games played, rounded to three places, and it is what the table is sorted on. In JSON it is a
number, so 0.390 arrives as 0.39. Pad it yourself if you want the newspaper look.
Games behind, and how half a game happens
gamesBehind is the distance to the leader of the group:
const gamesBehind = (leader, team) =>
((leader.won - team.won) + (team.lost - leader.lost)) / 2;
gamesBehind({ won: 60, lost: 22 }, { won: 56, lost: 26 }); // 4
It averages two gaps: how many more games the leader has won, and how many more the team has lost.
When both have played the same number of games the two gaps are equal and the result is a whole
number. During the season they rarely have, and a team one win and no losses behind is 0.5 games
back. It can even go below zero: a 15-4 team is a game and a half "ahead" of a 10-2 leader by this
sum, and still below it on percentage, .789 to .833. That is why the order follows
winPct, not this column.
The zone column is the play-in
In each conference, places 1 to 6 go straight to the playoffs and carry
"zone": "playoffs". Places 7 to 10 play the play-in tournament for the last two
playoff spots and carry "play-in". Everyone else is null. The EuroLeague
works the same way across its single table of 20: 1 to 6 to the playoffs, 7 to 10 to its play-in.
Draw a line under 6 and 10 and you have the picture people expect.
The zone is set by position, so before the last day it shows where a team stands, not what it has clinched.
Two teams on the same record
Philadelphia and Orlando both finished 45-37. Over the season Orlando outscored its opponents by 52 points and Philadelphia was outscored by 15, yet Philadelphia is 7th. It won two of the three games between them, and the head-to-head record is the first tie-break. Point difference comes after the division and conference records in the NBA's list, so it rarely gets a say.
Note what head-to-head means in the NBA: games won, not points. Philadelphia was outscored by 17 over those three meetings and still holds the tie. The EuroLeague differs here. It looks at head-to-head wins first too, but when those are level the point difference in the games between the tied teams decides, before anything over the whole season.
Further up, Toronto and Atlanta finished 46-36 and Toronto is 5th: it won all four meetings. Both made the top six, so that tie settled first-round opponents rather than the play-in.
Other tables
The default NBA table is by conference, which is what the play-in follows. Add
?table=division for the six divisions or ?table=league for all 30 teams in
one list. Each row also carries home and away records,
lastTen, form (up to the last ten results, oldest first) and
streak, such as W3, so a full standings page needs no second call.