1. Welcome! Please take a few seconds to create your free account to post threads, make some friends, remove a few ads while surfing and much more. ClutchFans has been bringing fans together to talk Houston Sports since 1996. Join us!

Why our Players are FAR from their peaks -- Statistical Analysis and Hypothesis

Discussion in 'Houston Rockets: Game Action & Roster Moves' started by torocan, Dec 3, 2012.

  1. torocan

    torocan Member

    Joined:
    Oct 15, 2012
    Messages:
    4,228
    Likes Received:
    436
    Ok, so there's a lot of posts flying around talking about how some of our rookies are garbage, others are at their peak, or are going to regress, or are limited.

    The fact is, our team is the Youngest in the league. However, what DOES that mean?

    Well, it means that they're at the earliest stages of their skills curve. Not only will nearly ALL of them improve (barring injury), but they should all improve Substantially for several more Years.

    This kind gentleman was kind enough to compile the data from the 10,000 player seasons since the introduction of the 3 point line, and then crunch each player's career performance as they aged.

    He then broke them down into the following 4 categories of age.


    "very young" (18-21)
    "young" (22-25)
    "prime" (26-29)
    "old" (30+)

    And synthesized them into a chart on this blog page.

    http://www.d3coder.com/thecity/2012...ce-metric-to-create-aging-curves-for-the-nba/

    A quick summary of the conclusions...

    -Young players develop until the age of 26 where they typically first hit their "peak"
    -Players then decline on average, and then his a SECOND peak at age 30
    -Players who enter the NBA at 18 tend to peak at age 21

    Here is the current age of our roster...

    Aldrich - 24
    Asik - 26
    Cook - 25
    Delfino - 30
    Douglas - 26
    Harden - 23
    Jones - 20
    Lin - 24
    Machado - 22
    Morris - 23
    Montejunas - 22
    Parsons - 24
    Patterson - 23
    Smith - 21
    White - 21

    Now, the unusual distribution of Peaks (21, 26 and 30) begs several questions. Why do players peak at different points?

    My hypothesis is that it works in the following way...

    1) The first peak a player undergoes directly related to experience. Players who first enter are directly gaining skills and basic understanding of the game. They are learning the rules, learning to habitualize the play sets, working on their glaring weaknesses, and becoming basically familiar with their team mates and opponents.

    2) The second peak I suspect is more linked to declines in athleticism. The average NBA career lasts 5-6 seasons. A substantial number of players just simply wear out as their bodies decline. Those who can last longer either have games that reduce the amount of wear and tear on their bodies, avoid catastrophic injury, are smaller in size (easier on the joints), but more importantly are the ones who can ADJUST to their changing bodies.

    In order for players to sustain a long career, at some point players must ALTER their games to be able to perform WITHOUT relying on pure athletic ability. Athletic jumpers learn to shoot and play post up. Quick and speedy guards learn to better judge angles and spacing as well as develop more of a range game. Quick on ball defenders learn to become crafty so they don't need to rely so much on lateral movement.

    Here's the link in term of average career length in the NBA.

    http://weaksideawareness.wordpress.com/2011/11/22/average-nba-career-length-for-players-details/

    Anyway, long story short, I believe we haven't seen anywhere near the best of the Rockets. And I fully expect the vast majority of the players on this team who ARE able to make the transition to the NBA, and more specifically our Starters to develop significantly over the next 2-3 years.

    We may very well grow into a very formidable team, and that's before we add another Star to the puzzle.

    No matter how you cut it, it's going to be a heck of a ride. :grin:
     
    2 people like this.
  2. douglasreedy1

    douglasreedy1 Member

    Joined:
    Jul 1, 2002
    Messages:
    463
    Likes Received:
    46
    Good read. I'm intrigued by the positions in sports like center, pitcher, and goalkeeper that seem to last longer than others. With regards to asik, I'm ecstatic watching him develop at a rapid pace.
     
    1 person likes this.
  3. heypartner

    heypartner Contributing Member

    Joined:
    Oct 27, 1999
    Messages:
    62,496
    Likes Received:
    56,074
    Cool article. thanks for posting.

    There is one flaw in your thinking, though: Some of our players won't even remain in the league to see their "peak" age of 26. They will get cut. So you can't really say all our young players will improve. Yeah, their improvement curve might go up, but not as fast as the people competing for their job.

    Also, the data from that site is only based on scoring metrics. No other criteria is used, it appears.
     
    1 person likes this.
  4. lookabove

    lookabove Contributing Member

    Joined:
    Jul 25, 2012
    Messages:
    621
    Likes Received:
    17
    So who doesn't make the peak?

    White - Can't get over his disorder.
    Aldrich - bleh
    Harden - Retires and create the first franchise strip club corp at 25.
    TD - Already peaked yikes!
    Delfino - Already in the old group.
     
