Last week, Ben Golliver from the Blazer's Edge blog asked if I could put together some numbers looking at the theory, popular in Portland, that the Blazers' defensive woes were due in large part to the team's youth. When I looked at the historical data for a post Monday, I was surprised by the strength of the relationship between age and defensive success at the team level. With that conclusion in mind, I decided to take a deeper look at age in the NBA.
Before going any further, I should note that when I talk about team age, I mean effective age--the team's average age weighted by minutes played. When the league calculates average age, it includes everyone on the roster, like Portland's 32-year-old Raef LaFrentz (better known as Raef LaFrentz's Expiring Contract), who has not been with the team this year. While other veterans like Dikembe Mutombo might offer leadership with their presence, it seems silly to count them the same in terms of quantifying a team's age as starters and players in the rotation.
Using that method, here are the league's youngest and oldest teams this season (all 2008-09 stats through Feb. 26):
Youngest Age Oldest Age
Memphis 23.9 San Antonio 31.4
Portland 24.5 Phoenix 30.0
Chicago 25.3 Dallas 29.5
Oklahoma City 25.5 Detroit 29.3
Minnesota 25.7 New Orleans 29.2
The Bulls are really the only surprising entrant to me amongst the youngest teams, taking advantage of starting lineups that have included Derrick Rose (20), Tyrus Thomas (22) and Joakim Noah (24). On the aged side, the urgency for Phoenix and Dallas to make the playoffs this year is obvious, while Detroit's run appears to be nearly complete (at least with this group of players). The Boston Celtics (average age 28.6) are noteworthy in their absence from the second list. The shocking bit of information is just how old the Hornets have become. Because Chris Paul is still very young and David West is in his prime, New Orleans doesn't feel like an old team. However, most of the Hornets' role players are in their 30s, which has to be a concern. (This is also kind of funny in that last year New Orleans was criticized for its lack of experience.)
I calculated effective ages for every team dating back to the 1979-80 season, a span of three full decades. In this context, the Blazers' youth is evident: They are the eighth-youngest team in modern NBA history, with the Grizzlies ranking an impressive third (the 2005-06 Atlanta Hawks and 2002-03 Cleveland Cavaliers being the lone more youthful squads). At the other end of the spectrum, the Spurs check in as the 15th-most senior team in that timespan. It was last year's Spurs that were historically ancient, placing fourth in effective age behind the 2000-01 Utah Jazz and 1997-98 Chicago Bulls and Houston Rockets. This year, George Hill and Roger Mason represent at least something of a youth movement in San Antonio.
One other bit of interesting information is the average age year by year. As I discovered in a column for Hoopsworld.com looking at the impact of the age limit way back in 2004, the counterintuitive result when you look at the data is that despite a trend toward younger players, the league is still older as a whole now than it was during much of the 1980s. From 1982-83 to its high point in 1998-99 (the lockout season), the league's average age went up from 26.8 to 28.6. It quickly dropped to 27.3 in 2005-06, the last year before the age limit was enacted. Since then, average age has been relatively stable around 27.3. Here's how that looks graphically:
Let's return to the team level and look at how team age relates to offensive and defensive performance. My suspicion going in was that age would be more meaningful at the defensive end of the floor, where experience is theoretically a bigger factor. On offense, pure talent can overcome youthful mistakes. Alas, when you take a look at the numbers, you find isn't the case.
Note that this chart shows Offensive and Defensive Ratings adjusted for league, so that 100 is average and higher is better for both offense and defense. Both sides of the floor show a slightly positive trend. To my naked eye, it seems like offense is a little more variable. However, the correlations are virtually the same at either end of the floor (age with Offensive Rating: .382, Defensive Rating .388). It appears that the more tangible benefits of experience on offense (improved shooting, fewer turnovers, etc.) have as much impact as experience does at the defensive end.
Now let's put the two sides of the floor together and look at overall team performance by effective age.
Here, the correlation appears to be much stronger, and the math supports that conclusion. The correlation between team age and winning percentage is .515, which is reasonably strong (a correlation of 1 or -1 means two variables are perfectly in sync, while a correlation of 0 means no relationship whatsoever). There simply aren't a lot of teams in the upper left quadrant (good young teams) or in the bottom right (bad veteran teams). Apparently this trend holds even over a much longer sample, according to a similar study by poster PoliSam on Blazer's Edge that I came across after completing this work.
