In 1996-97 the San Antonio Spurs won 20 games. The next season, with the additions of Tim Duncan and a healthy David Robinson, the Spurs won 56 games. This 36 game improvement stands as the greatest leap forward in NBA history.
Last season the Boston Celtics won 24 games. This off-season the Celtics added Kevin Garnett and Ray Allen. If this team wins 61 games in 2007-08, the new look Celtics will break the record set by the Spurs. Is this likely to happen?
The Early Returns in Boston
After nine games the Celtics are 8-1. When we look at point differential, we see that the Celtics have outscored their opponents by 13.2 points per game. To put that in perspective, the Chicago Bulls of 1995-96 outscored their opponents by 12.2 points per game en route to winning 72 games.
When we look at Boston’s Wins Produced, we see a team that at the moment (and I emphasize the phrase “at the moment”] rivals the Bulls of 95-96. To see this point, consider the top trio on that edition of the Bulls.
Michael Jordan: 24.9 Wins Produced
Dennis Rodman: 18.1 Wins Produced
Scottie Pippen: 15.8 Wins Produced
Combined this trio produced 58.7 wins. With such output from three players it’s not surprising this team won an NBA record 72 games.
As noted, after nine games the Celtics have a better point differential than the fabled Bulls. And like the Bulls, the Celtics in 07-08 are also led by a dominant trio. Here is how many wins the Celtics Trio is on pace to produce after just nine contests:
Kevin Garnett: 31.3 Wins Produced
Paul Pierce: 15.3 Wins Produced
Ray Allen: 13.7 Wins Produced
When we add this together, we see the Celtics Trio is on pace to produce 60.3 wins. And when we look at the entire team, we see a franchise that is on pace to win 76.5 games.
Table One: The Celtics after Nine Game
Now it’s very important to note, this is just after nine games. Let me repeat this point. The Celtics have only played nine games. Such a small sample is not enough to tell us the Celtics in 07-08 are the greatest team in NBA history. Nor would I want anyone to think this team is going to win 77 games.
In fact, one will note that the last two columns in Table One tell a slightly different story. In these columns I look at how many wins the Celtics could expect if each player Boston employs (except for rookies Glen Davis and Gabe Pruitt) perform as they did in 2006-07 (in the minutes allocated in 2007-08). This exercise reveals the Celtics could expect close to 65 wins.
Remember, last year this team only won 24 games. So if these players maintain what they did last year, the Celtics should expect to post the largest improvement in NBA history. No, they wouldn’t be the greatest team in NBA history. But they would be damn good.
Again, thus far this team has collectively done better than what we would expect looking at last year. Where has this improvement come from? One great benefit from looking at the box score data – the Wins Produced way – is that we can see specifically who improved. In other words, we don’t have to depend on such arguments as “the team has just come together.”
Now it could be the case this team has “just come together.” If so, as Table One reports, this “coming together” has manifested itself in the improved play of Kevin Garnett, Ray Allen, and Eddie House. In other words, the “coming together” effect has not really touched the other players on the team.
One suspects that what we see from Eddie House after nine games is not what we will see when the season is over. House has never played this well in the past. Still, even if House, Garnett, and Ray Allen return to what we saw last year, the Celtics – based solely on these players did last year – look like a great team. In fact, they are the class of a weak Eastern Conference.
Boston’s Eastern Competition
Thus far the competition the Celtics face in the East comes from Orlando, Detroit, and maybe the defending conference champion, Cavaliers. Let’s look at how many wins the leading trios on each team are on pace to produce in 07-08.
The Orlando Magic (after 11 games)
Dwight Howard: 25.7 Wins Produced
Jameer Nelson: 12.2 Wins Produced
Rashard Lewis: 7.9 Wins Produced
This trio is on pace to produce 45.8 wins and the team is on track – again after 11 games (not including the last game against the Hornets) – to win 61 games.
The Detroit Pistons (after 10 games)
Chauncey Billups: 12.2 Wins Produced
Tayshaun Prince: 11.6 Wins Produced
Rasheed Wallace: 7.8 Wins Produced
Detroit’s trio is on pace to produce 31.5 wins and the team is on track to win 51 games.
The Cleveland Cavaliers (after 10 games)
LeBron James: 21.1 Wins Produced
Zydrunas Ilgauskas: 13.0 Wins Produced
Daniel Gibson: 6.9 Wins Produced
This trio is on pace to produce 41.0 wins. The team, though, is incredibly weak after these three. The entire team – as their negative point differential indicates – is only on pace to win 35 games.
Brian Windhorst – of the Akron Beacon Journal- argued yesterday that the Cavaliers trio of Lebron, Ilgauskas, and Drew Gooden (on pace for 4.4 Wins Produced) rivaled the Celtics Trio. His argument rested on three stats – points, rebounds, and assists. When we consider all the box score statistics, we see that Cleveland does not have any threesome that comes close to matching Garnett, Pierce, and Allen. And that will also be true when Anderson Varejao – part of Cleveland’s top trio in 2006-07 – finally makes an appearance.