  5. torocan

    torocan Member

    Joined:
    Oct 15, 2012
    Messages:
    4,228
    Likes Received:
    436
    I did say that it would depend on IF they could stay in the NBA.

    As for the emphasis on scoring metrics, I do agree it's a specific metric however it is also a universal metric. Using a metric like Assists would create extremely uneven results given the impact of position and role, while measuring Defense is very difficult from a statistical point of view due to the complexity of Defense.

    However, as scoring improvement is heavily skill based (IE, it's hard to argue everyone's jump shot will improve because they bench press more), I'm reasonably confident there's a similar correlation to other skill based efforts such as defense, reduction of turnovers, etc.
     
  6. beta ticket

    beta ticket Member

    Joined:
    Sep 19, 2012
    Messages:
    82
    Likes Received:
    3
    Wow, a good read. It's always great to read an analytical posting in a sports fan blog.
     
    1 person likes this.
  7. just a word

    just a word Member

    Joined:
    Nov 14, 2012
    Messages:
    951
    Likes Received:
    39
    This is part of the reason why this Rockets team is so exciting to me even when they lose. There's only upside.

    AND, it's upside on a team that Wages to Wins has recently predicted will get 7th's seed in the playoffs, taking into account strength of schedule and strength of opponent, and then adjusted for season schedule. :grin: Playoffs against the OKC?

    Quote from the site: "Houston is very intriguing given their youth. This is looking to very possibly be a very good team for a very long time."
     
    1 person likes this.
  8. prs325

    prs325 Member

    Joined:
    Jun 25, 2002
    Messages:
    1,104
    Likes Received:
    583
    Or if you're Luther Head...
     
    1 person likes this.
  9. Entropy

    Entropy Member

    Joined:
    Jun 30, 2002
    Messages:
    5,039
    Likes Received:
    1,241
    Golden State not in the playoffs, despite the way they've been playing?
     
  10. Mr. Clutch

    Mr. Clutch Contributing Member

    Joined:
    Nov 8, 2002
    Messages:
    46,550
    Likes Received:
    6,131
    Yes, this is the most intriguing aspect of our team.

    A lot of our players are very good, but they haven't even peaked yet.
     
  11. sidestep

    sidestep Member

    Joined:
    Jul 30, 2012
    Messages:
    777
    Likes Received:
    35
    For a brief comparison of the relationship between age and performance in baseball and bball, I'll paste from Berri's book Stumbling on Wins.
    The focus of this section is on the performance peak in bball. See spoiler.

    The analysis of basketball requires some context. J.C. Bradbury
    examined 86 years of data in baseball and found that the peak performance
    for a hitter or pitcher occurred at about 29 years of age. Although
    the peak for a baseball player’s overall performance occurred in his late
    twenties, the peak with respect to specific tasks is different. A player’s
    ability to draw a walk—or for a pitcher to prevent walks—peaks around
    32 years of age. In contrast, a pitcher’s ability to throw strike-outs peaks
    at about 24 years of age. These differences are explained by noting that
    the location of the peak is related to the physical skills an athlete
    requires. Strike-outs primarily depend on pitch speed, or the physical
    skills of the pitcher. In contrast, a batter needs plate discipline to draw
    walks, discipline that improves as a batter gains more experience.

    The study of hitters and pitchers can be related to a study of the
    age at which athletes set records in track and field, swimming, tennis,
    and golf. In sports that depended more on pure physical ability—
    such as swimming, short-distance running, and tennis—performance
    appears to peak in the early-to-mid twenties. When a sport depends
    more on experience and knowledge—such as in long-distance running
    and golf—the peak is closer to 30 years of age.

    Is basketball like tennis or golf? In other words, does success in
    basketball depend more on physical skills or experience and knowledge?
    Although veteran players like to believe the latter, anyone
    watching an NBA game must be impressed by the rare physical gifts
    it takes to play in the Association. The study of performance and age
    confirms this impression. Performance in the NBA appears to hit its
    maximum at 24.4 years of age.

    Although performance peaks around 24, the drop-off in performance
    is not dramatic until a player approaches 30 years of age. To
    illustrate, let’s imagine a team employs an average player at 24 years
    of age, or a player with a 0.100 WP48. Table 8.3 reports what one can
    expect will happen22 if nothing else changes (i.e., all other factors that
    impact performance are unchanged) as he ages.

    At the age of 25—if nothing else changes—his WP48 will be
    essentially the same. After this point, though, performance starts to
    decline. By the age of 30, our average player at 24 is now posting a
    0.071 WP48. At the age of 32, this value is now down to 0.046; and at
    35 his performance has dipped into the negative range.

