Dunk Base

By now, I’m sure you’ve heard the stats that the Miami Heat are only 6-13 in games decided by 5 points or less, and that they are 1-18 on shots that would either tie or win the game in the 4th Quarter or Overtime.

On the surface, these seem like pretty damning statistics. After all, teams that can’t close out close games don’t go on to postseason success, right?

Actually, wrong.

There are a number of problems with the two measurements mentioned above. The first is what is known as an arbitrary endpoint. Yes, the Heat are just 5-13 in games decided by 5 points or less. But what exactly does that mean? Why only games decided by 5 points or less? Why not 6? Is there some magical cutoff line that says a game decided by 5 points or less is a “close game,” but a game decided by 6 points isn’t? Because I really don’t see a distinction. After all, a 6-point game is still a 2-possession game. It would seem like those should be included under the umbrella of “close games.”

In fact, if you take the 5 or less measurement, and simply add one possession onto it (to make it games decided by 8 or less), Miami’s record jumps from 6-13 to 17-15. That’s right, in games decided by 6-8 points, Miami is 11-2. Choosing 5-point games as the cutoff line is a classic case of picking an arbitrary number just because it looks good on a page.

To illustrate my point, I’ll give you two test cases. The first will be a game between Dallas and Minnesota on March 7th. With 10 seconds left in the game, Dallas led 106-99. The game was well in hand. Teams usually don’t make up seven point leads in 10 seconds. So what happened? Minnesota made a 3 (four point game), fouled Dirk, who made both free throws (six point game), and then nailed a half-court heave at the buzzer to make the final score 108-105. So basically, according to the statistic, Dallas gets credit for winning a “close” game, despite the fact that the result was never in question. Doesn’t really make sense, does it?

Want another example? How about one using the very team under the microscope – Miami? Let’s use their March 10th game against the Lakers. The game was tied at 88 with under two minutes to play. Miami forced a turnover and gets a layup in transition, then forces three straight misses on the other end. Back on offense, Wade makes a pretty move for a layup. Miami is up 4 with 46 seconds left. Still anybody’s game. After trading turnovers, Los Angeles gets the ball back, still down 4 with 25 seconds left. Kobe misses a three, and Ron Artest fouls LeBron, and James knocks down both freebies. Six point game. The game ends 20 seconds later with a final score of 94-88. Again, according to this statistic, Miami didn’t win a “close” game, despite the fact that the game was tied with 1:29 remaining before Miami went on a run to seal the victory. And again, that doesn’t totally make sense.

On top of that, Miami has had a number of games like the one against LA. January 9th vs. Portland, Miami trialed 91-84 with 1:46 remaining, went on a quick 9-2 run, forced overtime, and in overtime held at least a 2 possession lead for the last 2:12, and ended up winning by 7. But I suppose that wasn’t a “close” game.

Furthermore, the idea that teams that don’t perform well in “close” games don’t perform well in the Playoffs simply isn’t true. Take a look at the following teams, their record in “close” games, and how they ended up doing in the Playoffs:

And before you guys come back with a bunch of examples where teams had poor records in close games and then went on to get bounced in Round 1, go ahead and look up what the scoring differential for those teams were. For the most part, I’m betting it wasn’t that great. For example, the 2006 Pacers were pretty abysmal in close games (8-22), and lost in Round 1, but that team wasn’t even good to begin with. They went 41-41 overall and their point differential was only +1.9. Similar story with the 1996 Heat (6-15 in close games). They went 42-40 overall, and their differential was only +1.5. This Miami team has a differential of +7.2. At the end of the day, there isn’t much of a correlation between “underperforming in close games” and “underperforming in the Playoffs.”

The other statistic I mentioned in the beginning has even less of a correlation. Yes, Miami is the worst in the league at converting game-winning shots with the clock winding down. But guess which teams are near the top of the list? Golden State, Cleveland, Phoenix, and Milwaukee make up four of the top 8. None of those teams are playoff teams. Conversely, New Orleans, Portland, Philadelphia, and New York are four of the nine WORST teams at converting last-second shots, and all of those teams are headed for the postseason.

Furthermore, the league AVERAGE shooting percentage in those situations is only around 24% to begin with. So basically everyone in the entire league, with the exception of a handful of players, “choke” in the clutch. Or maybe the stats just don’t tell us that much because of the miniscule sample size. These are the same stats that tell us that Charlie Villanueva converts more of his last-second shots than Kobe does. Obviously, Charlie Villanueva is not a better clutch shooter than Kobe Bryant. But at the same time, we can’t have the stats apply to the Heat but not apply to Villanueva and Kobe. They either apply to both or they apply to neither. I’m leaning towards neither. When we’re talking about 18 shots, each within the last five seconds of a game, that’s 90 seconds out of a regular season that’s over 200,000 seconds long. So pardon me if I don’t put much stock into those numbers.

On the other side of the coin, you have teams like the ’07 Mavericks. They went 67-15 during the regular season, but did disproportionately well in close games and in last-second situations (20-4 in games decided by 5 points or less). They had the best record in the league by six games, but their scoring differential was only 3rd best in the league. As we know now, Dallas went on to get upset in Round 1 by an inferior Golden State team and lost three of the four games by double digits.

If you want to take an accurate snapshot of a team’s quality, you can’t just pick and choose which games matter more than most. It’s better to look at numbers that encompass the entirety of the season. For instance, Miami leads the league in scoring differential (and all the stats that are based off of it, like Pythagorean Win% and Simple Rating System). They lead the league in points per shot differential. They’re 2nd in the league in FG% differential. They’re tied for 2nd in the league in 3PT% differential. Those are the types of numbers that correlate to postseason success, not leading statistics that become largely invalid because of a small sample size or arbitrary endpoints.

Comments (4)