When we look at the early numbers, it looks like the only serious competition the Celtics will face will come from Orlando. And although Orlando handed Boston its first loss on Sunday, right now it looks like Boston is the better team.
So will the Celtics set a record in 2007-08? Again it’s really early. Injuries can happen. But if this team stays healthy, there is a good chance that Boston will beat the record for team improvement set by San Antonio a decade ago. And there is also a good chance that Boston will defeat the Spurs next June to take the 2008 NBA title.
Okay, just to cover all bases, there is also a chance Garnett, Pierce, and/or Allen will get hurt and this whole prediction will go to hell. Remember, the Celtics have only played nine games.
– DJ
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
andrew
November 20, 2007
Dave,
Thanks for looking at the Cavs. I guess I should have assumed gooden wasn’t performing that well, as evidenced by Ira Newble (!) Getting the crunch time minutes at the 4 against Utah on Friday.
Andrew
Ben
November 20, 2007
It’s interesting that (through 9 games) Allen, KG, and Pierce are all playing better than last year. Looks like the “law of diminshing returns” may not be true here. Or maybe it is similar to Tom Brady having a career year because of the addition of Randy Moss (among others). That players can be play more effeciently when playing with effecient teammates. I imagine the same thing happened to Jordan and Pippen with the addition of Rodman.
dberri
November 20, 2007
Ben,
I thought that was interesting also. Diminishing returns is real, but not a very large factor. And as we see here, not always apparent. It is still early, but it will be interesting to see if these players continue to play better together.
Baba O'Riley
November 20, 2007
Devin Harris isn’t on the all-start ballot. This is ridiculous. I’m going to write him in– http://www.blogmaverick.com/2007/11/17/devin-harris-is-not-on-the-all-star-ballot/
Pete
November 20, 2007
Baba,
Good luck with your write-in campaign. I think it’s slightly more likely that you turn into a giraffe than it is that Devin Harris makes the all-star team.
jdawg
November 20, 2007
Dave,
I asked this before, but could you explain how your stuff differs from Dean Oliver’s? It seems to me that your WP48 and his player win % often reach similar conclusions. Also your WP and Win Shares based on DeanOs stuff (available on basketball-reference.com) are pretty close (if you divide Win Shares by 3).
Pete
November 20, 2007
Jdawg:
Is this the Joseph Dawg that co-runs the East Brunswick fantasy basketball league with Mike Lerra? This is my third year in it. Last year I came in 4th place in Cohort C so just missed being in the playoffs. I have a feq questions to ask you off-line.
Baba O'Riley
November 21, 2007
Pete:
I don’t think Devin Harris would make the all-star team but he at least deserves a chance. It’s not fair otherwise. He’s the starting forward for Dallas.
Oren
November 21, 2007
There are two things that catch my eye. One is when you look at Boston’s performance is shooting percentage.
Pierce is shooting 49.6% from the field this year. His career average is 44%. His Free Throw Percentage is the same as his career average.
KG is shooting 56.3% from the field. His average is 49.2%. His FT% is much higher then his career average, but not so much higher then his average for the last four years.
Ray Allen is disappointing. He’s shooting 46.5% and his career average is 44.6%. Of course, 46.5% would be his second most efficient year in his career.
Rondo’s FG% is more then ten percent higher this year then last year, even when controlling for the fact that he hasn’t taken any three pointers this year.
Perkins’s FG% is just less then ten percent higher then his average.
Eddie House is shooting 43.4%, which is 2.5% better then his average. But he’s been shooting around there the last few years.
All of the starters seem to be shooting fewer shots… but they’re keeping up their scoring averages because they’re shooting extremely efficiently.
I know this is only one sample, and there have only been nine games to boot so it’s not even a reliable sample. But I can’t help wonder if this can be used to show that points per game isn’t affected by the law of diminishing returns.
It seems to me that this hints that the better shooters you have on your team, the fewer shots you take and the higher percentage that one makes. The above is obvious, but what isn’t obvious is that this also hints that it evens out to a point where ppg is minimally affected.
Oren
November 21, 2007
I would think that you need to look at the law of diminishing returns category by category.
It seems to me that most of the main players on Boston are shooting more effectively this year then last year. Therefore, there aren’t diminshing returns in points. There are fewer shots per player, but a higher percent of them are made.
However, with the exception of Eddie House and KG, I think every other player on Boston is averaging fewer rebounds then last year. Seems to me like there are diminishing returns there.
Rasta
November 21, 2007
Um, what happened to that Magic article? It’s gone.
Bobbie
November 21, 2007
The celts have played weak competition. It’s not suprising that they are playing well.
vince
November 23, 2007
I,m sure the boston championships is going to improve from 16 to 17.
Lindsey Strickland
December 3, 2007
How many games have the Celts won this season so far? I NEED the info!
Lindsey Strickland
December 3, 2007
I agree with Vince.
Gerry
December 6, 2007
Yeah,… It’s #17 or bust!…