League Continuity
Posted by Neil Paine on September 7, 2010
Several times in the past, I've looked at what I called "Team Continuity" -- that is, the amount of minutes/possessions/etc. that a team gave to players who had been on their roster the year before. Today, I want to extend the concept to the NBA as a whole and examine league continuity, specifically the 5-year periods since the merger in which the league had the biggest influx of new talent.
How do we measure that, though? Just like with teams, I can look at the percentage of league minutes that were given to players who were still in the NBA 5 years later:
| Year | Y+5 | Total MP | In NBA Y+5 | Pct |
|---|---|---|---|---|
| 1995 | 2000 | 535610 | 383165 | 71.5% |
| 1994 | 1999 | 533810 | 364200 | 68.2% |
| 1996 | 2001 | 574570 | 385047 | 67.0% |
| 1993 | 1998 | 535160 | 356796 | 66.7% |
| 1992 | 1997 | 535310 | 346011 | 64.6% |
| 1987 | 1992 | 455590 | 289844 | 63.6% |
| 1991 | 1996 | 535260 | 339311 | 63.4% |
| 1997 | 2002 | 575170 | 362846 | 63.1% |
| 1990 | 1995 | 534760 | 336529 | 62.9% |
| 2004 | 2009 | 574770 | 358285 | 62.3% |
| 2000 | 2005 | 574270 | 357141 | 62.2% |
| 1985 | 1990 | 455340 | 280490 | 61.6% |
| 1986 | 1991 | 455640 | 278572 | 61.1% |
| 1988 | 1993 | 455040 | 277283 | 60.9% |
| 2005 | 2010 | 595000 | 362081 | 60.9% |
| 1989 | 1994 | 495150 | 300563 | 60.7% |
| 2003 | 2008 | 575420 | 347481 | 60.4% |
| 1998 | 2003 | 575120 | 343240 | 59.7% |
| 2002 | 2007 | 574670 | 342554 | 59.6% |
| 1999 | 2004 | 350650 | 207428 | 59.2% |
| 2001 | 2006 | 575520 | 335154 | 58.2% |
| 1984 | 1989 | 456490 | 260887 | 57.2% |
| 1983 | 1988 | 455090 | 252374 | 55.5% |
| 1982 | 1987 | 455690 | 250565 | 55.0% |
| 1981 | 1986 | 455340 | 237135 | 52.1% |
| 1978 | 1983 | 436160 | 220196 | 50.5% |
| 1980 | 1985 | 436260 | 217396 | 49.8% |
| 1979 | 1984 | 434960 | 213680 | 49.1% |
| 1977 | 1982 | 434810 | 209727 | 48.2% |
In other words, the 1995 NBA had the most 5-year continuity of any season since the merger, as 71.5% of the league's minutes that year were given to players who were still active in the 2000 season. Conversely, the very first postmerger season (1977) had the lowest 5-year continuity -- only 48.2% of NBA minutes that season were given to players who were still around in 1982.
Now, here's another take on the same data, this time looking backward instead of forward... This table gives the percentage of league minutes earned by players who were in the league 5 years earlier:
| Year | Y-5 | Total MP | In NBA Y-5 | Pct |
|---|---|---|---|---|
| 2000 | 1995 | 574270 | 309072 | 53.8% |
| 1999 | 1994 | 350650 | 185699 | 53.0% |
| 2001 | 1996 | 575520 | 303804 | 52.8% |
| 2002 | 1997 | 574670 | 301041 | 52.4% |
| 2004 | 1999 | 574770 | 299010 | 52.0% |
| 2003 | 1998 | 575420 | 293927 | 51.1% |
| 2010 | 2005 | 594498 | 303364 | 51.0% |
| 2005 | 2000 | 595000 | 300877 | 50.6% |
| 1996 | 1991 | 574570 | 289972 | 50.5% |
| 1997 | 1992 | 575170 | 289683 | 50.4% |
| 2009 | 2004 | 594656 | 294468 | 49.5% |
| 2007 | 2002 | 595767 | 294986 | 49.5% |
| 2008 | 2003 | 594114 | 293967 | 49.5% |
| 1998 | 1993 | 575120 | 283731 | 49.3% |
| 1994 | 1989 | 533810 | 260906 | 48.9% |
| 2006 | 2001 | 595555 | 286895 | 48.2% |
| 1995 | 1990 | 535610 | 256656 | 47.9% |
| 1991 | 1986 | 535260 | 256444 | 47.9% |
| 1992 | 1987 | 535310 | 254208 | 47.5% |
| 1990 | 1985 | 534760 | 245933 | 46.0% |
| 1989 | 1984 | 495150 | 225214 | 45.5% |
| 1993 | 1988 | 535160 | 235165 | 43.9% |
| 1988 | 1983 | 455040 | 197407 | 43.4% |
| 1987 | 1982 | 455590 | 196131 | 43.0% |
| 1986 | 1981 | 455640 | 196076 | 43.0% |
| 1984 | 1979 | 456490 | 178340 | 39.1% |
| 1985 | 1980 | 455340 | 174062 | 38.2% |
| 1982 | 1977 | 455690 | 168716 | 37.0% |
| 1983 | 1978 | 455090 | 166874 | 36.7% |
Again, you interpret this by saying that 53.8% of NBA minutes in the 2000 season were given to players who had been active 5 years before, in 1995.
