Sunday, June 12, 2011
Analyzing the closeness of the NBA Finals
By Dean Oliver and Alok Pattani
ESPN Stats & Info
As the 2011 NBA Finals come to a close with Game 6 on Sunday and a possible Game 7 on Tuesday, basketball fans everywhere are bemoaning the fact that such a competitive and compelling series must come to an end. Through five games, the Miami Heat and the Dallas Mavericks have battled with the score rarely being even remotely one-sided, and the outcome in doubt until virtually the final seconds in each game.
Just how competitive has this series been? Let’s start with some basic facts:
• Games 1-5 have featured 50 ties and 44 lead changes.
• The largest lead for either team at any point was Miami’s 15-point lead in the fourth quarter of Game 2 – a game that the Heat went on to lose in stunning fashion.
• The second largest lead in the series took place in Game 3, when Miami led by 14 towards the end of the second quarter. The Heat did go on to win that game, but by two points.
• Neither team has led by double digits at any point in each of the last two games. In fact, even though they are leading the series three games to two, the Mavericks have yet to hold a double-digit lead at any point in any game!
While these facts are illustrative of how close the series has been, a more nuanced calculation using play-by-play data can help show literally how close the games have been at every second. This can be done by looking at the absolute score margin – the positive difference between the two teams’ point totals at every point in the game – and weighting it for how long that margin persisted on the game clock.
For example, if the score was within five points with two minutes left, went down to two points at 1:30, increased to four with 1:00 left, and then jumped to seven with 0:30 left with no scoring after that, we’d have 30 seconds each of score margins of two, five, four, and seven. The average time-weighted score margin for this two-minute period would be the average of those four numbers (since each accounted for an equal 30 seconds of game), or 4.5 points. Extending this calculation for the entire game – or an entire series of games – we can see just how close the score has been throughout, on average.
It should be noted that the average playoff game over the last nine years has a time-weighted scoring margin of 7.44 points. This means that at any given time in a playoff game, the teams are separated by 7.44 points on the scoreboard, on average. The chart provides a look at how close the score has been throughout the first five games of this year’s Finals.
Each game has had an average score margin less than five points, and the average margin over the course of the series is an anemic 3.62 points.
The next question is, of course, how does this series stack up when compared to other really competitive series in playoff history? Due to the limitations of play-by-play data, we are only able to calculate this average in-game scoring margin for playoff games back to the 2003 postseason, which also happens to be the first year that all playoff series went to seven games. That gives us a total of 135 playoff series for which the average score margin at every second was calculated.
As you can see, this series through Game 5 ranks second in our sample, trailing only the 2007 Eastern Conference Finals between the Detroit Pistons and the Cleveland Cavaliers. That was the series when LeBron James famously led Cleveland to an upset of Detroit in six games on the strength of an amazing performance in a double-overtime Game 5.
In fact, the most amazing part of this list is that James has been involved in each one of the closest four series over the last nine years! In addition to the top two series already mentioned, this group includes the Heat’s victory over the Chicago Bulls just a few weeks ago, as well as the Cavs’ six-game victory over the Washington Wizards in LeBron’s first career playoff series back in 2006. Keep in mind that while “The King” has been widely criticized for his lack of clutch play in the current NBA Finals, his team did come out on top in the other three super-close series shown above, at least partially due to his ability to come through in “crunch time.”
While looking at average score margin does tell you which series was closest from second to second of each game, it doesn’t actually account for how close the series itself was in terms of both teams winning games. For example, the Bulls-Heat series that comes in third was a 4-1 victory for the Heat, and retrospectively doesn’t seem to be that “close” even if each of the games were since it only went five of a possible seven games. None of the series on the list above even went to a Game 7. Where is that Celtics-Bulls classic first-round series with so many overtimes from just a couple years ago?!
Fortunately, there is a good way to adjust for that too, by accounting for how many minutes were played within such a close scoring range in the series. In a more advanced version of this calculation, a series with a lower than average score margin that went seven games would be given a higher “closeness” score than one that went five games with the same average margin because the first series had more basketball played at the same per-minute (or per-second) level of closeness.
We’ve converted this value to a “Series Closeness Rating” on a 0-100 scale to help ease the interpretation. A series with a score closer to 100 was really close, a series with a score approaching 0 was not close at all, and a series with a score around 50 had a scoring margin around the league postseason average of 7.44. The chart to the right shows the five closest playoff series in our sample using the “Series Closeness Rating.”
While the top spot still belongs to the 2007 Eastern Conference Finals, that memorable Bulls-Celtics series from two years ago jumps all the way up to second place. The average margin in that series was 4.81 points – a full point above where this year’s Finals stands now – but the fact that the teams played an incredible 371 minutes (the equivalent of 7.7 regulation-length games) was enough to surpass the current Heat-Mavericks series, at least for the moment. You can also see that the Mavericks-Spurs Conference Semifinals – another extremely competitive seven-game series in recent playoff history that was ranked third-best since the merger by ESPN.com’s John Hollinger just last week – rises into the top five using this rating.
In the end, the numbers support what the eyes have told us from just watching the games: this year’s NBA Finals has been one of the closest postseason series in recent memory. The best part is that it’s not even over yet.