Review of cutting down inefficient 3-point habits

Final idea with last dataset and posts inspired by certain players is presented to you thanks to LeBron James.
Before this season he attempted over 2500 three-point shots and made 33% of them which considering his ability to get to the rim seems like a waste of possessions. He reportedly vowed to cut down on long ball this season and so far it holds true: right now he averages only around 2 three-point attempts per game with a career mark over 4. It got me thinking, will he be able to keep it up? What are his chances?
Which other players did try to cut down inefficient three-point habits? Have they succeed or failed?

Let’s try to find out, I started again with a file from basketballreference.com and then I searched for pair of seasons next to each other where:
– player played at least 1000 minutes in both years,
– he attempted at least 70 threes in a previous season,
– in both seasons his career three point percentage was below 33,34%,
– player’s career rate of three point attempts per 36 minutes dropped at least by 6% from minimum 1,2.
[It will happen to LeBron if he continues current pace]

I used to-date-career point of view because I was interested in career’s path changes and I wanted to weed out as many fluke seasons as possible especially those at the beginning of player’s career or those which happened mostly due to injuries.

Also because list wasn’t long and I was interested in multiple things at once I decided to review each of them individually. Here are the results [in alphabetical order of player’s surname]:

In 1993 Mahmoud Abdul-rauf (his 4th season) lowered his career rate from 2,45 to 2,25
He returned to the old habit next season but to his credit in 5th season he improved efficiency to 38+% so maybe it was a transitional year for him? Interestingly his eFG% was at the same level in both seasons.

In 1983 Mark Aguirre (his 3rd season) lowered his career rate from 1,24 to 1,02
He stayed at this level for the next 4 years and then he got back to old habit.
Note that his three best eFG% seasons happened during that window.

In 2004 Carmelo Anthony (his 2nd season) lowered his career rate from 2,57 to 2,39
In 2005 Carmelo Anthony (his 3rd season) lowered his career rate from 2,39 to 2,21
Those two seasons has become the lowest marks in his career so far so this change didn’t last long.
To his credit in 5th season he improved efficiency to 35% though 2005 is still his 2nd best season in eFG%.

 

Investigation of recent snubs in NBA rotations

Recently I ranted on Twitter about Ryan Anderson’s playing time because virtually all advanced metrics suggest he should play more yet for some reason he doesn’t. It got me thinking about players in those situations.
How often does it happen at similar level? What does happen to those players in following years?
In other words, I wanted automated version of the strangest rotation decisions.

To investigate I used excellent set of data collected by Alex at sportskeptic.wordpress.com with multiple advanced metrics in one file and a following method: for each team in the last 5 years I sorted players based on average minutes played [with minimum of 1000 total] and their position according to five different metrics [PER, Win Shares, Wins Produced, new RAPM, ASPM – btw, those were the only ones with data for each of last 5 regular season]. Then I simply compared average ranking for all metrics to the ranking based on average minutes and I focused on the Top10 players in each season to find possible explanations.

Let’s start this excercise with the 2010/11 season…

10) JaVale McGee, ranked 6th in minutes per game but 2.2 according to average for 5 advanced metrics.
Two big guys averaged more minutes – Rashard Lewis and Andray Blatche… so it simply means Wizards viewed Blatche as a better option at center than McGee. Strange… but still true because in current season average minutes suggest the same thing and it’s not like McGee has too many fouls…

9) Amir Johnson, ranked 5th in minutes per game but 1st[!] according to PER, Win Shares, Wins Produced, new RAPM and ASPM.
He had [and still has] some problems with fouls so it can’t be blamed entirely on coaches’ decisions… but he also may be the victim of foul trouble treatment. Just play him in every game until he collects 6 fouls and I won’t complain about it ;-)

 

NBA History of Blowouts, New Record in 2011-12?

On December 28th in one of the e-mails my friend Leszczur mentioned “we are going to see a lot of random NBA scores this year” to which I automatically replied “every year there are blowouts” but it immediately got me thinking about them. Will this season actually lead to more blowouts than usual?
What exactly is the usual number? What happened in the previous shortened season?.
I’ve waited a week or so for more data and here we are…

I defined blowout as 15 or more point differential and using all scores in the NBA history from basketball-reference.com I compiled a historical perspective… note that 2012 includes only 125 games played so far.

 

Journey through Miles of NBA Travel History

In my last post about Distance Traveled by NBA Teams during a Season I offered an innocent curiosity… but quite typically for me it lead to more questions plus more serious questions… and man, I went big this time.