    Now let’s go back to the study of how coaches allocate minutes.
    As noted in the previous chapter, a player’s minutes will peak at about
    28 years of age. This suggests that coaches not only have problems
    allocating minutes—remember the focus on scoring—coaches in
    the NBA are also confusing basketball with baseball. The study of
    aging and performance suggests there is not much of a penalty
    imposed for thinking a player at 28 is close to his peak. After all, what
    an average player will do at 28 (WP48 of 0.088) is not far from the
    actual peak. However, one suspects that teams might think that the
    difference between 24 and 28 would be similar to the difference
    between 28 and 32, and that doesn’t appear to be true. An average
    basketball player in his early 30s is well past his peak and can expect
    to offer much less with each passing year.

    TABLE 8.3 The Impact of Age on the Performance of an Average
    NBA Player—Peak Age Is 24 Years

    Age WP48 (Wins Produced per 48 minutes)
    18 0.061
    19 0.072
    20 0.081
    21 0.089
    22 0.095
    23 0.098
    24 0.100
    25 0.100
    26 0.098
    27 0.094
    28 0.088
    29 0.080
    30 0.071
    31 0.059
    32 0.046
    33 0.030
    34 0.013
    35 –0.006

    To drive this point home, let’s compare the number of players still
    employed in the NBA and MLB after age 30. About 30% of baseball
    players are between 30 and 34 years of age, and 10% of baseball
    seasons are played by players who are 35 or older. In contrast, only
    20% of basketball players are in the 30 to 34 age group. And only 4%
    of basketball seasons played are logged by players 35 or older. Given
    how quickly players exit basketball after the age of 28, it seems
    unlikely that this age is the peak on the hardwood.
     
    #11 sidestep, Dec 3, 2012
    Last edited: Dec 3, 2012
    1 person likes this.
  12. LCAhmed

    LCAhmed Contributing Member

    Joined:
    Nov 10, 2009
    Messages:
    11,034
    Likes Received:
    1,631
    Rockets will be Optimal Prime in 4 or 5 years ;)
     
  13. Ynnis888

    Ynnis888 Member

    Joined:
    Aug 11, 2012
    Messages:
    3,466
    Likes Received:
    55
    Interesting.
     
  14. Akim523

    Akim523 Member

    Joined:
    May 17, 2008
    Messages:
    3,733
    Likes Received:
    162
    Good read, as long as our young core keeps improving in content
     
  15. IzakDavid13

    IzakDavid13 Contributing Member

    Joined:
    Jan 2, 2011
    Messages:
    9,958
    Likes Received:
    801
    Interesting that Toney Douglas is in his Prime...scary.
     
  16. jtr

    jtr Contributing Member

    Joined:
    Dec 4, 2011
    Messages:
    7,470
    Likes Received:
    275
    torocan, I looked at the cited article and started tracking back the articles in that post. I do not have the interest needed to put in the time to understand the math behind the cited statistical model. Given the number of assumptions in the model presented, the conclusions are probably accurate to within 1 - 2 years. I mentally assigned a large margin of error, ie a LCI and it is good enough I guess.

    I do not understand your conclusions about the "mini" peaks in a players career at 21, 26 and 30 years of age. Perhaps I am not intuitive enough. I guess one statisticians "peaks" are another statisticians "small sample size - within the margin of error".

    But plugging the starters into a basic excel document and applying the projections in the cited article's "Distance is the scoring metric developed in a previous post" graph it looks to me like they peak in 2014 - 2015 time frame and then hold steady until after 2018 when they start to decline. The decline will be slow until the latter half of 2019 and then will accelerate rapidly.
     
    #16 jtr, Dec 3, 2012
    Last edited: Dec 3, 2012
  17. dachuda86

    dachuda86 Member

    Joined:
    May 3, 2008
    Messages:
    16,308
    Likes Received:
    3,580
    peaks for bigs happen later than guards... think about it and we're in even better shape.
     
  18. jocar

    jocar Member

    Joined:
    Nov 10, 2007
    Messages:
    4,869
    Likes Received:
    614

    [​IMG]
     
  19. JK1

    JK1 Member

    Joined:
    Nov 8, 2012
    Messages:
    146
    Likes Received:
    6
    As always, good stuff Torocan. Rockets gonna shock the world in 2015 :), lol
     
  20. Honey Bear

    Honey Bear Contributing Member

    Joined:
    May 8, 2006
    Messages:
    5,102
    Likes Received:
    555
    OK but why did jadakiss cry? Have a report on my desk by 5 pronto. This is groundbreaking analysis and I'm blessed to be a part of it.
     

Share This Page

  • About ClutchFans

    Since 1996, ClutchFans has been loud and proud covering the Houston Rockets, helping set an industry standard for team fan sites. The forums have been a home for Houston sports fans as well as basketball fanatics around the globe.

  • Support ClutchFans!

    If you find that ClutchFans is a valuable resource for you, please consider becoming a Supporting Member. Supporting Members can upload photos and attachments directly to their posts, customize their user title and more. Gold Supporters see zero ads!


    Upgrade Now