Close games, the ones decided by five points or fewer, are where we go looking for luck. The logic is straightforward: in a contest that tight, the outcome often hinges on a single bounced rebound, a borderline foul call, or whether a 38-percent three-point shooter happens to hit one more shot than usual. Over a full season, a team's record in these games tends to regress toward .500. The further a team drifts from that line, the more likely variance, not virtue, is doing the work.
This season's close-game data is, to put it gently, sparse. We are working with ones and twos. That is not a flaw in the analysis. It is the analysis.
Most Cursed by Close Games
These teams won their nail-biters. Which means, historically, they banked results they're unlikely to replicate at the same rate.
1. Memphis Grizzlies - A pristine 1-0 in close games. One game. One win. One data point doing a remarkable amount of narrative lifting.
2. Milwaukee Bucks - Also 1-0. The Bucks grabbed their lone five-point-or-fewer contest, which sounds great until you remember the coin hasn't been flipped enough times to mean anything.
3. Philadelphia 76ers - 1-0 in the clutch. For a franchise whose fans have memorized every possible way a season can go sideways, winning the only close game on the ledger is a small, fragile gift.
4. Boston Celtics - 1-0. The defending-era Celtics played one game decided by a possession or two and won it. The sample is so thin you could read a newspaper through it.
5. Oklahoma City Thunder - The outlier at the top, going 2-0. Twice the data of everyone above them, which still amounts to roughly nothing. But in a landscape of single-game records, two wins is a relative mountain.
Most Blessed by Close Games
These teams lost every close game they played. That sounds bad. Historically, it means the universe owes them a few.
1. Miami Heat - An 0-2 close-game record leads the blessed list. Two losses in two tight finishes suggests the Heat left a win or so on the table through sheer randomness.
2. Chicago Bulls - 0-1. One close game, one loss. The basketball gods took their shot and missed, which is fitting.
3. Dallas Mavericks - 0-1. A single close-game loss. In a vacuum, meaningless. In a regression model, a quiet little promise.
4. Brooklyn Nets - 0-2, matching Miami's futility in the margins. Two games that could have swung either way both swung away.
5. Washington Wizards - 0-1. Washington lost its only close contest, which for this franchise barely registers on the suffering index.
What This Tells Us
The entire range of this season's close-game signal spans from 2-0 to 0-2. That is four games of separation between the most cursed and most blessed teams in the league. In any normal year with a fuller sample, you would see teams at 12-4 or 3-9 in close games, and the regression stories would write themselves. Here, the samples are almost comically small.
But the principle holds regardless of volume. Research going back decades shows that close-game records are among the least sticky metrics in basketball. Teams that go on heaters in tight finishes one year tend to cool off the next, and teams that lose a string of heartbreakers tend to start grabbing a few. A .500 close-game record is not a ceiling or a floor. It is gravity.
So what does this tell us heading into the playoffs? Mostly that Oklahoma City's 2-0 is the softest foundation to build confidence on, and Miami's 0-2 is the softest foundation for despair. The numbers are real. The sample just isn't.
