Sunday, February 27, 2011

All-Star Break Analysis, Pt. 2: Winning Streaks After the Trade Deadline

There was a lot of hype surrounding the Miami Heat before the season started and the obvious question after the All-Star break is this, "How did they do living up to the hype?"

There were a few predictions hyped on this blog:

  1. Total wins for the season
  2. Consecutive wins during the season
  3. Attendance & playoff seed

This article is the second in a three-part series that will evaluate how the team has measured up to the pre-season hype.

Before the season started, ABC's Jeff Van Gundy made the following prediction about the Miami Heat:

"They will break the single-season win record [of 72]," Jeff Van Gundy said. "And I think they have a legit shot at the Lakers' 33-game [winning] streak [in 1971-72], as well."

Why the Heat won't win 70+ wins was analyzed in Part 1. As for winning 33-games in a row, that didn't happen either as they only won 12 in a row. However, as I said in the original post discussing Van Gundy's prediction, it wasn't realistic anyway.

The more interesting winning streak for the Heat was 13 games. According to Hot Hot Hoops, the over/under at Bodog was 12.5 for the Heat's longest win streak of the regular season.

With 23 games left in the season, do the Heat have a good chance of winning 13 in a row?

This article will use Win Score and Wins Produced, statistical models created by Professor David Berri from the Wages of Wins Journal, to measure how much a player's box score statistics contributed to their team's efficiency differential and wins. An average player produces 0.100 wins per 48 minutes (WP48), a star player produces 0.200+ WP48 and a superstar produces 0.300+ WP48. More information on these stats can be found at the following links:

Simple Models of Player Performance
Wins Produced vs. Win Score
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics

A formula developed by Arturo Galletti to measure the probability of a team winning a single game will be used to evaluate the Heat's chances of running off a 13-game winning streak before the season ends. The great thing about Arturo's formula is that it can use the average WP48 of each team's players to calculate their probability of winning and that means it will be simple to measure each team's strength after the trade deadline.

The Heat go into their final 23 games with an average WP48 of 0.151 and play 14 teams that made trades shortly before the trade deadline. This spreadsheet shows the probability of the Heat beating each of those teams.

There are three dates in the Heat schedule where they have a 50 percent chance or less of winning:
  • Game 62 on Fri, Mar 4, 2011 @ San Antonio Spurs (34 percent chance)
  • Game 69 on Fri, Mar 18, 2011 @ Atlanta Hawks (50 percent chance)
  • Game 81 on Mon, Apr 11, 2011 @ Atlanta Hawks (43 percent chance)

The longest winning streak the Heat could be expected to put together with those three potholes in the schedule is 11 games. Expecting a streak longer than that would likely require an upset of the Hawks in Atlanta on March 18th.

Of course, in order to take advantage of an upset, the Heat have to win tough games against the Magic (56 percent chance) and Spurs (58 percent chance) in Miami during the month of March.

Before the season started, it looked like the Heat were a lock to win 13 or more games in a row, but a loss to the Dallas Mavericks on December 18 forced them to settle for 21 wins in 22 games. After the trade deadline, it looks like the Heat will have a 50/50 shot to beat an improved Hawks team on the road and live up to the hype.

All of the Wins Produced stats used in this article were Powered by Nerdnumbers. Roster and trade information was taken from NBA schedules were taken from

No comments:

Post a Comment