Instinctively I wrote about Portland’s loneliness on the US map with recent team’s relocations from Seattle and Vancouver and then I hit me… wait a second, I can check the influence of those changes!
Average journey for each NBA team in a single season? That’s nice… but how about the whole history?
I could finally settle the issue how amount of travel actually has changed through league’s history!

So while I was fired up I got my hands dirty, this time with schedules from basketball-reference.com
[but I still used latitude and longitude of U.S. and Canadian Cities from infoplease.com and an algorithm converting it into distance in miles from meridianworlddata.com] but before we look at the results a couple of technical notes which you should keep in mind:

– I totally ignored random games in temporary locations but I included changes which lasted at least a season. So for example, Nets spent one year in New York so I accounted for that in calculations but if the same Nets played a game in Kansas due to some difficulties I assumed they played at home. In recent decades this point probably doesn’t matter but in early days of the NBA it could have some impact.

– I totally ignored exhibition games like those in London or Tokyo and I assumed they were played in team’s arena. There were few of them and I prefered to save some time than play with such details.

– This batch of numbers in some cases is slightly different than in my previous post because there I used scheduled games and here I used games when they actually happened [so any postponed games made a tiny difference here].

OK, that’s enough, here’s what I dug out from that goldmine of information so far…

Let’s start with a graph of Total Journeys by NBA Teams Over 800 Miles per Season [which is easier to digest version of Total Miles Traveled per team because there’s a 0.99 correlation between those two]:

 

Distance Traveled by NBA Teams during a Season

Have you ever wondered how many miles teams/players travel during typical NBA season?
If you have… well, you can join this one-member club! ;-)

Because even though it probably doesn’t matter now (with private jets and great service provided to the players can it have any effect on their performance?) it’s a trivia which has made me curious for some time…

So to calculate average distance traveled by NBA teams during a typical season I used schedule from insidehoops.com, latitude and longitude of U.S. and Canadian Cities from infoplease.com and an algorithm converting it into distance in miles from meridianworlddata.com. Additionally, I wanted to include some information about long-ish journeys so based on distribution I set a bar at 800 miles.

Results for the last 2 years are…
1.344.475 miles traveled by all NBA Teams in 2010/11 [44816 miles per team] and
1.360.641 miles traveled by all NBA Teams in 2009/10 [45355 miles per team].
To put it in context, every team travels around the world 11,5 times per season… and it’s without playoffs!

Here’s a more detailed breakdown for each team [numbers are rounded to the integer]…

 

20-Year History of Missed Games in the NBA

Because of a lockout this year there will be an unusually packed schedule in the NBA which undoubtedly contributed to a theory that we are going to see more injuries this season. Don’t get me wrong, I think this idea sounds plausible because… well, I know how I feel after playing basketball too often in short period (muscle injury or ligament tear waiting to happen). But could it be also a reason why we overrate this issue? I’m not an athlete and I don’t usually play any back-to-backs so maybe that’s why I feel this way when I do…

Beside this point, have you ever wondered do players miss more games now than they used to?
Which teams’ players missed the most and least games in the last decade?

Well, I have so here’s how I tackled those topics:
I considered counting injuries at prosportstransactions.com but there’s one huge problem with this approach:
5 or even 10 day-to-day injuries are usually less significant than one semi-serious knee injury because of time needed for healing and rehabilitation. So why not try to measure the main effect of injuries which is missed games? That’s exactly what I’ve tried to accomplish. I downloaded data from last 20 seasons from dougstats.com, then in each season I sorted it by team’s name and then by players’ minutes per game.
I deleted all players who were 7th or lower on their own team and counted how many games such group played versus how many games they would have played given full 82 games schedule [or 50 in 1998-99].
My rationale here is “let’s count starters and 6th man, the rest doesn’t matter”.

In other words, for players considered the most important by their own teams what percentage of available regular season games did they miss?

Unfortunately it’s not a perfect tool, for example it totally ignores tanking or players who missed entire season because they were injured in the offseason, so if you have a better idea let me know in the comments.

With all that in mind here are the results:

 

Average NBA Career Length for Players – Details

Thanks to the lockout you’ve probably read or heard how long average career of NBA player really is, and you can easily google it but… it’s not up-to-date information and there are very few details about it…

For example, how average NBA career length has changed through league’s history?
Is it so much longer now than it used to be? How exactly is it calculated?
How does this average career span differs for NBA starters and deep reserves?

I’ll answer those and many more questions in this post… but let’s start with the description. I collected data from basketball-reference.com for every season from 1946-47 to 2010-11, then I removed multiple occurrences in one season [usually due to trades] and finally counted how many times player appeared on the list.
Simple, right? Such sequence gave me an average NBA career length of 4.869 seasons for 3668 players.

But you’ve already knew that so let’s begin the fun part…