Any time the Golden State Warriors lose, it’s big news — that’s what happens when a team is battling ghosts, trying to stay ahead of the Chicago Bulls’ record pace from two decades ago. But for the loss to come against the lowly Los Angeles Lakers? They of the 12-51 record, a full 44½ games worse than Golden State’s tally going into the matchup? It bordered on the unthinkable, particularly because basketball sees fewer upsets than other sports.
Yet this kind of loss does happen from time to time in the NBA, and to teams favored even more heavily than Golden State was on Sunday. According to our Elo ratings, which measure the strength of each NBA team and can be used to estimate the odds of victory when any two teams face off, the Lakers went into the game with a 6.5 percent chance of upsetting the Warriors. In the entire history of the league, that’s only the 23rd-lowest pregame win probability for a team that won:
| WINNER | LOSER | ||||||
|---|---|---|---|---|---|---|---|
| DATE | TEAM | ELO | TEAM | ELO | SCORE | CHANCE OF UPSET | |
| 1 | 4/24/93 | Mavericks | 1160 | Rockets | 1682 | 128-123 | 2.7% |
| 2 | 4/6/93 | Mavericks | 1131 | SuperSonics | 1646 | 109-107 | 2.8 |
| 3 | 3/21/92 | Magic | 1288 | Bulls | 1782 | 111-108 | 3.2 |
| 4 | 3/21/90 | Hornets | 1235 | Suns | 1678 | 115-114 | 4.2 |
| 5 | 2/22/72 | Pistons | 1298 | Lakers | 1737 | 135-134 | 4.3 |
| 6 | 1/1/60 | Royals | 1320 | Celtics | 1728 | 128-115 | 5.1 |
| 7 | 4/16/14 | 76ers | 1215 | Heat | 1620 | 100-87 | 5.2 |
| 8 | 12/1/94 | Timberwolves | 1235 | Jazz | 1634 | 96-94 | 5.4 |
| 9 | 2/13/98 | Mavericks | 1279 | Pacers | 1677 | 85-82 | 5.4 |
| 10 | 2/6/02 | Bulls | 1299 | Lakers | 1693 | 97-89 | 5.5 |
| 11 | 12/7/48 | Steamrollers | 1191 | Royals | 1583 | 90-89 | 5.6 |
| 12 | 2/18/12 | Nets | 1306 | Bulls | 1694 | 97-85 | 5.7 |
| 13 | 4/12/94 | Mavericks | 1210 | Blazers | 1597 | 108-107 | 5.7 |
| 14 | 12/29/85 | Warriors | 1361 | Lakers | 1747 | 130-122 | 5.8 |
| 15 | 3/27/81 | Pistons | 1303 | Celtics | 1679 | 115-90 | 6.1 |
| 16 | 4/16/14 | Lakers | 1357 | Spurs | 1732 | 113-100 | 6.1 |
| 17 | 11/4/80 | Pistons | 1281 | Bucks | 1656 | 98-96 | 6.1 |
| 18 | 11/23/67 | SuperSonics | 1267 | Celtics | 1641 | 133-106 | 6.2 |
| 19 | 4/13/15 | Knicks | 1249 | Hawks | 1619 | 112-108 | 6.3 |
| 20 | 2/18/72 | 76ers | 1356 | Bucks | 1723 | 132-126 | 6.4 |
| 21 | 11/29/69 | Pistons | 1346 | Knicks | 1712 | 110-98 | 6.4 |
| 22 | 4/19/97 | Grizzlies | 1200 | Suns | 1566 | 121-107 | 6.4 |
| 23 | 3/6/16 | Lakers | 1259 | Warriors | 1823 | 112-95 | 6.5 |
| 24 | 11/29/09 | Timberwolves | 1285 | Nuggets | 1645 | 106-100 | 6.6 |
| 25 | 2/17/53 | Bullets | 1280 | Lakers | 1638 | 74-70 | 6.7 |
| 26 | 3/25/02 | Grizzlies | 1296 | Blazers | 1654 | 103-100 | 6.7 |
| 27 | 2/27/10 | Nets | 1237 | Celtics | 1595 | 104-96 | 6.7 |
| 28 | 2/2/93 | Timberwolves | 1281 | Blazers | 1637 | 104-91 | 6.8 |
| 29 | 4/8/08 | Knicks | 1302 | Pistons | 1657 | 98-94 | 6.8 |
| 30 | 2/9/07 | Hawks | 1374 | Suns | 1728 | 120-111 | 6.8 |
| 31 | 4/19/00 | Clippers | 1174 | SuperSonics | 1524 | 114-106 | 7.0 |
| 32 | 1/22/71 | Hawks | 1350 | Bucks | 1696 | 117-110 | 7.1 |
| 33 | 1/4/00 | Grizzlies | 1258 | Heat | 1601 | 91-87 | 7.2 |
| 34 | 1/18/91 | Nets | 1297 | Celtics | 1640 | 111-106 | 7.3 |
| 35 | 4/18/01 | Bulls | 1237 | 76ers | 1579 | 92-86 | 7.3 |
| 36 | 11/23/90 | Nets | 1281 | Suns | 1621 | 116-114 | 7.3 |
| 37 | 4/11/99 | Clippers | 1282 | Blazers | 1622 | 89-83 | 7.4 |
| 38 | 1/23/72 | Hawks | 1374 | Bucks | 1714 | 118-113 | 7.4 |
| 39 | 4/16/12 | Wizards | 1335 | Bulls | 1675 | 87-84 | 7.4 |
| 40 | 4/12/89 | Hornets | 1228 | Knicks | 1566 | 104-99 | 7.4 |
A couple of times in April 1993, the disastrously bad Dallas Mavericks entered games with greater than a 97 percent probability of losing — and won! In those games, the Mavs averaged a win probability nearly three times lower than the Lakers had against the Warriors on Sunday. A couple of percentage points may not seem to matter much, but they add up when we’re talking about the likelihoods of extremely rare events.
It’s also worth noting that LA’s victory over Golden State is unique in one other important respect: The Lakers are the only home team on our list. Every other top-40 upset was pulled off away from a team’s home court, which makes sense because home-court advantage is baked into the Elo formula. Starting the game on the road is a significant built-in disadvantage — between two otherwise evenly matched teams, Elo would expect the road team to win only 36 percent of the time — and, conversely, being at home is such an advantage that it’s difficult to be a very heavy underdog there. There’s a case to be made that this formulation shortchanges home underdogs by not giving them enough credit for their upset wins, but the gap between the Warriors and Lakers was so great that, even though LA was at home, it still managed to have a 93.5 percent likelihood of losing.
And somehow, the Lakers still won. Absurdly implausible results like that don’t happen often in the NBA, but they still occur frequently enough to remind us that even the most unbeatable of teams has to lose every once in a while.