What's the takeaway from this? Well, for one thing, the mid-to-late 1990s (and, to a lesser extent, the 2000s-2010s) were a great time for NBA job security and continuity among players. Both methods show that if you were an active player during that era, there was better than a 50-50 minute-weighted chance you'd be drawing an NBA playcheck 5 years later. As for the reasons why this was the case, anecdotally we saw stars enter the league earlier (often from high school) and enjoy more staying power thanks to modern medical practices that were absent from the league's earlier periods. But perhaps the biggest factor was simply expansion: the NBA went from 23 teams in 1988 to 29 in 1996, providing nearly 100 new jobs to fill (and maintain) that didn't exist in the past.
At the other end of the spectrum, all methods point to the immediate post-merger era as the span when continuity was at its lowest. Obviously new players joined the league directly from the merger in 1977, but this study doesn't even include that break in continuity; instead, it's the influx of players even after the Spurs, Pacers, Nets, & Nuggets were added that's being captured. The standard (read: lazy) explanation would be that the acceptance of the ABA led to an adoption of new playing styles, including a faster pace and more trapping, pressing defenses, which in turn created less continuity as the NBA struggled to find players who could adapt to the changes in play. The problem with that explanation, though, is that it's not even necessarily true: pace actually went significantly down within 5 years of the merger, from 106.5 possessions per game in 1977 to 100.9 in 1982!
So what's your take? Why do you think the league saw such major personnel turnover from 1977-1982/83?
September 7th, 2010 at 10:03 am
That's an interesting question, and I don't really know that we can find one right answer. If you consider the major influx of talent plus the expansion, you might just see a lot of extended reshuffling as GMs try to balance rosters after the initial wave of ABA hirings.
I'd say there were factors off the court as well. I wasn't around back then, but if it's true that Magic and Bird "saved" the NBA, then that must have been the time at which it was closest to dying. Therefore there may have been a lot of turnover because teams were working hard not to have to pay guys, plus whenever things are bad changes in GM and coach are common side-effects that often lead to player turnover.
September 7th, 2010 at 11:46 am
The main reason might be the advances in conditioning, weight training, and medical technology that have been made since then. Look at the number of players whose careers ended or went into serious decline at age 30 back then compared to now. As the NBA became more popular in the 1980s, it also became much more lucrative to play, so there was additional motivation to get the best out of yourself by committing year-round to maintaining your body.
The late '70s/early '80s was also the height of the cocaine era. There's no telling just how many careers were cut short because of drug addiction in those days.
You also have the lingering ABA factor. A ton of guys came into the NBA in 1977 who had played exclusively, or almost exclusively, in the ABA. It took teams awhile to figure out which of those guys could play in the NBA. Some (Dr. J, Artis Gilmore) were no-brainers. Others (Bird Averitt, Darnell Hillman), not so much.
September 7th, 2010 at 3:26 pm
cocaine is a powerful drug.
September 7th, 2010 at 5:43 pm
Yeah, I knew the drug angle would be front and center... I don't think we'll ever be able to pinpoint precisely how many guys fell out of the league simply because of that (vs. just not being good enough), but it's telling that the numbers show the "Cocaine Era" as the time with by far the least continuity. I feel like that's the conventional wisdom on the high personnel turnover, at least.
Also, Jason and Chuck both mentioned this idea of teams needing time to figure out if players could play at the NBA level or not, which makes sense. It would be akin to having a third of today's rookies go off and play in some alternate league that you knew was better than the NBDL but worse than the NBA. How long would it take to figure out who could play and who couldn't? Probably less time than you can tell with guys coming straight out of college, but since we're working with a 5-year window it's going to treat all of the early washouts the same.
September 7th, 2010 at 9:57 pm
There may be something to the ABA and drug theories, but I think much of the difference is simply the flip side of why recent times have had the greatest stability. The length of NBA careers has tended to increase over time, and since 1977-83 is the earliest time period covered in the survey, players from that era had the shortest careers of any period covered in the study. Among the factors (some repetition of items already cited):
--Conditioning, weight training and medical technology have improved over time, allowing players to extend their careers.
--Player salaries have increased dramatically, providing extra incentive to play for a few more years.
--Players now enter the pros at younger ages, theoretically allowing them to tack extra years onto the front ends of their careers (I'm not sure how true this actually is in practice, though).
--There are far more NBA jobs available now than there were in the late '70s and early '80s. To begin with, there were fewer teams back then, and only one new expansion team was added in the 12 years between the merger and 1988. Roster sizes were also smaller. Between a point early in the 1977-78 season and the end of the 1980-81 season, NBA active rosters were reduced to 11, instead of the usual 12. Even after that, through at least the 1984-85 season, teams were still given the option to carry only 11. In addition, whether rosters were 11 or 12, teams in that era did not typically stash extra players on IL on a permanent basis as became the case in the '90s and '00s (when it became common for teams to carry 15 players at all times).
The last point above may actually be the single biggest factor. With the ABA gone, rosters cut to 11, and further expansion on the slow track, I think the late '70s and early '80s was the toughest era for a marginal/aging player to stick in the NBA at any time since the explosion in the number of high-level pro basketball teams in the mid-to-late '60s. (The early-to-mid '80s weren't much easier, although the addition of one expansion team and the restoration of the 12th roster spot helped a little.)
September 9th, 2010 at 9:29 am
Hi... I don't know if you've been making changes, but your pages aren't displaying correctly for me. The margins look all screwed up. I all worked fine the last time I was here. I don't know if it's on my end or if you've made a change... Just thought you might want to look at it. Thanks! Sidney Berrong