One notable outlier in the former category? The Portland Trail Blazers, natch. They're represented by the lone blue dot above the .600 line to the far left. In the big picture, Portland's success with such a young team is remarkable. To find the next-youngest team which won more than half of its games (two very young teams, including last year's Blazers, were exactly .500), you have to go all the way to the 1981-82 Blazers, 1982-83 Kansas City Kings and 1997-98 Cavaliers, all of them with an average age of 25.4--nearly a year older than this year's Portland squad.
Using this data, we can project the average record for teams of a given age. Every year of age, all other things equal, improves a team's winning percentage by .050--the equivalent of four wins over the course of the season. The difference between the expected record for the venerable 2001-01 Utah Jazz, given the team's age, is a full 38 games better than that of the youthful 2005-06 Hawks.
We can then look at the teams that have most overachieved and underachieved, given their age. Naturally, the former list includes the Blazers, though in a surprise it's another 2008-09 team that leads the list.
Year Team Age Win%
2009 L.A. Lakers 27.7 .828
2005 Phoenix 26.6 .756
2009 Cleveland 27.2 .786
1998 L.A. Lakers 26.4 .744
1983 Philadelphia 27.5 .793
1985 Milwaukee 26.3 .720
1992 Chicago 28.3 .817
1980 L.A. Lakers 26.9 .732
2009 Portland 24.5 .614
1981 Boston 27.4 .756
Yes, the 2008-09 L.A. Lakers are the greatest proof that age is nothing but a number. With a weighted team age of 27.7, the Lakers are barely above league average, yet they continue to win at an 80-percent-plus clip. Cleveland is also overachieving its age, given the Cavaliers as a whole are slightly younger than league average. Most of the teams on this list tend to fall into the same category--teams of average age and uncommon success. Four of the seven pre-2009 teams on the list ended the season with a championship. Portland, then, is a bit of an oddity. No other young team even comes close.
Year Team Age Win%
1983 Houston 29.5 .171
1997 San Antonio 30.6 .244
1998 Denver 27.2 .134
1996 Vancouver 28.0 .183
1991 Dallas 31.2 .341
2008 Miami 27.8 .183
1999 Chicago 29.2 .260
2002 New York 31.0 .366
1992 Minnesota 27.1 .183
Congrats to the 1982-83 Rockets, the most terrible old team in NBA history. There are two different types of teams on this list--those like Houston who saw successful runs come to an abrupt end and rebuilding teams with too few youngsters in the lineup. The 1996-97 Spurs have a good excuse, a variety of injuries (most notably to David Robinson) that put the team in position to draft Tim Duncan. This year's greatest underachievers in terms of age are the Sacramento Kings, at 27.5 slightly older than league average with a record that puts them firmly in the lottery.
Now, all of this needs to be understood in context. The data shows a correlation between age and team success, but this does not mean the former leads to the latter. Going out and signing a 40-year-old free agent and putting him in the lineup won't improve your team. Much of the explanation for the relationship between age and success comes from self-selection--teams competing for a championship will fill out their bench with veterans looking for a ring, while younger teams are giving heavy minutes to youngsters in the name of player development. The strength of the team often determines its age, not the other way around.
There are still some valuable pieces of information. Again, I was surprised by the strength of the relationship between team performance and team age. Because of self-selection and the development of individual players through peak age, I anticipated older teams would fare better, but not as dramatically as was the case. I also anticipated finding a point at which additional age was a negative factor. Going all the way back to the 1950s, the study by PoliSam on Blazer's Edge did find such some evidence of diminishing returns to age, but it does not show up in the data over the last three decades. The ten oldest teams in that span averaged a .622 winning percentage and include a champion (the 1997-98 Chicago Bulls).
The bottom line is a conclusion that hardly needs to be quantified: A team's performance must be understood in the context of its age. When a young team struggles, it is to be expected. When a veteran team can't put things together, it's time to rebuild. And when a young team, like the Blazers squad that inspired this study, enjoys precocious success, it's certainly worth taking notice.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.