< Back to front page Text size +

Rondo under the microscope: Are the Celtics better without him?

Posted by Andrew Mooney  February 28, 2013 10:15 PM

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

On January 27th of this year, Celtics fans across the country watched their favorite team beat the rival Miami Heat in an overtime thriller. Not many people celebrated, however, as the team's star point guard, the future of the franchise, discovered that he had a torn ACL and would be out for the season. Analysts and fans alike wrote the Celtics off for the season. How could this 21-23 team make a playoff run? Would Danny Ainge blow up his team and give up on this season? The team has gone 9-4 since Rondo tore his ACL, putting them at 12-6 without him on the season. To put this into perspective, with Rondo as a starter, the team has gone 18-21. If the win-loss ratio of the team without Rondo is much higher than it is with him, are the Celtics a better team without him?

The idea of a team being better without its "best" player is not a novel idea—Bill Simmons termed it the "Ewing Theory." Before he went down, Rondo was the league-leader in assists and triple-doubles (he still leads the league in both categories). However, does this really mean he is one of the most valuable players in the league? The team with the league-leader in assists per game hasn't won an NBA championship since 1987, when the Lakers accomplished the feat.

But, how can we determine if the team needs Rondo?

First, let's take a look at the team overall and what exactly they have done better without him. The following stats have been pulled from http://www.nbawowy.com: With Rondo on the court, this season the team has averaged 1.011 points per possession. Without him on the court, the team has averaged 1.048. The team averages .5 fewer turnovers per 100 possessions without Rondo than they did with him. When it comes to rebounding, this Celtics team is much better without Rondo, averaging 9.6 offensive rebounds per 100 possessions, 34.6 defensive rebounds per 100 possessions, and 44.2 total rebounds per 100 possessions. With Rondo, the team averaged 8.6, 32.9, and 41.5 in the same categories.

Obviously some key numbers have fallen with Rondo on the sidelines. The team averages 1.6 fewer assists per 100 possessions and 0.9 fewer steals per 100 possessions.

How much do these differences impact the team overall? It's hard to determine whether or not there is a statistically significant difference in the core statistics. The best way to determine whether the team has any significantly different performance is to run a t-test on several players’ plus-minus with and without Rondo as well as the team's plus-minus. The results were very interesting.


At the .05 significance level, Paul Pierce, Jason Terry, Kevin Garnett and the team as a whole are all playing statistically better (individually and as a team) when Rondo is not playing.

But does this really mean that the team is better without Rondo? The sample size is really too small right now to suggest that the team is better. However, is the team playing better right now? Statistically, yes. It seems that the old veterans on the team want to make sure that they do not go down without a fight.

“National TV Rondo”

One criticism leveled at Rondo over the past year and a half or so is that he only "tries" during Nationally Televised games. The claim is backed by evidence that the majority of his triple-doubles occur when he plays on national television. I decided to test this theory by compiling data over the past two years and running a t-test using his non-nationally televised stats and his nationally televised stats at the .05 significance level to see if his stats were actually "better" when he played on national television. The results seemed counter-intuitive to what the critics claim.


Although all of his averages are clearly better in nationally televised games, we statistically cannot say that there is a difference in how well he plays. But though the p-values are not technically “statistically significant,” they are low enough to warrant a closer look. Perhaps with more data—more nationally televised games—in the sample, the effect would be more visible.

This blog is not written or edited by Boston.com or the Boston Globe.
The author is solely responsible for the content.

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

Stats Driven is powered by David Sabino, who over the last two decades has been a source of statistical analysis on the pages of Sports Illustrated, New York Times, and Chicago Tribune. David has written about all seven recent Boston-area championships for Sports Illustrated Presents commemorative issues, was the creator of such long time features as SI’s Player Value Ranking, NBA Player Rating and long running fantasy football and baseball columns.

He has also authored or made contributions to many books, including the Sports Illustrated’s 100 Fenway: A Fascinating First Century.

Now living in Marblehead, he’s focusing his attention on the Boston sports scene, specifically delving into the numbers affecting the Red Sox, Patriots, Celtics and Bruins, with the goal of informing and entertaining real fans. You can follow him on Twitter at @SabinoSports.

More community voices

[an error occurred while processing this directive]

Child in Mind

Corner Kicks

Dirty Old Boston

Mortal Matters

On Deck

TEDx Beacon Street


browse this blog

by category