Like in every other sport, NBA players tend to come into the league, develop their game to a certain peak point, then slowly decline until the point that they decide to retire. That long career arc can be seen many times and is not a subject which needs a great deal of debate currently about its general validity. From year to year, however, how well do player's games hold up? When they reach a peak versus the rest of the league's ability do they decline or continue to improve the next year? That's our question for the day. To limit our scope so this may be covered in a single article (there is alot stuff inherent to this discussion that could be expanded, but I feel there are better issues to talk about in the coming days), we will look at how well over the last 10 years players who finished in the league top 10 in PER (Player Efficiency Rating) fared their next season. So without further ado, let's begin.Over the course of the last 10 years many great players of the 90's faded away into retirement and others rose to the top of the heap and have now begun their downfall. The average NBA Top 10 PER over this time has lost approximately .62 points from their PER in the following season, meaning they have gotten slightly worse; however, this number is can be very misleading. When player's are posting extremely high PERs (25+ in general) it becomes extrememly hard to continue to improve your performance while there is limitless free fall potential for getting worse. Numerous players had a drop of over 5 PER points from one year to the next for various reasons (Shawn Kemp won the contest with a drop 0f 6.4 points from 99 to 00). Meanwhile very few players were able to post substantial increases in their PER after posting a Top 10 PER the season before (Tracy McGrady went up 5.2 points from one season to the next while in Orlando). This leads to what I feel is an unfair downward skewing which I don't feel represents how they players truly react from one year to the next. There is a way to cure this though.
If we simply add up the total number of people who improved or lowered their PER from one year to the next the result is extremely interesting. 47 players over the last 10 years got worse, 49 got better, 1 stayed the same, and 3 retired or spent much of the year injured (Amare Stoudamire, Grant Hill, and Michael Jordan). What this tells the statistician is that on average about half the players will get better and half will get worse, so trying to predict how well they will play next year is a toss-up on this information alone. It is necessary to look at the more unique circumstances of each player then to truly decide how likely it is they will improve or lose some of their individual ability. The gathered PER numbers can tell us some interesting things besides simply saying whether a player will get better of worse though.
-Michael Jordan was the only player to post a top 10 PER and then retire the next season. When he returned to play for the Wizards his PER dropped over 8 points from his last Bull's season.-Shaquille O'Neal in a true showing of his dominance was the top PER player for five straight seasons (97-97 to 01-02) before falling behind Tracy McGrady by 1 point in 02-03.
-The Top PER player has only won 3 championships during the last 10 years, all of those championships went to Shaquille O'Neal who posted 30 plus PERs in all of those season, an amazing feat.
-The Top PER player for a given year has only won one MVP, Kevin Garnett in (03-04). The fact that Shaquille O'Neal has never won an MVP award should be a blight upon every writer who votes for the award.
-The Detroit Pistons are the only team in the last 10 years to win an NBA title without a Top 5 PER player in their championship season. Star players do matter.
-One of the most amazing PER feats, something to which no parallel can be drawn, is Lebron James' 28.1 PER in his third season at age 21 (2nd in the league). No player in the last 10 years has come close to posting a PER that good, that young. Tracy McGrady at age 24 posted a 30.1 during his sixth year in the league.
There are a limitless number of observations of the data that can be made, but the one thing that is clear that we set out to determine was that from year to year more variables must be taken into account when determining if a player will improve than if they were really, really good the year before. Age, motivation, team situation, nagging injuries, and athletic potential
all likely play a role (I would like to test this data with some of these factors taken into account, but alas I have not been able to do so yet) in whether a top tier player will continue to improve from year to year. What is clear though is that there is no simple, easy way of making the determination based only on singular past results with any accuracy. It's always nice when things turn out simple, but that usually means that what they tell is not very valuable. If anyone would like to further analyze all these situations just hit up Basketball Reference and drop me a line at tmx117@gmail.com to tell me what you found. Until then, I'm back at the drawing board applying what we learned about the NBA the past few weeks to the NFL for tomorrow.
